| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9983434566537824, | |
| "eval_steps": 500, | |
| "global_step": 452, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0022087244616234127, | |
| "grad_norm": 5.535903453826904, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 6.0457, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.004417448923246825, | |
| "grad_norm": 6.602200984954834, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 8.2904, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.006626173384870237, | |
| "grad_norm": 6.935258865356445, | |
| "learning_rate": 3e-06, | |
| "loss": 8.4242, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.00883489784649365, | |
| "grad_norm": 6.811558723449707, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 8.2953, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.011043622308117063, | |
| "grad_norm": 6.635751247406006, | |
| "learning_rate": 5e-06, | |
| "loss": 8.186, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.013252346769740474, | |
| "grad_norm": 7.359472751617432, | |
| "learning_rate": 6e-06, | |
| "loss": 8.2423, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.015461071231363888, | |
| "grad_norm": 7.092770099639893, | |
| "learning_rate": 7.000000000000001e-06, | |
| "loss": 8.6333, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0176697956929873, | |
| "grad_norm": 7.17343282699585, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 8.2698, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.019878520154610713, | |
| "grad_norm": 7.231985092163086, | |
| "learning_rate": 9e-06, | |
| "loss": 8.0394, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.022087244616234125, | |
| "grad_norm": 7.38832426071167, | |
| "learning_rate": 1e-05, | |
| "loss": 8.3605, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.024295969077857537, | |
| "grad_norm": 7.6347551345825195, | |
| "learning_rate": 1.1000000000000001e-05, | |
| "loss": 8.4294, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.02650469353948095, | |
| "grad_norm": 7.681893825531006, | |
| "learning_rate": 1.2e-05, | |
| "loss": 8.2903, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.02871341800110436, | |
| "grad_norm": 7.411649227142334, | |
| "learning_rate": 1.3000000000000001e-05, | |
| "loss": 8.5525, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.030922142462727776, | |
| "grad_norm": 7.768363952636719, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 8.7325, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.03313086692435119, | |
| "grad_norm": 8.075197219848633, | |
| "learning_rate": 1.5e-05, | |
| "loss": 8.48, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0353395913859746, | |
| "grad_norm": 7.757429122924805, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 8.3817, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.03754831584759801, | |
| "grad_norm": 8.141379356384277, | |
| "learning_rate": 1.7000000000000003e-05, | |
| "loss": 8.5635, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.03975704030922143, | |
| "grad_norm": 8.649480819702148, | |
| "learning_rate": 1.8e-05, | |
| "loss": 8.6707, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.041965764770844835, | |
| "grad_norm": 8.826072692871094, | |
| "learning_rate": 1.9e-05, | |
| "loss": 8.8567, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.04417448923246825, | |
| "grad_norm": 8.904451370239258, | |
| "learning_rate": 2e-05, | |
| "loss": 8.5523, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.04638321369409166, | |
| "grad_norm": 9.640382766723633, | |
| "learning_rate": 2.1e-05, | |
| "loss": 8.3306, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.048591938155715074, | |
| "grad_norm": 9.720635414123535, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 8.4766, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.05080066261733849, | |
| "grad_norm": 10.471412658691406, | |
| "learning_rate": 2.3000000000000003e-05, | |
| "loss": 8.4523, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0530093870789619, | |
| "grad_norm": 10.258910179138184, | |
| "learning_rate": 2.4e-05, | |
| "loss": 8.5165, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.05521811154058531, | |
| "grad_norm": 10.779662132263184, | |
| "learning_rate": 2.5e-05, | |
| "loss": 8.4274, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.05742683600220872, | |
| "grad_norm": 10.864973068237305, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 8.1579, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.05963556046383214, | |
| "grad_norm": 11.23272705078125, | |
| "learning_rate": 2.7000000000000002e-05, | |
| "loss": 8.3138, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.06184428492545555, | |
| "grad_norm": 11.67050838470459, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 8.3039, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.06405300938707896, | |
| "grad_norm": 11.947738647460938, | |
| "learning_rate": 2.9e-05, | |
| "loss": 8.0931, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.06626173384870238, | |
| "grad_norm": 12.407663345336914, | |
| "learning_rate": 3e-05, | |
| "loss": 7.9797, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06847045831032579, | |
| "grad_norm": 12.427129745483398, | |
| "learning_rate": 3.1e-05, | |
| "loss": 8.3415, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.0706791827719492, | |
| "grad_norm": 13.105308532714844, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 7.8873, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.07288790723357261, | |
| "grad_norm": 12.776379585266113, | |
| "learning_rate": 3.3e-05, | |
| "loss": 7.837, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.07509663169519602, | |
| "grad_norm": 13.840560913085938, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 7.5128, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.07730535615681944, | |
| "grad_norm": 12.904688835144043, | |
| "learning_rate": 3.5e-05, | |
| "loss": 7.5659, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.07951408061844285, | |
| "grad_norm": 12.567195892333984, | |
| "learning_rate": 3.6e-05, | |
| "loss": 7.7097, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.08172280508006627, | |
| "grad_norm": 11.897880554199219, | |
| "learning_rate": 3.7e-05, | |
| "loss": 7.2553, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.08393152954168967, | |
| "grad_norm": 13.644888877868652, | |
| "learning_rate": 3.8e-05, | |
| "loss": 7.6672, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.08614025400331309, | |
| "grad_norm": 12.215312004089355, | |
| "learning_rate": 3.9000000000000006e-05, | |
| "loss": 7.321, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.0883489784649365, | |
| "grad_norm": 11.68522834777832, | |
| "learning_rate": 4e-05, | |
| "loss": 7.67, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09055770292655992, | |
| "grad_norm": 11.208578109741211, | |
| "learning_rate": 4.1e-05, | |
| "loss": 7.6815, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.09276642738818332, | |
| "grad_norm": 11.380876541137695, | |
| "learning_rate": 4.2e-05, | |
| "loss": 7.7252, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.09497515184980673, | |
| "grad_norm": 10.951228141784668, | |
| "learning_rate": 4.3e-05, | |
| "loss": 7.2269, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.09718387631143015, | |
| "grad_norm": 10.96963882446289, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 6.5408, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.09939260077305356, | |
| "grad_norm": 11.59006404876709, | |
| "learning_rate": 4.5e-05, | |
| "loss": 6.9152, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10160132523467698, | |
| "grad_norm": 11.817084312438965, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 6.6265, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.10381004969630038, | |
| "grad_norm": 12.219128608703613, | |
| "learning_rate": 4.7e-05, | |
| "loss": 6.6341, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.1060187741579238, | |
| "grad_norm": 12.786672592163086, | |
| "learning_rate": 4.8e-05, | |
| "loss": 6.7037, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.10822749861954721, | |
| "grad_norm": 12.815557479858398, | |
| "learning_rate": 4.9e-05, | |
| "loss": 6.7343, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.11043622308117063, | |
| "grad_norm": 12.737327575683594, | |
| "learning_rate": 5e-05, | |
| "loss": 7.3387, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11264494754279404, | |
| "grad_norm": 6.819960594177246, | |
| "learning_rate": 5.1000000000000006e-05, | |
| "loss": 4.7037, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.11485367200441744, | |
| "grad_norm": 12.15744686126709, | |
| "learning_rate": 5.2000000000000004e-05, | |
| "loss": 6.148, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.11706239646604086, | |
| "grad_norm": 11.386445045471191, | |
| "learning_rate": 5.300000000000001e-05, | |
| "loss": 6.0605, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.11927112092766427, | |
| "grad_norm": 11.03603458404541, | |
| "learning_rate": 5.4000000000000005e-05, | |
| "loss": 6.1902, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.12147984538928769, | |
| "grad_norm": 10.725998878479004, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 5.8361, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1236885698509111, | |
| "grad_norm": 8.622870445251465, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 5.865, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.12589729431253452, | |
| "grad_norm": 7.68513298034668, | |
| "learning_rate": 5.6999999999999996e-05, | |
| "loss": 6.1753, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.12810601877415792, | |
| "grad_norm": 7.354230880737305, | |
| "learning_rate": 5.8e-05, | |
| "loss": 5.8207, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.13031474323578135, | |
| "grad_norm": 6.177774906158447, | |
| "learning_rate": 5.9e-05, | |
| "loss": 5.7267, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.13252346769740475, | |
| "grad_norm": 6.288235187530518, | |
| "learning_rate": 6e-05, | |
| "loss": 5.7419, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13473219215902815, | |
| "grad_norm": 6.973496437072754, | |
| "learning_rate": 6.1e-05, | |
| "loss": 5.8684, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.13694091662065158, | |
| "grad_norm": 6.646786689758301, | |
| "learning_rate": 6.2e-05, | |
| "loss": 5.6718, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.13914964108227498, | |
| "grad_norm": 7.017954349517822, | |
| "learning_rate": 6.3e-05, | |
| "loss": 5.6098, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.1413583655438984, | |
| "grad_norm": 7.299660682678223, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 5.6966, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.1435670900055218, | |
| "grad_norm": 8.129739761352539, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 5.8456, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14577581446714521, | |
| "grad_norm": 8.626437187194824, | |
| "learning_rate": 6.6e-05, | |
| "loss": 5.9174, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.14798453892876864, | |
| "grad_norm": 9.129312515258789, | |
| "learning_rate": 6.7e-05, | |
| "loss": 5.7032, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.15019326339039205, | |
| "grad_norm": 8.709549903869629, | |
| "learning_rate": 6.800000000000001e-05, | |
| "loss": 5.8617, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.15240198785201547, | |
| "grad_norm": 9.02973747253418, | |
| "learning_rate": 6.9e-05, | |
| "loss": 5.8936, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.15461071231363888, | |
| "grad_norm": 8.83764934539795, | |
| "learning_rate": 7e-05, | |
| "loss": 5.7209, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15681943677526228, | |
| "grad_norm": 9.076760292053223, | |
| "learning_rate": 7.1e-05, | |
| "loss": 5.7627, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.1590281612368857, | |
| "grad_norm": 9.358622550964355, | |
| "learning_rate": 7.2e-05, | |
| "loss": 5.9458, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.1612368856985091, | |
| "grad_norm": 8.883482933044434, | |
| "learning_rate": 7.3e-05, | |
| "loss": 5.6288, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.16344561016013254, | |
| "grad_norm": 9.588410377502441, | |
| "learning_rate": 7.4e-05, | |
| "loss": 5.5858, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.16565433462175594, | |
| "grad_norm": 10.137778282165527, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 5.5368, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16786305908337934, | |
| "grad_norm": 10.220065116882324, | |
| "learning_rate": 7.6e-05, | |
| "loss": 5.7181, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.17007178354500277, | |
| "grad_norm": 9.702897071838379, | |
| "learning_rate": 7.7e-05, | |
| "loss": 5.6469, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.17228050800662617, | |
| "grad_norm": 9.3972806930542, | |
| "learning_rate": 7.800000000000001e-05, | |
| "loss": 5.4397, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.17448923246824957, | |
| "grad_norm": 10.074295043945312, | |
| "learning_rate": 7.900000000000001e-05, | |
| "loss": 5.397, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.176697956929873, | |
| "grad_norm": 10.400197982788086, | |
| "learning_rate": 8e-05, | |
| "loss": 5.5235, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1789066813914964, | |
| "grad_norm": 10.35517406463623, | |
| "learning_rate": 8.1e-05, | |
| "loss": 5.6897, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.18111540585311983, | |
| "grad_norm": 11.12540054321289, | |
| "learning_rate": 8.2e-05, | |
| "loss": 5.5081, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.18332413031474323, | |
| "grad_norm": 11.132303237915039, | |
| "learning_rate": 8.3e-05, | |
| "loss": 5.8882, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.18553285477636663, | |
| "grad_norm": 10.35818099975586, | |
| "learning_rate": 8.4e-05, | |
| "loss": 5.1969, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.18774157923799006, | |
| "grad_norm": 11.621879577636719, | |
| "learning_rate": 8.5e-05, | |
| "loss": 5.6639, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.18995030369961347, | |
| "grad_norm": 11.88983154296875, | |
| "learning_rate": 8.6e-05, | |
| "loss": 5.7786, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.1921590281612369, | |
| "grad_norm": 11.17719841003418, | |
| "learning_rate": 8.7e-05, | |
| "loss": 5.5071, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.1943677526228603, | |
| "grad_norm": 11.159234046936035, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 5.4638, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.1965764770844837, | |
| "grad_norm": 10.719842910766602, | |
| "learning_rate": 8.900000000000001e-05, | |
| "loss": 5.3603, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.19878520154610713, | |
| "grad_norm": 10.188541412353516, | |
| "learning_rate": 9e-05, | |
| "loss": 5.7355, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20099392600773053, | |
| "grad_norm": 10.070486068725586, | |
| "learning_rate": 9.1e-05, | |
| "loss": 5.1332, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.20320265046935396, | |
| "grad_norm": 10.740564346313477, | |
| "learning_rate": 9.200000000000001e-05, | |
| "loss": 5.6564, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.20541137493097736, | |
| "grad_norm": 11.773510932922363, | |
| "learning_rate": 9.300000000000001e-05, | |
| "loss": 5.3035, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.20762009939260076, | |
| "grad_norm": 10.705671310424805, | |
| "learning_rate": 9.4e-05, | |
| "loss": 5.4218, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.2098288238542242, | |
| "grad_norm": 10.64268684387207, | |
| "learning_rate": 9.5e-05, | |
| "loss": 5.2128, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2120375483158476, | |
| "grad_norm": 11.824028015136719, | |
| "learning_rate": 9.6e-05, | |
| "loss": 5.5192, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.21424627277747102, | |
| "grad_norm": 10.851103782653809, | |
| "learning_rate": 9.7e-05, | |
| "loss": 5.6141, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.21645499723909442, | |
| "grad_norm": 11.880955696105957, | |
| "learning_rate": 9.8e-05, | |
| "loss": 5.7783, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.21866372170071782, | |
| "grad_norm": 11.984313011169434, | |
| "learning_rate": 9.900000000000001e-05, | |
| "loss": 5.1242, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.22087244616234125, | |
| "grad_norm": 12.726852416992188, | |
| "learning_rate": 0.0001, | |
| "loss": 5.1561, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22308117062396465, | |
| "grad_norm": 7.531022548675537, | |
| "learning_rate": 9.999800863190707e-05, | |
| "loss": 4.6879, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.22528989508558808, | |
| "grad_norm": 9.076756477355957, | |
| "learning_rate": 9.999203468625017e-05, | |
| "loss": 5.4269, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.22749861954721148, | |
| "grad_norm": 7.9871954917907715, | |
| "learning_rate": 9.998207863888228e-05, | |
| "loss": 5.3304, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.22970734400883489, | |
| "grad_norm": 7.338669300079346, | |
| "learning_rate": 9.99681412828496e-05, | |
| "loss": 5.1078, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.23191606847045831, | |
| "grad_norm": 5.404487609863281, | |
| "learning_rate": 9.995022372832836e-05, | |
| "loss": 4.879, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23412479293208172, | |
| "grad_norm": 5.911509037017822, | |
| "learning_rate": 9.992832740253645e-05, | |
| "loss": 4.5742, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.23633351739370514, | |
| "grad_norm": 5.688644886016846, | |
| "learning_rate": 9.990245404961963e-05, | |
| "loss": 4.9074, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.23854224185532855, | |
| "grad_norm": 5.671441078186035, | |
| "learning_rate": 9.987260573051269e-05, | |
| "loss": 4.8472, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.24075096631695195, | |
| "grad_norm": 6.069337368011475, | |
| "learning_rate": 9.983878482277522e-05, | |
| "loss": 4.866, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.24295969077857538, | |
| "grad_norm": 6.0655694007873535, | |
| "learning_rate": 9.98009940204023e-05, | |
| "loss": 4.9271, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24516841524019878, | |
| "grad_norm": 6.612443447113037, | |
| "learning_rate": 9.975923633360985e-05, | |
| "loss": 5.0706, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.2473771397018222, | |
| "grad_norm": 6.772550582885742, | |
| "learning_rate": 9.971351508859488e-05, | |
| "loss": 5.1031, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.2495858641634456, | |
| "grad_norm": 6.7489800453186035, | |
| "learning_rate": 9.96638339272705e-05, | |
| "loss": 5.3257, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.25179458862506904, | |
| "grad_norm": 6.750669002532959, | |
| "learning_rate": 9.961019680697593e-05, | |
| "loss": 5.0159, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.25400331308669244, | |
| "grad_norm": 7.035401821136475, | |
| "learning_rate": 9.955260800016113e-05, | |
| "loss": 4.9488, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.25621203754831584, | |
| "grad_norm": 7.1767473220825195, | |
| "learning_rate": 9.949107209404665e-05, | |
| "loss": 4.7798, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.25842076200993924, | |
| "grad_norm": 7.472834587097168, | |
| "learning_rate": 9.942559399025803e-05, | |
| "loss": 4.9628, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.2606294864715627, | |
| "grad_norm": 7.51600456237793, | |
| "learning_rate": 9.935617890443557e-05, | |
| "loss": 5.4082, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.2628382109331861, | |
| "grad_norm": 7.743865489959717, | |
| "learning_rate": 9.928283236581873e-05, | |
| "loss": 5.2246, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.2650469353948095, | |
| "grad_norm": 7.967807292938232, | |
| "learning_rate": 9.92055602168058e-05, | |
| "loss": 5.0226, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2672556598564329, | |
| "grad_norm": 8.627487182617188, | |
| "learning_rate": 9.912436861248847e-05, | |
| "loss": 5.1528, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.2694643843180563, | |
| "grad_norm": 8.125624656677246, | |
| "learning_rate": 9.903926402016153e-05, | |
| "loss": 5.3868, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.27167310877967976, | |
| "grad_norm": 8.606098175048828, | |
| "learning_rate": 9.895025321880776e-05, | |
| "loss": 5.1796, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.27388183324130316, | |
| "grad_norm": 8.127717971801758, | |
| "learning_rate": 9.885734329855798e-05, | |
| "loss": 4.8976, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.27609055770292656, | |
| "grad_norm": 8.626016616821289, | |
| "learning_rate": 9.87605416601262e-05, | |
| "loss": 4.9745, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.27829928216454997, | |
| "grad_norm": 8.542702674865723, | |
| "learning_rate": 9.865985601422018e-05, | |
| "loss": 4.7371, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.28050800662617337, | |
| "grad_norm": 8.690094947814941, | |
| "learning_rate": 9.855529438092724e-05, | |
| "loss": 5.1957, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.2827167310877968, | |
| "grad_norm": 8.771077156066895, | |
| "learning_rate": 9.844686508907537e-05, | |
| "loss": 4.7499, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.2849254555494202, | |
| "grad_norm": 8.793961524963379, | |
| "learning_rate": 9.833457677556987e-05, | |
| "loss": 5.3736, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.2871341800110436, | |
| "grad_norm": 9.016656875610352, | |
| "learning_rate": 9.821843838470534e-05, | |
| "loss": 5.0202, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.28934290447266703, | |
| "grad_norm": 8.994014739990234, | |
| "learning_rate": 9.809845916745321e-05, | |
| "loss": 4.9291, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.29155162893429043, | |
| "grad_norm": 8.716841697692871, | |
| "learning_rate": 9.797464868072488e-05, | |
| "loss": 4.9976, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.2937603533959139, | |
| "grad_norm": 9.860023498535156, | |
| "learning_rate": 9.784701678661045e-05, | |
| "loss": 4.553, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.2959690778575373, | |
| "grad_norm": 9.492415428161621, | |
| "learning_rate": 9.77155736515932e-05, | |
| "loss": 5.1861, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.2981778023191607, | |
| "grad_norm": 9.687310218811035, | |
| "learning_rate": 9.758032974573972e-05, | |
| "loss": 5.3625, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.3003865267807841, | |
| "grad_norm": 9.293684959411621, | |
| "learning_rate": 9.744129584186598e-05, | |
| "loss": 5.2148, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.3025952512424075, | |
| "grad_norm": 9.503033638000488, | |
| "learning_rate": 9.729848301467919e-05, | |
| "loss": 5.5468, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.30480397570403095, | |
| "grad_norm": 9.93913459777832, | |
| "learning_rate": 9.715190263989561e-05, | |
| "loss": 5.2361, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.30701270016565435, | |
| "grad_norm": 8.820573806762695, | |
| "learning_rate": 9.700156639333451e-05, | |
| "loss": 4.6691, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.30922142462727775, | |
| "grad_norm": 9.540395736694336, | |
| "learning_rate": 9.68474862499881e-05, | |
| "loss": 4.5415, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.31143014908890115, | |
| "grad_norm": 9.923493385314941, | |
| "learning_rate": 9.668967448306758e-05, | |
| "loss": 4.8208, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.31363887355052456, | |
| "grad_norm": 10.182249069213867, | |
| "learning_rate": 9.652814366302568e-05, | |
| "loss": 5.0238, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.315847598012148, | |
| "grad_norm": 9.629402160644531, | |
| "learning_rate": 9.636290665655524e-05, | |
| "loss": 5.072, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.3180563224737714, | |
| "grad_norm": 10.206043243408203, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 5.0429, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.3202650469353948, | |
| "grad_norm": 10.271145820617676, | |
| "learning_rate": 9.602136702612794e-05, | |
| "loss": 5.6498, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.3224737713970182, | |
| "grad_norm": 9.493379592895508, | |
| "learning_rate": 9.584509160741599e-05, | |
| "loss": 4.8689, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.3246824958586416, | |
| "grad_norm": 10.454998016357422, | |
| "learning_rate": 9.566516441059827e-05, | |
| "loss": 5.7323, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.3268912203202651, | |
| "grad_norm": 11.290060997009277, | |
| "learning_rate": 9.548159976772592e-05, | |
| "loss": 5.1861, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.3290999447818885, | |
| "grad_norm": 10.804366111755371, | |
| "learning_rate": 9.529441230058987e-05, | |
| "loss": 5.0733, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.3313086692435119, | |
| "grad_norm": 11.59500503540039, | |
| "learning_rate": 9.510361691955608e-05, | |
| "loss": 5.185, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3335173937051353, | |
| "grad_norm": 5.955521106719971, | |
| "learning_rate": 9.490922882237791e-05, | |
| "loss": 3.7304, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.3357261181667587, | |
| "grad_norm": 6.95432710647583, | |
| "learning_rate": 9.471126349298556e-05, | |
| "loss": 5.0306, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.3379348426283821, | |
| "grad_norm": 7.6052937507629395, | |
| "learning_rate": 9.450973670025264e-05, | |
| "loss": 4.8507, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.34014356709000554, | |
| "grad_norm": 7.37522029876709, | |
| "learning_rate": 9.430466449674014e-05, | |
| "loss": 4.9886, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.34235229155162894, | |
| "grad_norm": 6.831510543823242, | |
| "learning_rate": 9.409606321741775e-05, | |
| "loss": 4.9264, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.34456101601325234, | |
| "grad_norm": 6.073673248291016, | |
| "learning_rate": 9.388394947836279e-05, | |
| "loss": 4.7457, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.34676974047487574, | |
| "grad_norm": 5.337328910827637, | |
| "learning_rate": 9.366834017543651e-05, | |
| "loss": 4.7442, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.34897846493649914, | |
| "grad_norm": 5.276707649230957, | |
| "learning_rate": 9.344925248293837e-05, | |
| "loss": 4.9261, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.3511871893981226, | |
| "grad_norm": 5.147680759429932, | |
| "learning_rate": 9.322670385223797e-05, | |
| "loss": 4.4692, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.353395913859746, | |
| "grad_norm": 5.277919292449951, | |
| "learning_rate": 9.300071201038503e-05, | |
| "loss": 4.8624, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3556046383213694, | |
| "grad_norm": 5.418294906616211, | |
| "learning_rate": 9.277129495869726e-05, | |
| "loss": 4.6649, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.3578133627829928, | |
| "grad_norm": 5.715703964233398, | |
| "learning_rate": 9.253847097132655e-05, | |
| "loss": 5.2021, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.3600220872446162, | |
| "grad_norm": 5.969396591186523, | |
| "learning_rate": 9.230225859380326e-05, | |
| "loss": 5.0423, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.36223081170623966, | |
| "grad_norm": 5.960832118988037, | |
| "learning_rate": 9.206267664155907e-05, | |
| "loss": 4.8568, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.36443953616786307, | |
| "grad_norm": 6.366528034210205, | |
| "learning_rate": 9.181974419842818e-05, | |
| "loss": 4.7735, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.36664826062948647, | |
| "grad_norm": 6.627102375030518, | |
| "learning_rate": 9.157348061512727e-05, | |
| "loss": 4.8802, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.36885698509110987, | |
| "grad_norm": 7.070224761962891, | |
| "learning_rate": 9.1323905507714e-05, | |
| "loss": 5.1596, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.37106570955273327, | |
| "grad_norm": 7.091405868530273, | |
| "learning_rate": 9.107103875602459e-05, | |
| "loss": 4.9509, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.3732744340143567, | |
| "grad_norm": 7.415458679199219, | |
| "learning_rate": 9.08149005020903e-05, | |
| "loss": 4.395, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.3754831584759801, | |
| "grad_norm": 7.86855936050415, | |
| "learning_rate": 9.055551114853295e-05, | |
| "loss": 5.0404, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.37769188293760353, | |
| "grad_norm": 7.8068718910217285, | |
| "learning_rate": 9.02928913569399e-05, | |
| "loss": 4.9543, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.37990060739922693, | |
| "grad_norm": 8.077149391174316, | |
| "learning_rate": 9.002706204621803e-05, | |
| "loss": 4.5707, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.38210933186085033, | |
| "grad_norm": 7.964366912841797, | |
| "learning_rate": 8.975804439092765e-05, | |
| "loss": 4.6822, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.3843180563224738, | |
| "grad_norm": 8.550042152404785, | |
| "learning_rate": 8.94858598195958e-05, | |
| "loss": 4.664, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.3865267807840972, | |
| "grad_norm": 8.32715129852295, | |
| "learning_rate": 8.921053001300928e-05, | |
| "loss": 4.5238, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3887355052457206, | |
| "grad_norm": 8.401163101196289, | |
| "learning_rate": 8.893207690248776e-05, | |
| "loss": 5.061, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.390944229707344, | |
| "grad_norm": 8.160713195800781, | |
| "learning_rate": 8.865052266813685e-05, | |
| "loss": 4.4293, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.3931529541689674, | |
| "grad_norm": 8.723557472229004, | |
| "learning_rate": 8.836588973708129e-05, | |
| "loss": 4.7296, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.39536167863059085, | |
| "grad_norm": 8.8028564453125, | |
| "learning_rate": 8.807820078167856e-05, | |
| "loss": 4.9615, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.39757040309221425, | |
| "grad_norm": 8.326678276062012, | |
| "learning_rate": 8.778747871771292e-05, | |
| "loss": 4.9342, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.39977912755383765, | |
| "grad_norm": 9.141505241394043, | |
| "learning_rate": 8.749374670257005e-05, | |
| "loss": 5.0855, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.40198785201546106, | |
| "grad_norm": 8.820999145507812, | |
| "learning_rate": 8.719702813339248e-05, | |
| "loss": 5.3793, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.40419657647708446, | |
| "grad_norm": 8.986492156982422, | |
| "learning_rate": 8.689734664521583e-05, | |
| "loss": 5.0518, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.4064053009387079, | |
| "grad_norm": 8.636345863342285, | |
| "learning_rate": 8.659472610908627e-05, | |
| "loss": 5.2209, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.4086140254003313, | |
| "grad_norm": 8.594338417053223, | |
| "learning_rate": 8.628919063015898e-05, | |
| "loss": 5.1603, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4108227498619547, | |
| "grad_norm": 8.710060119628906, | |
| "learning_rate": 8.598076454577814e-05, | |
| "loss": 5.2937, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.4130314743235781, | |
| "grad_norm": 8.30070972442627, | |
| "learning_rate": 8.566947242353827e-05, | |
| "loss": 4.9801, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.4152401987852015, | |
| "grad_norm": 8.696635246276855, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 4.7529, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.417448923246825, | |
| "grad_norm": 8.520990371704102, | |
| "learning_rate": 8.503838947535179e-05, | |
| "loss": 5.344, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.4196576477084484, | |
| "grad_norm": 8.845208168029785, | |
| "learning_rate": 8.471864891814304e-05, | |
| "loss": 4.7962, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4218663721700718, | |
| "grad_norm": 8.896949768066406, | |
| "learning_rate": 8.439614285654689e-05, | |
| "loss": 5.1538, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.4240750966316952, | |
| "grad_norm": 8.677151679992676, | |
| "learning_rate": 8.407089697969457e-05, | |
| "loss": 5.0412, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.4262838210933186, | |
| "grad_norm": 9.09294319152832, | |
| "learning_rate": 8.374293719495651e-05, | |
| "loss": 4.8806, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.42849254555494204, | |
| "grad_norm": 8.677190780639648, | |
| "learning_rate": 8.341228962587882e-05, | |
| "loss": 4.5531, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.43070127001656544, | |
| "grad_norm": 9.5293550491333, | |
| "learning_rate": 8.307898061010221e-05, | |
| "loss": 5.1039, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.43290999447818884, | |
| "grad_norm": 9.922966957092285, | |
| "learning_rate": 8.274303669726426e-05, | |
| "loss": 5.0978, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.43511871893981224, | |
| "grad_norm": 9.41407585144043, | |
| "learning_rate": 8.240448464688453e-05, | |
| "loss": 4.5112, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.43732744340143564, | |
| "grad_norm": 10.871786117553711, | |
| "learning_rate": 8.206335142623305e-05, | |
| "loss": 4.7271, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.4395361678630591, | |
| "grad_norm": 11.003656387329102, | |
| "learning_rate": 8.171966420818228e-05, | |
| "loss": 4.8793, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.4417448923246825, | |
| "grad_norm": 12.43280029296875, | |
| "learning_rate": 8.13734503690426e-05, | |
| "loss": 5.1866, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4439536167863059, | |
| "grad_norm": 5.124542713165283, | |
| "learning_rate": 8.102473748638173e-05, | |
| "loss": 3.8602, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.4461623412479293, | |
| "grad_norm": 7.500517845153809, | |
| "learning_rate": 8.067355333682798e-05, | |
| "loss": 4.7929, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.4483710657095527, | |
| "grad_norm": 7.736271858215332, | |
| "learning_rate": 8.031992589385777e-05, | |
| "loss": 4.9299, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.45057979017117616, | |
| "grad_norm": 8.110419273376465, | |
| "learning_rate": 7.996388332556735e-05, | |
| "loss": 4.9557, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.45278851463279957, | |
| "grad_norm": 6.666763782501221, | |
| "learning_rate": 7.960545399242916e-05, | |
| "loss": 4.5717, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.45499723909442297, | |
| "grad_norm": 5.63977575302124, | |
| "learning_rate": 7.924466644503265e-05, | |
| "loss": 4.2633, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.45720596355604637, | |
| "grad_norm": 5.997104644775391, | |
| "learning_rate": 7.88815494218103e-05, | |
| "loss": 4.6012, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.45941468801766977, | |
| "grad_norm": 5.42753267288208, | |
| "learning_rate": 7.85161318467482e-05, | |
| "loss": 4.9035, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.4616234124792932, | |
| "grad_norm": 5.311511039733887, | |
| "learning_rate": 7.814844282708239e-05, | |
| "loss": 4.4711, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.46383213694091663, | |
| "grad_norm": 5.2506256103515625, | |
| "learning_rate": 7.777851165098012e-05, | |
| "loss": 4.6818, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.46604086140254003, | |
| "grad_norm": 5.2805280685424805, | |
| "learning_rate": 7.7406367785207e-05, | |
| "loss": 4.9481, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.46824958586416343, | |
| "grad_norm": 5.382137298583984, | |
| "learning_rate": 7.703204087277988e-05, | |
| "loss": 4.5646, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.47045831032578683, | |
| "grad_norm": 5.792073726654053, | |
| "learning_rate": 7.665556073060552e-05, | |
| "loss": 4.8119, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.4726670347874103, | |
| "grad_norm": 6.007907867431641, | |
| "learning_rate": 7.627695734710564e-05, | |
| "loss": 4.687, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.4748757592490337, | |
| "grad_norm": 5.8332109451293945, | |
| "learning_rate": 7.589626087982816e-05, | |
| "loss": 4.1969, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4770844837106571, | |
| "grad_norm": 6.367125511169434, | |
| "learning_rate": 7.5513501653045e-05, | |
| "loss": 4.3616, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.4792932081722805, | |
| "grad_norm": 6.897268772125244, | |
| "learning_rate": 7.512871015533658e-05, | |
| "loss": 4.9368, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.4815019326339039, | |
| "grad_norm": 7.102772235870361, | |
| "learning_rate": 7.474191703716339e-05, | |
| "loss": 4.8659, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.48371065709552735, | |
| "grad_norm": 7.861514568328857, | |
| "learning_rate": 7.435315310842436e-05, | |
| "loss": 4.5714, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.48591938155715075, | |
| "grad_norm": 8.072993278503418, | |
| "learning_rate": 7.396244933600285e-05, | |
| "loss": 4.668, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.48812810601877415, | |
| "grad_norm": 8.46921443939209, | |
| "learning_rate": 7.35698368412999e-05, | |
| "loss": 4.932, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.49033683048039756, | |
| "grad_norm": 7.927035331726074, | |
| "learning_rate": 7.317534689775528e-05, | |
| "loss": 4.6539, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.49254555494202096, | |
| "grad_norm": 8.272632598876953, | |
| "learning_rate": 7.277901092835649e-05, | |
| "loss": 4.7136, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.4947542794036444, | |
| "grad_norm": 8.767943382263184, | |
| "learning_rate": 7.238086050313563e-05, | |
| "loss": 4.7382, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.4969630038652678, | |
| "grad_norm": 8.93777084350586, | |
| "learning_rate": 7.198092733665487e-05, | |
| "loss": 5.0617, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.4991717283268912, | |
| "grad_norm": 8.695636749267578, | |
| "learning_rate": 7.157924328548002e-05, | |
| "loss": 4.9294, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.5013804527885146, | |
| "grad_norm": 8.272305488586426, | |
| "learning_rate": 7.117584034564329e-05, | |
| "loss": 4.5938, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.5035891772501381, | |
| "grad_norm": 8.093917846679688, | |
| "learning_rate": 7.077075065009433e-05, | |
| "loss": 4.6757, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.5057979017117614, | |
| "grad_norm": 8.725174903869629, | |
| "learning_rate": 7.036400646614094e-05, | |
| "loss": 4.9074, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.5080066261733849, | |
| "grad_norm": 8.211441040039062, | |
| "learning_rate": 6.995564019287871e-05, | |
| "loss": 4.7963, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5102153506350083, | |
| "grad_norm": 8.903889656066895, | |
| "learning_rate": 6.954568435861033e-05, | |
| "loss": 4.9038, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.5124240750966317, | |
| "grad_norm": 8.311030387878418, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 4.8634, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.5146327995582551, | |
| "grad_norm": 8.491503715515137, | |
| "learning_rate": 6.872113475074486e-05, | |
| "loss": 5.043, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.5168415240198785, | |
| "grad_norm": 8.99098014831543, | |
| "learning_rate": 6.830660665641898e-05, | |
| "loss": 5.0784, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.5190502484815019, | |
| "grad_norm": 8.576278686523438, | |
| "learning_rate": 6.789062035439767e-05, | |
| "loss": 4.4813, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5212589729431254, | |
| "grad_norm": 8.29208755493164, | |
| "learning_rate": 6.747320897995493e-05, | |
| "loss": 4.9514, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.5234676974047487, | |
| "grad_norm": 9.072782516479492, | |
| "learning_rate": 6.705440578187842e-05, | |
| "loss": 5.0796, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.5256764218663722, | |
| "grad_norm": 9.66883373260498, | |
| "learning_rate": 6.663424411982121e-05, | |
| "loss": 4.745, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.5278851463279955, | |
| "grad_norm": 8.428522109985352, | |
| "learning_rate": 6.621275746164437e-05, | |
| "loss": 4.599, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.530093870789619, | |
| "grad_norm": 9.09389591217041, | |
| "learning_rate": 6.578997938075125e-05, | |
| "loss": 4.443, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5323025952512425, | |
| "grad_norm": 8.505534172058105, | |
| "learning_rate": 6.536594355341305e-05, | |
| "loss": 4.5598, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.5345113197128658, | |
| "grad_norm": 9.29316520690918, | |
| "learning_rate": 6.494068375608646e-05, | |
| "loss": 4.5639, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.5367200441744893, | |
| "grad_norm": 8.853930473327637, | |
| "learning_rate": 6.451423386272312e-05, | |
| "loss": 4.6652, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.5389287686361126, | |
| "grad_norm": 9.70859146118164, | |
| "learning_rate": 6.408662784207149e-05, | |
| "loss": 5.0798, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.5411374930977361, | |
| "grad_norm": 9.851663589477539, | |
| "learning_rate": 6.3657899754971e-05, | |
| "loss": 4.6318, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5433462175593595, | |
| "grad_norm": 9.574539184570312, | |
| "learning_rate": 6.322808375163895e-05, | |
| "loss": 4.541, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.5455549420209829, | |
| "grad_norm": 10.078091621398926, | |
| "learning_rate": 6.279721406895038e-05, | |
| "loss": 4.6311, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.5477636664826063, | |
| "grad_norm": 11.278042793273926, | |
| "learning_rate": 6.236532502771078e-05, | |
| "loss": 5.0703, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.5499723909442297, | |
| "grad_norm": 10.999835968017578, | |
| "learning_rate": 6.193245102992244e-05, | |
| "loss": 5.322, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.5521811154058531, | |
| "grad_norm": 10.701493263244629, | |
| "learning_rate": 6.149862655604404e-05, | |
| "loss": 4.7289, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5543898398674766, | |
| "grad_norm": 4.466903209686279, | |
| "learning_rate": 6.106388616224419e-05, | |
| "loss": 3.7511, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.5565985643290999, | |
| "grad_norm": 6.9619526863098145, | |
| "learning_rate": 6.062826447764883e-05, | |
| "loss": 4.6559, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.5588072887907234, | |
| "grad_norm": 6.754791736602783, | |
| "learning_rate": 6.019179620158294e-05, | |
| "loss": 4.7606, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.5610160132523467, | |
| "grad_norm": 7.005521774291992, | |
| "learning_rate": 5.9754516100806423e-05, | |
| "loss": 4.4367, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.5632247377139702, | |
| "grad_norm": 7.08423376083374, | |
| "learning_rate": 5.93164590067449e-05, | |
| "loss": 4.5616, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5654334621755936, | |
| "grad_norm": 6.051517009735107, | |
| "learning_rate": 5.887765981271518e-05, | |
| "loss": 4.6624, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.567642186637217, | |
| "grad_norm": 6.105447292327881, | |
| "learning_rate": 5.84381534711458e-05, | |
| "loss": 4.5011, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.5698509110988405, | |
| "grad_norm": 6.139936447143555, | |
| "learning_rate": 5.799797499079301e-05, | |
| "loss": 4.5359, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.5720596355604638, | |
| "grad_norm": 5.590278148651123, | |
| "learning_rate": 5.7557159433952034e-05, | |
| "loss": 4.6673, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.5742683600220873, | |
| "grad_norm": 5.5569586753845215, | |
| "learning_rate": 5.7115741913664264e-05, | |
| "loss": 4.2704, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5764770844837107, | |
| "grad_norm": 5.2458319664001465, | |
| "learning_rate": 5.6673757590920306e-05, | |
| "loss": 4.0604, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.5786858089453341, | |
| "grad_norm": 5.579367637634277, | |
| "learning_rate": 5.62312416718593e-05, | |
| "loss": 4.4609, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.5808945334069575, | |
| "grad_norm": 5.590094566345215, | |
| "learning_rate": 5.578822940496451e-05, | |
| "loss": 4.7488, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.5831032578685809, | |
| "grad_norm": 5.870893955230713, | |
| "learning_rate": 5.534475607825566e-05, | |
| "loss": 4.5398, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.5853119823302043, | |
| "grad_norm": 6.134853839874268, | |
| "learning_rate": 5.490085701647805e-05, | |
| "loss": 4.6951, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5875207067918278, | |
| "grad_norm": 6.254935264587402, | |
| "learning_rate": 5.44565675782888e-05, | |
| "loss": 4.5561, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.5897294312534511, | |
| "grad_norm": 6.584839344024658, | |
| "learning_rate": 5.401192315344039e-05, | |
| "loss": 4.8534, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.5919381557150746, | |
| "grad_norm": 7.013578414916992, | |
| "learning_rate": 5.3566959159961615e-05, | |
| "loss": 4.6348, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.5941468801766979, | |
| "grad_norm": 7.184816837310791, | |
| "learning_rate": 5.312171104133646e-05, | |
| "loss": 4.271, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.5963556046383214, | |
| "grad_norm": 7.353315830230713, | |
| "learning_rate": 5.2676214263680754e-05, | |
| "loss": 4.462, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5985643290999448, | |
| "grad_norm": 8.102008819580078, | |
| "learning_rate": 5.223050431291729e-05, | |
| "loss": 4.7596, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.6007730535615682, | |
| "grad_norm": 7.693032264709473, | |
| "learning_rate": 5.178461669194903e-05, | |
| "loss": 4.7534, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.6029817780231916, | |
| "grad_norm": 8.456910133361816, | |
| "learning_rate": 5.133858691783123e-05, | |
| "loss": 4.9262, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.605190502484815, | |
| "grad_norm": 8.400936126708984, | |
| "learning_rate": 5.08924505189423e-05, | |
| "loss": 4.4905, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.6073992269464384, | |
| "grad_norm": 8.744009971618652, | |
| "learning_rate": 5.0446243032153875e-05, | |
| "loss": 4.5625, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.6096079514080619, | |
| "grad_norm": 8.910327911376953, | |
| "learning_rate": 5e-05, | |
| "loss": 5.2096, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.6118166758696852, | |
| "grad_norm": 8.787812232971191, | |
| "learning_rate": 4.9553756967846144e-05, | |
| "loss": 4.8109, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.6140254003313087, | |
| "grad_norm": 8.353130340576172, | |
| "learning_rate": 4.9107549481057696e-05, | |
| "loss": 4.6905, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.616234124792932, | |
| "grad_norm": 8.456555366516113, | |
| "learning_rate": 4.866141308216878e-05, | |
| "loss": 5.3734, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.6184428492545555, | |
| "grad_norm": 8.801351547241211, | |
| "learning_rate": 4.821538330805098e-05, | |
| "loss": 5.1219, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.620651573716179, | |
| "grad_norm": 8.711759567260742, | |
| "learning_rate": 4.7769495687082725e-05, | |
| "loss": 5.0409, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.6228602981778023, | |
| "grad_norm": 8.708940505981445, | |
| "learning_rate": 4.7323785736319244e-05, | |
| "loss": 4.6526, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.6250690226394258, | |
| "grad_norm": 8.809032440185547, | |
| "learning_rate": 4.687828895866355e-05, | |
| "loss": 4.8223, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.6272777471010491, | |
| "grad_norm": 9.175921440124512, | |
| "learning_rate": 4.643304084003839e-05, | |
| "loss": 4.8745, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.6294864715626726, | |
| "grad_norm": 8.915994644165039, | |
| "learning_rate": 4.598807684655963e-05, | |
| "loss": 5.0015, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.631695196024296, | |
| "grad_norm": 9.178176879882812, | |
| "learning_rate": 4.5543432421711205e-05, | |
| "loss": 4.9042, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.6339039204859194, | |
| "grad_norm": 8.78802490234375, | |
| "learning_rate": 4.509914298352197e-05, | |
| "loss": 5.0455, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.6361126449475428, | |
| "grad_norm": 9.149216651916504, | |
| "learning_rate": 4.4655243921744374e-05, | |
| "loss": 4.8976, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.6383213694091662, | |
| "grad_norm": 9.336906433105469, | |
| "learning_rate": 4.4211770595035496e-05, | |
| "loss": 4.7158, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.6405300938707896, | |
| "grad_norm": 8.577296257019043, | |
| "learning_rate": 4.3768758328140705e-05, | |
| "loss": 4.5859, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6427388183324131, | |
| "grad_norm": 9.4404878616333, | |
| "learning_rate": 4.3326242409079705e-05, | |
| "loss": 4.8293, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.6449475427940364, | |
| "grad_norm": 9.113490104675293, | |
| "learning_rate": 4.288425808633575e-05, | |
| "loss": 4.9918, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.6471562672556599, | |
| "grad_norm": 9.25007438659668, | |
| "learning_rate": 4.2442840566047964e-05, | |
| "loss": 4.7895, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.6493649917172832, | |
| "grad_norm": 9.28573226928711, | |
| "learning_rate": 4.2002025009206994e-05, | |
| "loss": 4.7842, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.6515737161789067, | |
| "grad_norm": 9.722284317016602, | |
| "learning_rate": 4.1561846528854214e-05, | |
| "loss": 4.6345, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6537824406405301, | |
| "grad_norm": 9.87505054473877, | |
| "learning_rate": 4.1122340187284846e-05, | |
| "loss": 4.5099, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.6559911651021535, | |
| "grad_norm": 10.381636619567871, | |
| "learning_rate": 4.068354099325511e-05, | |
| "loss": 4.8878, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.658199889563777, | |
| "grad_norm": 10.358430862426758, | |
| "learning_rate": 4.0245483899193595e-05, | |
| "loss": 4.7117, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.6604086140254003, | |
| "grad_norm": 10.86618423461914, | |
| "learning_rate": 3.980820379841708e-05, | |
| "loss": 4.8503, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.6626173384870238, | |
| "grad_norm": 11.921982765197754, | |
| "learning_rate": 3.937173552235117e-05, | |
| "loss": 5.1041, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6648260629486472, | |
| "grad_norm": 4.309291362762451, | |
| "learning_rate": 3.893611383775582e-05, | |
| "loss": 3.8454, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.6670347874102706, | |
| "grad_norm": 5.9008097648620605, | |
| "learning_rate": 3.850137344395598e-05, | |
| "loss": 4.865, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.669243511871894, | |
| "grad_norm": 5.888951778411865, | |
| "learning_rate": 3.806754897007759e-05, | |
| "loss": 4.3102, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.6714522363335174, | |
| "grad_norm": 6.510290145874023, | |
| "learning_rate": 3.763467497228922e-05, | |
| "loss": 4.4619, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.6736609607951408, | |
| "grad_norm": 6.538812637329102, | |
| "learning_rate": 3.720278593104963e-05, | |
| "loss": 4.5339, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6758696852567642, | |
| "grad_norm": 6.5220136642456055, | |
| "learning_rate": 3.677191624836106e-05, | |
| "loss": 4.7157, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.6780784097183876, | |
| "grad_norm": 6.203037261962891, | |
| "learning_rate": 3.634210024502903e-05, | |
| "loss": 4.4569, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.6802871341800111, | |
| "grad_norm": 6.234178066253662, | |
| "learning_rate": 3.591337215792852e-05, | |
| "loss": 4.6308, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.6824958586416344, | |
| "grad_norm": 6.102400302886963, | |
| "learning_rate": 3.5485766137276894e-05, | |
| "loss": 4.6082, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.6847045831032579, | |
| "grad_norm": 6.148480415344238, | |
| "learning_rate": 3.5059316243913554e-05, | |
| "loss": 4.6146, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6869133075648812, | |
| "grad_norm": 5.826088905334473, | |
| "learning_rate": 3.463405644658695e-05, | |
| "loss": 4.6873, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.6891220320265047, | |
| "grad_norm": 5.779773235321045, | |
| "learning_rate": 3.421002061924876e-05, | |
| "loss": 4.452, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.6913307564881281, | |
| "grad_norm": 5.726971626281738, | |
| "learning_rate": 3.378724253835564e-05, | |
| "loss": 4.0362, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.6935394809497515, | |
| "grad_norm": 5.8791422843933105, | |
| "learning_rate": 3.336575588017881e-05, | |
| "loss": 4.5656, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.6957482054113749, | |
| "grad_norm": 5.984518527984619, | |
| "learning_rate": 3.294559421812158e-05, | |
| "loss": 4.2669, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6979569298729983, | |
| "grad_norm": 6.276485919952393, | |
| "learning_rate": 3.2526791020045086e-05, | |
| "loss": 4.521, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.7001656543346217, | |
| "grad_norm": 6.756010055541992, | |
| "learning_rate": 3.210937964560234e-05, | |
| "loss": 4.2707, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.7023743787962452, | |
| "grad_norm": 6.564029216766357, | |
| "learning_rate": 3.1693393343581044e-05, | |
| "loss": 4.0805, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.7045831032578685, | |
| "grad_norm": 6.940414905548096, | |
| "learning_rate": 3.127886524925515e-05, | |
| "loss": 4.4171, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.706791827719492, | |
| "grad_norm": 7.548460006713867, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 4.8604, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.7090005521811154, | |
| "grad_norm": 7.360044002532959, | |
| "learning_rate": 3.045431564138969e-05, | |
| "loss": 4.6096, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.7112092766427388, | |
| "grad_norm": 8.025870323181152, | |
| "learning_rate": 3.004435980712129e-05, | |
| "loss": 4.4699, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.7134180011043623, | |
| "grad_norm": 7.835132598876953, | |
| "learning_rate": 2.9635993533859062e-05, | |
| "loss": 4.7898, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.7156267255659856, | |
| "grad_norm": 7.707442760467529, | |
| "learning_rate": 2.9229249349905684e-05, | |
| "loss": 4.6073, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.7178354500276091, | |
| "grad_norm": 7.932573318481445, | |
| "learning_rate": 2.8824159654356732e-05, | |
| "loss": 4.0933, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7200441744892324, | |
| "grad_norm": 8.290323257446289, | |
| "learning_rate": 2.842075671451996e-05, | |
| "loss": 4.4136, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.7222528989508559, | |
| "grad_norm": 8.424840927124023, | |
| "learning_rate": 2.801907266334516e-05, | |
| "loss": 4.5087, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.7244616234124793, | |
| "grad_norm": 8.157886505126953, | |
| "learning_rate": 2.7619139496864378e-05, | |
| "loss": 4.3176, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.7266703478741027, | |
| "grad_norm": 8.439778327941895, | |
| "learning_rate": 2.7220989071643533e-05, | |
| "loss": 4.386, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.7288790723357261, | |
| "grad_norm": 8.862435340881348, | |
| "learning_rate": 2.682465310224473e-05, | |
| "loss": 4.922, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7310877967973495, | |
| "grad_norm": 8.862676620483398, | |
| "learning_rate": 2.6430163158700115e-05, | |
| "loss": 4.3482, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.7332965212589729, | |
| "grad_norm": 8.734305381774902, | |
| "learning_rate": 2.603755066399718e-05, | |
| "loss": 4.5575, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.7355052457205964, | |
| "grad_norm": 8.912976264953613, | |
| "learning_rate": 2.5646846891575637e-05, | |
| "loss": 4.3409, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.7377139701822197, | |
| "grad_norm": 8.8455810546875, | |
| "learning_rate": 2.5258082962836614e-05, | |
| "loss": 4.3128, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.7399226946438432, | |
| "grad_norm": 9.515101432800293, | |
| "learning_rate": 2.4871289844663437e-05, | |
| "loss": 4.9582, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7421314191054665, | |
| "grad_norm": 9.195880889892578, | |
| "learning_rate": 2.4486498346955027e-05, | |
| "loss": 4.9631, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.74434014356709, | |
| "grad_norm": 8.75934886932373, | |
| "learning_rate": 2.410373912017185e-05, | |
| "loss": 4.5848, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.7465488680287135, | |
| "grad_norm": 9.588723182678223, | |
| "learning_rate": 2.3723042652894362e-05, | |
| "loss": 4.3137, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.7487575924903368, | |
| "grad_norm": 9.760534286499023, | |
| "learning_rate": 2.334443926939448e-05, | |
| "loss": 4.7105, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.7509663169519603, | |
| "grad_norm": 9.867559432983398, | |
| "learning_rate": 2.296795912722014e-05, | |
| "loss": 4.7552, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7531750414135836, | |
| "grad_norm": 10.123164176940918, | |
| "learning_rate": 2.2593632214792994e-05, | |
| "loss": 4.7314, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.7553837658752071, | |
| "grad_norm": 10.126588821411133, | |
| "learning_rate": 2.2221488349019903e-05, | |
| "loss": 4.284, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.7575924903368305, | |
| "grad_norm": 10.301033973693848, | |
| "learning_rate": 2.1851557172917624e-05, | |
| "loss": 4.4204, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.7598012147984539, | |
| "grad_norm": 9.673257827758789, | |
| "learning_rate": 2.148386815325179e-05, | |
| "loss": 4.5415, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.7620099392600773, | |
| "grad_norm": 10.736601829528809, | |
| "learning_rate": 2.111845057818972e-05, | |
| "loss": 4.5998, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7642186637217007, | |
| "grad_norm": 10.613544464111328, | |
| "learning_rate": 2.075533355496735e-05, | |
| "loss": 4.9721, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.7664273881833241, | |
| "grad_norm": 10.81152629852295, | |
| "learning_rate": 2.0394546007570857e-05, | |
| "loss": 4.6632, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.7686361126449476, | |
| "grad_norm": 11.234309196472168, | |
| "learning_rate": 2.0036116674432654e-05, | |
| "loss": 4.6279, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.7708448371065709, | |
| "grad_norm": 11.452018737792969, | |
| "learning_rate": 1.9680074106142238e-05, | |
| "loss": 4.8806, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.7730535615681944, | |
| "grad_norm": 13.051356315612793, | |
| "learning_rate": 1.9326446663172036e-05, | |
| "loss": 4.6126, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7752622860298177, | |
| "grad_norm": 3.898439884185791, | |
| "learning_rate": 1.8975262513618293e-05, | |
| "loss": 3.3373, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.7774710104914412, | |
| "grad_norm": 5.114429473876953, | |
| "learning_rate": 1.8626549630957396e-05, | |
| "loss": 4.1765, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.7796797349530646, | |
| "grad_norm": 5.5542497634887695, | |
| "learning_rate": 1.8280335791817733e-05, | |
| "loss": 4.5676, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.781888459414688, | |
| "grad_norm": 5.338506698608398, | |
| "learning_rate": 1.7936648573766957e-05, | |
| "loss": 4.2831, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.7840971838763114, | |
| "grad_norm": 6.327260971069336, | |
| "learning_rate": 1.759551535311548e-05, | |
| "loss": 4.238, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7863059083379348, | |
| "grad_norm": 5.520748138427734, | |
| "learning_rate": 1.725696330273575e-05, | |
| "loss": 4.406, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.7885146327995582, | |
| "grad_norm": 5.61647891998291, | |
| "learning_rate": 1.6921019389897795e-05, | |
| "loss": 4.5205, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.7907233572611817, | |
| "grad_norm": 6.23562479019165, | |
| "learning_rate": 1.6587710374121203e-05, | |
| "loss": 4.5157, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.792932081722805, | |
| "grad_norm": 5.914584159851074, | |
| "learning_rate": 1.6257062805043478e-05, | |
| "loss": 4.5219, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.7951408061844285, | |
| "grad_norm": 5.929191589355469, | |
| "learning_rate": 1.592910302030544e-05, | |
| "loss": 4.3534, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7973495306460519, | |
| "grad_norm": 5.869836330413818, | |
| "learning_rate": 1.5603857143453128e-05, | |
| "loss": 4.2266, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.7995582551076753, | |
| "grad_norm": 6.144409656524658, | |
| "learning_rate": 1.5281351081856974e-05, | |
| "loss": 4.194, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.8017669795692988, | |
| "grad_norm": 6.3378520011901855, | |
| "learning_rate": 1.4961610524648223e-05, | |
| "loss": 4.8193, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.8039757040309221, | |
| "grad_norm": 6.32310152053833, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 4.4801, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.8061844284925456, | |
| "grad_norm": 6.283787250518799, | |
| "learning_rate": 1.4330527576461728e-05, | |
| "loss": 4.4156, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.8083931529541689, | |
| "grad_norm": 6.670344352722168, | |
| "learning_rate": 1.4019235454221857e-05, | |
| "loss": 4.0886, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.8106018774157924, | |
| "grad_norm": 6.906120300292969, | |
| "learning_rate": 1.371080936984101e-05, | |
| "loss": 4.4984, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.8128106018774158, | |
| "grad_norm": 7.01516580581665, | |
| "learning_rate": 1.340527389091374e-05, | |
| "loss": 4.2786, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.8150193263390392, | |
| "grad_norm": 7.165297508239746, | |
| "learning_rate": 1.3102653354784177e-05, | |
| "loss": 4.765, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.8172280508006626, | |
| "grad_norm": 7.451290130615234, | |
| "learning_rate": 1.2802971866607521e-05, | |
| "loss": 4.4154, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.819436775262286, | |
| "grad_norm": 7.890896320343018, | |
| "learning_rate": 1.250625329742996e-05, | |
| "loss": 4.6839, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.8216454997239094, | |
| "grad_norm": 8.337573051452637, | |
| "learning_rate": 1.2212521282287092e-05, | |
| "loss": 4.0614, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.8238542241855329, | |
| "grad_norm": 7.795760154724121, | |
| "learning_rate": 1.192179921832145e-05, | |
| "loss": 4.3992, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.8260629486471562, | |
| "grad_norm": 7.948381423950195, | |
| "learning_rate": 1.1634110262918718e-05, | |
| "loss": 4.4898, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.8282716731087797, | |
| "grad_norm": 7.649637222290039, | |
| "learning_rate": 1.134947733186315e-05, | |
| "loss": 4.6223, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.830480397570403, | |
| "grad_norm": 8.012591361999512, | |
| "learning_rate": 1.1067923097512256e-05, | |
| "loss": 4.696, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.8326891220320265, | |
| "grad_norm": 8.13010025024414, | |
| "learning_rate": 1.078946998699073e-05, | |
| "loss": 4.3658, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.83489784649365, | |
| "grad_norm": 8.306947708129883, | |
| "learning_rate": 1.0514140180404204e-05, | |
| "loss": 4.9003, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.8371065709552733, | |
| "grad_norm": 8.446629524230957, | |
| "learning_rate": 1.0241955609072358e-05, | |
| "loss": 4.9169, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.8393152954168968, | |
| "grad_norm": 8.176511764526367, | |
| "learning_rate": 9.972937953781986e-06, | |
| "loss": 4.1928, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8415240198785201, | |
| "grad_norm": 8.192519187927246, | |
| "learning_rate": 9.707108643060115e-06, | |
| "loss": 4.0346, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.8437327443401436, | |
| "grad_norm": 8.429787635803223, | |
| "learning_rate": 9.444488851467042e-06, | |
| "loss": 4.3153, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.845941468801767, | |
| "grad_norm": 8.412949562072754, | |
| "learning_rate": 9.185099497909716e-06, | |
| "loss": 4.5692, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.8481501932633904, | |
| "grad_norm": 8.446182250976562, | |
| "learning_rate": 8.928961243975437e-06, | |
| "loss": 4.6245, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.8503589177250138, | |
| "grad_norm": 9.340173721313477, | |
| "learning_rate": 8.676094492286013e-06, | |
| "loss": 4.5607, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8525676421866372, | |
| "grad_norm": 8.916312217712402, | |
| "learning_rate": 8.426519384872733e-06, | |
| "loss": 4.6569, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.8547763666482606, | |
| "grad_norm": 8.926466941833496, | |
| "learning_rate": 8.180255801571824e-06, | |
| "loss": 4.8406, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.8569850911098841, | |
| "grad_norm": 8.984560012817383, | |
| "learning_rate": 7.937323358440935e-06, | |
| "loss": 4.5565, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.8591938155715074, | |
| "grad_norm": 9.738987922668457, | |
| "learning_rate": 7.697741406196757e-06, | |
| "loss": 4.5395, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.8614025400331309, | |
| "grad_norm": 9.285807609558105, | |
| "learning_rate": 7.461529028673464e-06, | |
| "loss": 4.595, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8636112644947542, | |
| "grad_norm": 9.441668510437012, | |
| "learning_rate": 7.228705041302741e-06, | |
| "loss": 4.5406, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.8658199889563777, | |
| "grad_norm": 9.5773344039917, | |
| "learning_rate": 6.999287989614972e-06, | |
| "loss": 3.9756, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.8680287134180011, | |
| "grad_norm": 9.876778602600098, | |
| "learning_rate": 6.773296147762037e-06, | |
| "loss": 4.7135, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.8702374378796245, | |
| "grad_norm": 10.206405639648438, | |
| "learning_rate": 6.5507475170616565e-06, | |
| "loss": 4.7689, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.8724461623412479, | |
| "grad_norm": 9.651083946228027, | |
| "learning_rate": 6.331659824563513e-06, | |
| "loss": 4.3547, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8746548868028713, | |
| "grad_norm": 9.921658515930176, | |
| "learning_rate": 6.116050521637218e-06, | |
| "loss": 4.5168, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.8768636112644947, | |
| "grad_norm": 11.268795013427734, | |
| "learning_rate": 5.903936782582253e-06, | |
| "loss": 4.4979, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.8790723357261182, | |
| "grad_norm": 10.472535133361816, | |
| "learning_rate": 5.69533550325988e-06, | |
| "loss": 4.4414, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.8812810601877415, | |
| "grad_norm": 11.554786682128906, | |
| "learning_rate": 5.4902632997473665e-06, | |
| "loss": 4.6124, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.883489784649365, | |
| "grad_norm": 13.761305809020996, | |
| "learning_rate": 5.288736507014435e-06, | |
| "loss": 5.3981, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8856985091109884, | |
| "grad_norm": 4.258291244506836, | |
| "learning_rate": 5.090771177622083e-06, | |
| "loss": 3.1499, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.8879072335726118, | |
| "grad_norm": 4.202602386474609, | |
| "learning_rate": 4.896383080443934e-06, | |
| "loss": 3.9556, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.8901159580342353, | |
| "grad_norm": 5.0006513595581055, | |
| "learning_rate": 4.705587699410136e-06, | |
| "loss": 4.307, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.8923246824958586, | |
| "grad_norm": 4.940323829650879, | |
| "learning_rate": 4.5184002322740785e-06, | |
| "loss": 4.4528, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.8945334069574821, | |
| "grad_norm": 5.05031681060791, | |
| "learning_rate": 4.3348355894017424e-06, | |
| "loss": 4.0433, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8967421314191054, | |
| "grad_norm": 5.438938617706299, | |
| "learning_rate": 4.154908392584017e-06, | |
| "loss": 4.2909, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.8989508558807289, | |
| "grad_norm": 5.099344253540039, | |
| "learning_rate": 3.978632973872065e-06, | |
| "loss": 4.1406, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.9011595803423523, | |
| "grad_norm": 5.504106521606445, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 4.3558, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.9033683048039757, | |
| "grad_norm": 5.649275302886963, | |
| "learning_rate": 3.6370933434447698e-06, | |
| "loss": 4.4172, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.9055770292655991, | |
| "grad_norm": 5.748392105102539, | |
| "learning_rate": 3.471856336974322e-06, | |
| "loss": 4.1043, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.9077857537272225, | |
| "grad_norm": 5.7138237953186035, | |
| "learning_rate": 3.310325516932422e-06, | |
| "loss": 4.5152, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.9099944781888459, | |
| "grad_norm": 5.844823837280273, | |
| "learning_rate": 3.1525137500119207e-06, | |
| "loss": 4.3754, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.9122032026504694, | |
| "grad_norm": 5.811467170715332, | |
| "learning_rate": 2.998433606665496e-06, | |
| "loss": 4.4152, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.9144119271120927, | |
| "grad_norm": 5.769085884094238, | |
| "learning_rate": 2.848097360104396e-06, | |
| "loss": 4.4248, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.9166206515737162, | |
| "grad_norm": 6.123106479644775, | |
| "learning_rate": 2.7015169853208235e-06, | |
| "loss": 4.4335, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.9188293760353395, | |
| "grad_norm": 6.143897533416748, | |
| "learning_rate": 2.5587041581340233e-06, | |
| "loss": 4.4991, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.921038100496963, | |
| "grad_norm": 6.4361677169799805, | |
| "learning_rate": 2.419670254260281e-06, | |
| "loss": 4.3968, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.9232468249585865, | |
| "grad_norm": 6.608926773071289, | |
| "learning_rate": 2.2844263484068096e-06, | |
| "loss": 4.4498, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.9254555494202098, | |
| "grad_norm": 7.144963264465332, | |
| "learning_rate": 2.152983213389559e-06, | |
| "loss": 4.6328, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.9276642738818333, | |
| "grad_norm": 7.143801212310791, | |
| "learning_rate": 2.0253513192751373e-06, | |
| "loss": 4.4744, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9298729983434566, | |
| "grad_norm": 7.323301315307617, | |
| "learning_rate": 1.9015408325467888e-06, | |
| "loss": 4.4851, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.9320817228050801, | |
| "grad_norm": 7.94163179397583, | |
| "learning_rate": 1.7815616152946523e-06, | |
| "loss": 4.705, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.9342904472667035, | |
| "grad_norm": 7.78623628616333, | |
| "learning_rate": 1.6654232244301294e-06, | |
| "loss": 4.4493, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.9364991717283269, | |
| "grad_norm": 7.809777736663818, | |
| "learning_rate": 1.5531349109246362e-06, | |
| "loss": 4.6519, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.9387078961899503, | |
| "grad_norm": 7.850180149078369, | |
| "learning_rate": 1.4447056190727725e-06, | |
| "loss": 4.3355, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9409166206515737, | |
| "grad_norm": 8.725530624389648, | |
| "learning_rate": 1.3401439857798292e-06, | |
| "loss": 4.4763, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.9431253451131971, | |
| "grad_norm": 7.9435834884643555, | |
| "learning_rate": 1.2394583398738114e-06, | |
| "loss": 4.0902, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.9453340695748206, | |
| "grad_norm": 8.458505630493164, | |
| "learning_rate": 1.1426567014420297e-06, | |
| "loss": 4.4556, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.9475427940364439, | |
| "grad_norm": 8.368391036987305, | |
| "learning_rate": 1.0497467811922346e-06, | |
| "loss": 4.6413, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.9497515184980674, | |
| "grad_norm": 8.806220054626465, | |
| "learning_rate": 9.607359798384785e-07, | |
| "loss": 4.2521, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9519602429596907, | |
| "grad_norm": 8.404871940612793, | |
| "learning_rate": 8.756313875115385e-07, | |
| "loss": 4.2418, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.9541689674213142, | |
| "grad_norm": 8.731966018676758, | |
| "learning_rate": 7.944397831941952e-07, | |
| "loss": 4.5369, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.9563776918829376, | |
| "grad_norm": 8.71829605102539, | |
| "learning_rate": 7.171676341812706e-07, | |
| "loss": 4.7399, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.958586416344561, | |
| "grad_norm": 8.327555656433105, | |
| "learning_rate": 6.438210955644452e-07, | |
| "loss": 4.5096, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.9607951408061844, | |
| "grad_norm": 8.708925247192383, | |
| "learning_rate": 5.74406009741979e-07, | |
| "loss": 4.5694, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9630038652678078, | |
| "grad_norm": 9.103087425231934, | |
| "learning_rate": 5.089279059533658e-07, | |
| "loss": 4.4516, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.9652125897294312, | |
| "grad_norm": 8.804319381713867, | |
| "learning_rate": 4.4739199983886425e-07, | |
| "loss": 4.4854, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.9674213141910547, | |
| "grad_norm": 8.990171432495117, | |
| "learning_rate": 3.8980319302407977e-07, | |
| "loss": 4.7322, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.969630038652678, | |
| "grad_norm": 9.266417503356934, | |
| "learning_rate": 3.361660727295046e-07, | |
| "loss": 4.8561, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.9718387631143015, | |
| "grad_norm": 9.650453567504883, | |
| "learning_rate": 2.8648491140513266e-07, | |
| "loss": 5.1407, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9740474875759249, | |
| "grad_norm": 9.014724731445312, | |
| "learning_rate": 2.407636663901591e-07, | |
| "loss": 4.7725, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.9762562120375483, | |
| "grad_norm": 9.816120147705078, | |
| "learning_rate": 1.9900597959770507e-07, | |
| "loss": 4.3926, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.9784649364991718, | |
| "grad_norm": 9.809698104858398, | |
| "learning_rate": 1.612151772247883e-07, | |
| "loss": 4.1767, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.9806736609607951, | |
| "grad_norm": 9.744946479797363, | |
| "learning_rate": 1.2739426948732424e-07, | |
| "loss": 4.2795, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.9828823854224186, | |
| "grad_norm": 9.93597412109375, | |
| "learning_rate": 9.754595038037328e-08, | |
| "loss": 4.1773, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9850911098840419, | |
| "grad_norm": 10.203986167907715, | |
| "learning_rate": 7.167259746355681e-08, | |
| "loss": 4.3431, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.9872998343456654, | |
| "grad_norm": 9.65555191040039, | |
| "learning_rate": 4.9776271671642115e-08, | |
| "loss": 4.5657, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.9895085588072888, | |
| "grad_norm": 11.336668014526367, | |
| "learning_rate": 3.185871715041255e-08, | |
| "loss": 5.1661, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.9917172832689122, | |
| "grad_norm": 11.222773551940918, | |
| "learning_rate": 1.7921361117734236e-08, | |
| "loss": 4.8351, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.9939260077305356, | |
| "grad_norm": 13.293135643005371, | |
| "learning_rate": 7.96531374983589e-09, | |
| "loss": 4.995, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.996134732192159, | |
| "grad_norm": 4.857739448547363, | |
| "learning_rate": 1.9913680929295465e-09, | |
| "loss": 3.7627, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.9983434566537824, | |
| "grad_norm": 8.479840278625488, | |
| "learning_rate": 0.0, | |
| "loss": 4.7239, | |
| "step": 452 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 452, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 239, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.7098768232585626e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |