| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.1235955056179776, | |
| "eval_steps": 500, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 4.5, | |
| "learning_rate": 0.00019902985447817174, | |
| "loss": 2.5508, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 5.5, | |
| "learning_rate": 0.00019802970445566834, | |
| "loss": 2.7598, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.0625, | |
| "learning_rate": 0.00019702955443316497, | |
| "loss": 2.4902, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.84375, | |
| "learning_rate": 0.0001960294044106616, | |
| "loss": 2.4863, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 0.00019502925438815824, | |
| "loss": 3.25, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 0.00019402910436565487, | |
| "loss": 3.0195, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 8.875, | |
| "learning_rate": 0.0001930289543431515, | |
| "loss": 3.6758, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 0.0001920288043206481, | |
| "loss": 2.1699, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 0.00019102865429814473, | |
| "loss": 2.4805, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 4.625, | |
| "learning_rate": 0.00019002850427564136, | |
| "loss": 2.6797, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 0.000189028354253138, | |
| "loss": 2.2031, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 4.78125, | |
| "learning_rate": 0.0001880282042306346, | |
| "loss": 2.4688, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 0.00018702805420813123, | |
| "loss": 2.668, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 0.00018602790418562786, | |
| "loss": 2.8516, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 0.0001850277541631245, | |
| "loss": 2.3086, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 4.5, | |
| "learning_rate": 0.0001840276041406211, | |
| "loss": 2.1504, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 6.0, | |
| "learning_rate": 0.00018302745411811772, | |
| "loss": 2.5664, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 0.00018202730409561436, | |
| "loss": 2.4023, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 0.00018102715407311096, | |
| "loss": 2.6562, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 0.0001800270040506076, | |
| "loss": 1.9863, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 0.00017902685402810422, | |
| "loss": 2.4648, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 6.375, | |
| "learning_rate": 0.00017802670400560085, | |
| "loss": 2.5781, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 0.00017702655398309746, | |
| "loss": 2.2715, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 8.125, | |
| "learning_rate": 0.00017602640396059409, | |
| "loss": 2.5195, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 0.00017502625393809074, | |
| "loss": 2.4141, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 0.00017402610391558735, | |
| "loss": 2.2051, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 0.00017302595389308398, | |
| "loss": 2.5781, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 0.0001720258038705806, | |
| "loss": 2.2676, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 0.00017102565384807724, | |
| "loss": 2.291, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 0.00017002550382557384, | |
| "loss": 1.8398, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 0.00016902535380307047, | |
| "loss": 2.1543, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 0.0001680252037805671, | |
| "loss": 2.8789, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 5.5, | |
| "learning_rate": 0.0001670250537580637, | |
| "loss": 2.1621, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 5.28125, | |
| "learning_rate": 0.00016602490373556034, | |
| "loss": 2.2969, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 0.00016502475371305697, | |
| "loss": 2.4668, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 0.0001640246036905536, | |
| "loss": 2.7734, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 5.5, | |
| "learning_rate": 0.0001630244536680502, | |
| "loss": 1.9102, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 9.75, | |
| "learning_rate": 0.00016202430364554684, | |
| "loss": 2.6055, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 0.00016102415362304347, | |
| "loss": 2.0996, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 0.0001600240036005401, | |
| "loss": 2.5117, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 8.125, | |
| "learning_rate": 0.0001590238535780367, | |
| "loss": 2.3262, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 7.375, | |
| "learning_rate": 0.00015802370355553333, | |
| "loss": 2.4004, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 6.75, | |
| "learning_rate": 0.00015702355353302996, | |
| "loss": 1.7539, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 0.00015602340351052657, | |
| "loss": 2.1699, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 0.0001550232534880232, | |
| "loss": 2.418, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 0.00015402310346551986, | |
| "loss": 2.1543, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 6.875, | |
| "learning_rate": 0.00015302295344301646, | |
| "loss": 2.2812, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 0.0001520228034205131, | |
| "loss": 2.2422, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 0.00015102265339800972, | |
| "loss": 2.7344, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 0.00015002250337550635, | |
| "loss": 2.1504, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 8.25, | |
| "learning_rate": 0.00014902235335300296, | |
| "loss": 2.6289, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 5.84375, | |
| "learning_rate": 0.00014802220333049959, | |
| "loss": 2.0312, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 9.75, | |
| "learning_rate": 0.00014702205330799622, | |
| "loss": 2.1367, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 0.00014602190328549282, | |
| "loss": 1.7891, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 0.00014502175326298945, | |
| "loss": 2.2305, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 9.625, | |
| "learning_rate": 0.00014402160324048608, | |
| "loss": 2.4395, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 0.0001430214532179827, | |
| "loss": 2.1445, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 0.00014202130319547932, | |
| "loss": 2.5586, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 0.00014102115317297595, | |
| "loss": 2.7031, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 0.00014002100315047258, | |
| "loss": 2.373, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 8.375, | |
| "learning_rate": 0.0001390208531279692, | |
| "loss": 2.582, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 6.75, | |
| "learning_rate": 0.0001380207031054658, | |
| "loss": 1.9961, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 0.00013702055308296244, | |
| "loss": 2.4199, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 0.00013602040306045908, | |
| "loss": 2.875, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 0.00013502025303795568, | |
| "loss": 2.5273, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 0.0001340201030154523, | |
| "loss": 2.2988, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 8.625, | |
| "learning_rate": 0.00013301995299294897, | |
| "loss": 2.9023, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 7.1875, | |
| "learning_rate": 0.00013201980297044557, | |
| "loss": 2.8047, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 9.625, | |
| "learning_rate": 0.0001310196529479422, | |
| "loss": 2.6328, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 0.00013001950292543883, | |
| "loss": 2.168, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 0.00012901935290293546, | |
| "loss": 2.5234, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 0.00012801920288043207, | |
| "loss": 2.8516, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 0.0001270190528579287, | |
| "loss": 2.5273, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 9.375, | |
| "learning_rate": 0.00012601890283542533, | |
| "loss": 3.0469, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 6.875, | |
| "learning_rate": 0.00012501875281292196, | |
| "loss": 2.543, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 9.25, | |
| "learning_rate": 0.00012401860279041856, | |
| "loss": 3.3125, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 6.96875, | |
| "learning_rate": 0.0001230184527679152, | |
| "loss": 2.1328, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 7.875, | |
| "learning_rate": 0.00012201830274541181, | |
| "loss": 2.875, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 8.25, | |
| "learning_rate": 0.00012101815272290844, | |
| "loss": 2.4727, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 0.00012001800270040506, | |
| "loss": 2.5703, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 0.00011901785267790169, | |
| "loss": 3.3555, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 0.00011801770265539831, | |
| "loss": 2.5352, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 5.03125, | |
| "learning_rate": 0.00011701755263289494, | |
| "loss": 2.0664, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 9.5, | |
| "learning_rate": 0.00011601740261039156, | |
| "loss": 2.9844, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 0.00011501725258788819, | |
| "loss": 2.6602, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 6.21875, | |
| "learning_rate": 0.0001140171025653848, | |
| "loss": 2.1367, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 0.00011301695254288145, | |
| "loss": 2.459, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 7.75, | |
| "learning_rate": 0.00011201680252037807, | |
| "loss": 2.1484, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 7.90625, | |
| "learning_rate": 0.0001110166524978747, | |
| "loss": 2.8867, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 0.00011001650247537131, | |
| "loss": 3.5586, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 6.375, | |
| "learning_rate": 0.00010901635245286794, | |
| "loss": 2.8594, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 0.00010801620243036456, | |
| "loss": 3.4688, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 8.5, | |
| "learning_rate": 0.00010701605240786119, | |
| "loss": 2.7031, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 10.0, | |
| "learning_rate": 0.00010601590238535781, | |
| "loss": 3.0859, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 8.5, | |
| "learning_rate": 0.00010501575236285444, | |
| "loss": 3.0352, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 0.00010401560234035106, | |
| "loss": 2.084, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 7.0, | |
| "learning_rate": 0.00010301545231784769, | |
| "loss": 2.0508, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 0.0001020153022953443, | |
| "loss": 2.9805, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 7.5625, | |
| "learning_rate": 0.00010101515227284092, | |
| "loss": 2.9766, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 8.125, | |
| "learning_rate": 0.00010001500225033755, | |
| "loss": 2.502, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 9.901485222783417e-05, | |
| "loss": 2.4844, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 9.80147022053308e-05, | |
| "loss": 2.4492, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 9.701455218282743e-05, | |
| "loss": 2.5742, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 9.601440216032405e-05, | |
| "loss": 2.4688, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 7.5625, | |
| "learning_rate": 9.501425213782068e-05, | |
| "loss": 2.6211, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 9.40141021153173e-05, | |
| "loss": 2.8027, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 9.301395209281393e-05, | |
| "loss": 2.5742, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 5.5625, | |
| "learning_rate": 9.201380207031055e-05, | |
| "loss": 2.4258, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 9.25, | |
| "learning_rate": 9.101365204780718e-05, | |
| "loss": 2.9141, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 9.00135020253038e-05, | |
| "loss": 2.2578, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 8.901335200280043e-05, | |
| "loss": 2.5312, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 8.801320198029704e-05, | |
| "loss": 2.1133, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 8.625, | |
| "learning_rate": 8.701305195779367e-05, | |
| "loss": 2.6523, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 8.60129019352903e-05, | |
| "loss": 2.9453, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 8.501275191278692e-05, | |
| "loss": 2.6816, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 8.401260189028355e-05, | |
| "loss": 2.7227, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 8.301245186778017e-05, | |
| "loss": 2.6191, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 7.5625, | |
| "learning_rate": 8.20123018452768e-05, | |
| "loss": 2.459, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 8.101215182277342e-05, | |
| "loss": 2.4297, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 8.375, | |
| "learning_rate": 8.001200180027005e-05, | |
| "loss": 2.4609, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 7.875, | |
| "learning_rate": 7.901185177776667e-05, | |
| "loss": 2.7539, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 7.801170175526328e-05, | |
| "loss": 2.2148, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 6.75, | |
| "learning_rate": 7.701155173275993e-05, | |
| "loss": 2.4512, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 7.601140171025655e-05, | |
| "loss": 2.5859, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 7.501125168775318e-05, | |
| "loss": 2.6797, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 9.125, | |
| "learning_rate": 7.401110166524979e-05, | |
| "loss": 3.168, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 7.301095164274641e-05, | |
| "loss": 2.1816, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 7.201080162024304e-05, | |
| "loss": 2.9766, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 6.625, | |
| "learning_rate": 7.101065159773966e-05, | |
| "loss": 2.1836, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 10.5, | |
| "learning_rate": 7.001050157523629e-05, | |
| "loss": 2.8086, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 6.90103515527329e-05, | |
| "loss": 2.4141, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 7.375, | |
| "learning_rate": 6.801020153022954e-05, | |
| "loss": 2.5742, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 7.46875, | |
| "learning_rate": 6.701005150772615e-05, | |
| "loss": 2.9883, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 6.600990148522279e-05, | |
| "loss": 2.418, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 6.500975146271942e-05, | |
| "loss": 2.7578, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 7.90625, | |
| "learning_rate": 6.400960144021603e-05, | |
| "loss": 3.0078, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 6.125, | |
| "learning_rate": 6.300945141771266e-05, | |
| "loss": 2.2012, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 8.0, | |
| "learning_rate": 6.200930139520928e-05, | |
| "loss": 3.0703, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 6.1009151372705906e-05, | |
| "loss": 2.7031, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 5.15625, | |
| "learning_rate": 6.000900135020253e-05, | |
| "loss": 1.9014, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 5.9008851327699154e-05, | |
| "loss": 2.3809, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 5.800870130519578e-05, | |
| "loss": 2.5469, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 6.0, | |
| "learning_rate": 5.70085512826924e-05, | |
| "loss": 3.0938, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 7.25, | |
| "learning_rate": 5.600840126018903e-05, | |
| "loss": 2.4395, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 5.500825123768566e-05, | |
| "loss": 2.5254, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 6.96875, | |
| "learning_rate": 5.400810121518228e-05, | |
| "loss": 2.7773, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 5.3007951192678905e-05, | |
| "loss": 2.6875, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 5.200780117017553e-05, | |
| "loss": 3.0898, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 5.100765114767215e-05, | |
| "loss": 2.3984, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 5.000750112516878e-05, | |
| "loss": 2.1445, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 4.90073511026654e-05, | |
| "loss": 2.8125, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 4.8007201080162025e-05, | |
| "loss": 2.3242, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 4.700705105765865e-05, | |
| "loss": 2.3965, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 4.6006901035155273e-05, | |
| "loss": 2.4824, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 4.50067510126519e-05, | |
| "loss": 2.6445, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 6.71875, | |
| "learning_rate": 4.400660099014852e-05, | |
| "loss": 2.1689, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 4.300645096764515e-05, | |
| "loss": 2.4219, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 4.2006300945141776e-05, | |
| "loss": 2.7227, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 6.25, | |
| "learning_rate": 4.10061509226384e-05, | |
| "loss": 2.7812, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 5.5, | |
| "learning_rate": 4.0006000900135024e-05, | |
| "loss": 2.3574, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 3.900585087763164e-05, | |
| "loss": 2.6016, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 3.800570085512827e-05, | |
| "loss": 2.7617, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 3.7005550832624897e-05, | |
| "loss": 2.2285, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 7.75, | |
| "learning_rate": 3.600540081012152e-05, | |
| "loss": 3.0, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 3.5005250787618145e-05, | |
| "loss": 2.6172, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 3.400510076511477e-05, | |
| "loss": 2.3242, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 6.25, | |
| "learning_rate": 3.300495074261139e-05, | |
| "loss": 2.0098, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 7.25, | |
| "learning_rate": 3.200480072010802e-05, | |
| "loss": 1.9238, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 5.84375, | |
| "learning_rate": 3.100465069760464e-05, | |
| "loss": 2.4082, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 5.0625, | |
| "learning_rate": 3.0004500675101265e-05, | |
| "loss": 1.8848, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 5.28125, | |
| "learning_rate": 2.900435065259789e-05, | |
| "loss": 2.293, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 8.5, | |
| "learning_rate": 2.8004200630094516e-05, | |
| "loss": 3.332, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 5.84375, | |
| "learning_rate": 2.700405060759114e-05, | |
| "loss": 2.1543, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 9.0, | |
| "learning_rate": 2.6003900585087765e-05, | |
| "loss": 3.0156, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 2.500375056258439e-05, | |
| "loss": 2.4844, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 2.4003600540081013e-05, | |
| "loss": 2.4414, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 2.3003450517577637e-05, | |
| "loss": 2.584, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 2.200330049507426e-05, | |
| "loss": 2.9062, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 2.1003150472570888e-05, | |
| "loss": 2.127, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 5.78125, | |
| "learning_rate": 2.0003000450067512e-05, | |
| "loss": 2.3066, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 6.75, | |
| "learning_rate": 1.9002850427564136e-05, | |
| "loss": 1.6729, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 1.800270040506076e-05, | |
| "loss": 2.0488, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 5.53125, | |
| "learning_rate": 1.7002550382557384e-05, | |
| "loss": 1.5322, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 6.125, | |
| "learning_rate": 1.600240036005401e-05, | |
| "loss": 1.4639, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 6.8125, | |
| "learning_rate": 1.5002250337550632e-05, | |
| "loss": 2.5938, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 1.4002100315047258e-05, | |
| "loss": 1.7109, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 5.875, | |
| "learning_rate": 1.3001950292543882e-05, | |
| "loss": 1.6211, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 6.25, | |
| "learning_rate": 1.2001800270040506e-05, | |
| "loss": 2.0449, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 1.100165024753713e-05, | |
| "loss": 1.6562, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 1.0001500225033756e-05, | |
| "loss": 2.041, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 5.5, | |
| "learning_rate": 9.00135020253038e-06, | |
| "loss": 1.8105, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 4.875, | |
| "learning_rate": 8.001200180027004e-06, | |
| "loss": 1.916, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 5.75, | |
| "learning_rate": 7.001050157523629e-06, | |
| "loss": 1.7607, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 6.000900135020253e-06, | |
| "loss": 1.6367, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 5.0, | |
| "learning_rate": 5.000750112516878e-06, | |
| "loss": 2.0332, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 4.000600090013502e-06, | |
| "loss": 1.7188, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 5.84375, | |
| "learning_rate": 3.0004500675101266e-06, | |
| "loss": 1.832, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 6.78125, | |
| "learning_rate": 2.000300045006751e-06, | |
| "loss": 2.0049, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 1.0001500225033755e-06, | |
| "loss": 1.6953, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 6.875, | |
| "learning_rate": 0.0, | |
| "loss": 1.8525, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "total_flos": 509004324495360.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |