| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9984871406959153, | |
| "eval_steps": 83, | |
| "global_step": 330, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0030257186081694403, | |
| "grad_norm": 1226236486586.7495, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.415, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.006051437216338881, | |
| "grad_norm": 1054830503169.1672, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.4114, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.009077155824508321, | |
| "grad_norm": 1939.9283815132105, | |
| "learning_rate": 6e-06, | |
| "loss": 1.4317, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.012102874432677761, | |
| "grad_norm": 459.3036525279639, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.4001, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.015128593040847202, | |
| "grad_norm": 182.80378684092744, | |
| "learning_rate": 1e-05, | |
| "loss": 1.3109, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.018154311649016642, | |
| "grad_norm": 142.7771377398994, | |
| "learning_rate": 1.2e-05, | |
| "loss": 1.3535, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.02118003025718608, | |
| "grad_norm": 14843.66613495735, | |
| "learning_rate": 1.4e-05, | |
| "loss": 1.1695, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.024205748865355523, | |
| "grad_norm": 39.59077433599819, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 0.8874, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.02723146747352496, | |
| "grad_norm": 110.02627986509732, | |
| "learning_rate": 1.8e-05, | |
| "loss": 0.9144, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.030257186081694403, | |
| "grad_norm": 26.57084146682424, | |
| "learning_rate": 2e-05, | |
| "loss": 0.9139, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03328290468986384, | |
| "grad_norm": 24.876997526308156, | |
| "learning_rate": 1.999951808959328e-05, | |
| "loss": 0.8463, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.036308623298033284, | |
| "grad_norm": 8.554621464049323, | |
| "learning_rate": 1.9998072404820648e-05, | |
| "loss": 0.7976, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.039334341906202726, | |
| "grad_norm": 89.40611753938879, | |
| "learning_rate": 1.9995663085020215e-05, | |
| "loss": 0.7918, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.04236006051437216, | |
| "grad_norm": 5.5056163156163365, | |
| "learning_rate": 1.9992290362407232e-05, | |
| "loss": 0.8059, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0453857791225416, | |
| "grad_norm": 5.01374788164094, | |
| "learning_rate": 1.9987954562051724e-05, | |
| "loss": 0.7991, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.048411497730711045, | |
| "grad_norm": 6.204198483647409, | |
| "learning_rate": 1.998265610184716e-05, | |
| "loss": 0.8032, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.05143721633888049, | |
| "grad_norm": 4.638968016768252, | |
| "learning_rate": 1.997639549247016e-05, | |
| "loss": 0.6983, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.05446293494704992, | |
| "grad_norm": 5.739965915947546, | |
| "learning_rate": 1.9969173337331283e-05, | |
| "loss": 0.7624, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.057488653555219364, | |
| "grad_norm": 2.7754505834252616, | |
| "learning_rate": 1.9960990332516875e-05, | |
| "loss": 0.7911, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.060514372163388806, | |
| "grad_norm": 11.166370321864145, | |
| "learning_rate": 1.995184726672197e-05, | |
| "loss": 0.7832, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06354009077155824, | |
| "grad_norm": 2.931204688651764, | |
| "learning_rate": 1.9941745021174284e-05, | |
| "loss": 0.7501, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.06656580937972768, | |
| "grad_norm": 2.0799413444232266, | |
| "learning_rate": 1.9930684569549265e-05, | |
| "loss": 0.7217, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.06959152798789713, | |
| "grad_norm": 2.3639993655231124, | |
| "learning_rate": 1.991866697787626e-05, | |
| "loss": 0.7942, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.07261724659606657, | |
| "grad_norm": 2.0945628241773657, | |
| "learning_rate": 1.990569340443577e-05, | |
| "loss": 0.7571, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.07564296520423601, | |
| "grad_norm": 1.8595656724210274, | |
| "learning_rate": 1.989176509964781e-05, | |
| "loss": 0.7207, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.07866868381240545, | |
| "grad_norm": 2.0672258707048354, | |
| "learning_rate": 1.9876883405951378e-05, | |
| "loss": 0.7039, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.08169440242057488, | |
| "grad_norm": 1.9192654802975393, | |
| "learning_rate": 1.9861049757675087e-05, | |
| "loss": 0.705, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.08472012102874432, | |
| "grad_norm": 1.7928257778342083, | |
| "learning_rate": 1.9844265680898917e-05, | |
| "loss": 0.7165, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.08774583963691376, | |
| "grad_norm": 2.1055042932269172, | |
| "learning_rate": 1.982653279330712e-05, | |
| "loss": 0.7285, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0907715582450832, | |
| "grad_norm": 2.227087071398814, | |
| "learning_rate": 1.9807852804032306e-05, | |
| "loss": 0.7241, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.09379727685325265, | |
| "grad_norm": 1.9268416294444393, | |
| "learning_rate": 1.9788227513490724e-05, | |
| "loss": 0.7079, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.09682299546142209, | |
| "grad_norm": 1.838940648447788, | |
| "learning_rate": 1.9767658813208725e-05, | |
| "loss": 0.6868, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.09984871406959153, | |
| "grad_norm": 1.8177983496932915, | |
| "learning_rate": 1.974614868564045e-05, | |
| "loss": 0.7, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.10287443267776097, | |
| "grad_norm": 1.8667341435101743, | |
| "learning_rate": 1.9723699203976768e-05, | |
| "loss": 0.7127, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.1059001512859304, | |
| "grad_norm": 1.7158856644485938, | |
| "learning_rate": 1.9700312531945444e-05, | |
| "loss": 0.6716, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.10892586989409984, | |
| "grad_norm": 1.6969998808568116, | |
| "learning_rate": 1.96759909236026e-05, | |
| "loss": 0.7097, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.11195158850226929, | |
| "grad_norm": 1.7083400048776565, | |
| "learning_rate": 1.9650736723115476e-05, | |
| "loss": 0.6875, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.11497730711043873, | |
| "grad_norm": 1.4880068549977044, | |
| "learning_rate": 1.9624552364536472e-05, | |
| "loss": 0.698, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.11800302571860817, | |
| "grad_norm": 1.5881964679952696, | |
| "learning_rate": 1.9597440371568576e-05, | |
| "loss": 0.6736, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.12102874432677761, | |
| "grad_norm": 1.682849873860893, | |
| "learning_rate": 1.956940335732209e-05, | |
| "loss": 0.6703, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.12405446293494705, | |
| "grad_norm": 1.578678343514631, | |
| "learning_rate": 1.9540444024062807e-05, | |
| "loss": 0.6706, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.12708018154311648, | |
| "grad_norm": 1.5462143076031016, | |
| "learning_rate": 1.9510565162951538e-05, | |
| "loss": 0.723, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.13010590015128592, | |
| "grad_norm": 1.4267236744974217, | |
| "learning_rate": 1.9479769653775107e-05, | |
| "loss": 0.6785, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.13313161875945537, | |
| "grad_norm": 1.2789502955418373, | |
| "learning_rate": 1.944806046466878e-05, | |
| "loss": 0.6714, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.1361573373676248, | |
| "grad_norm": 1.3741313955797672, | |
| "learning_rate": 1.941544065183021e-05, | |
| "loss": 0.6777, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.13918305597579425, | |
| "grad_norm": 1.3393661605480673, | |
| "learning_rate": 1.9381913359224844e-05, | |
| "loss": 0.6628, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.1422087745839637, | |
| "grad_norm": 1.3662584967324343, | |
| "learning_rate": 1.9347481818282927e-05, | |
| "loss": 0.6574, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.14523449319213314, | |
| "grad_norm": 1.4036879546235965, | |
| "learning_rate": 1.9312149347588035e-05, | |
| "loss": 0.6736, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.14826021180030258, | |
| "grad_norm": 1.41250853096852, | |
| "learning_rate": 1.9275919352557242e-05, | |
| "loss": 0.6888, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.15128593040847202, | |
| "grad_norm": 7.567170155719625, | |
| "learning_rate": 1.9238795325112867e-05, | |
| "loss": 0.6699, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.15431164901664146, | |
| "grad_norm": 1.2602363573311963, | |
| "learning_rate": 1.920078084334595e-05, | |
| "loss": 0.6389, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.1573373676248109, | |
| "grad_norm": 1.2874728755236298, | |
| "learning_rate": 1.916187957117136e-05, | |
| "loss": 0.679, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.16036308623298035, | |
| "grad_norm": 1.2123512496936566, | |
| "learning_rate": 1.9122095257974676e-05, | |
| "loss": 0.6425, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.16338880484114976, | |
| "grad_norm": 1.2312663265173853, | |
| "learning_rate": 1.9081431738250815e-05, | |
| "loss": 0.6443, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.1664145234493192, | |
| "grad_norm": 1.246737094409614, | |
| "learning_rate": 1.9039892931234434e-05, | |
| "loss": 0.664, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.16944024205748864, | |
| "grad_norm": 1.2911686216202691, | |
| "learning_rate": 1.8997482840522218e-05, | |
| "loss": 0.6156, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.17246596066565809, | |
| "grad_norm": 1.245319647881283, | |
| "learning_rate": 1.895420555368697e-05, | |
| "loss": 0.6799, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.17549167927382753, | |
| "grad_norm": 1.345410733599897, | |
| "learning_rate": 1.891006524188368e-05, | |
| "loss": 0.6881, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.17851739788199697, | |
| "grad_norm": 1.2311279741215253, | |
| "learning_rate": 1.8865066159447468e-05, | |
| "loss": 0.6471, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.1815431164901664, | |
| "grad_norm": 1.2741957566673983, | |
| "learning_rate": 1.881921264348355e-05, | |
| "loss": 0.6785, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.18456883509833585, | |
| "grad_norm": 1.2598163044475674, | |
| "learning_rate": 1.8772509113449243e-05, | |
| "loss": 0.6859, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.1875945537065053, | |
| "grad_norm": 1.341909442102569, | |
| "learning_rate": 1.8724960070727974e-05, | |
| "loss": 0.7012, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.19062027231467474, | |
| "grad_norm": 1.2303070109576786, | |
| "learning_rate": 1.8676570098195443e-05, | |
| "loss": 0.6645, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.19364599092284418, | |
| "grad_norm": 1.272712059792455, | |
| "learning_rate": 1.862734385977792e-05, | |
| "loss": 0.6789, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.19667170953101362, | |
| "grad_norm": 1.3565185037992962, | |
| "learning_rate": 1.8577286100002723e-05, | |
| "loss": 0.6745, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.19969742813918306, | |
| "grad_norm": 1.2626838317024291, | |
| "learning_rate": 1.8526401643540924e-05, | |
| "loss": 0.6684, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.2027231467473525, | |
| "grad_norm": 1.237315545474584, | |
| "learning_rate": 1.8474695394742345e-05, | |
| "loss": 0.679, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.20574886535552195, | |
| "grad_norm": 1.1924234259827424, | |
| "learning_rate": 1.8422172337162865e-05, | |
| "loss": 0.6499, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.2087745839636914, | |
| "grad_norm": 1.1901494622357276, | |
| "learning_rate": 1.8368837533084092e-05, | |
| "loss": 0.6541, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.2118003025718608, | |
| "grad_norm": 1.2913463212344467, | |
| "learning_rate": 1.8314696123025456e-05, | |
| "loss": 0.6882, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.21482602118003025, | |
| "grad_norm": 1.2325296438799676, | |
| "learning_rate": 1.825975332524873e-05, | |
| "loss": 0.6975, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.2178517397881997, | |
| "grad_norm": 1.3107985662728308, | |
| "learning_rate": 1.8204014435255136e-05, | |
| "loss": 0.7037, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.22087745839636913, | |
| "grad_norm": 1.2813654912731927, | |
| "learning_rate": 1.8147484825274895e-05, | |
| "loss": 0.6445, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.22390317700453857, | |
| "grad_norm": 1.339160808607977, | |
| "learning_rate": 1.8090169943749477e-05, | |
| "loss": 0.6871, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.22692889561270801, | |
| "grad_norm": 1.1902505573236317, | |
| "learning_rate": 1.803207531480645e-05, | |
| "loss": 0.6459, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.22995461422087746, | |
| "grad_norm": 1.3379756086738064, | |
| "learning_rate": 1.797320653772707e-05, | |
| "loss": 0.6349, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.2329803328290469, | |
| "grad_norm": 1.184265960372134, | |
| "learning_rate": 1.7913569286406606e-05, | |
| "loss": 0.6163, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.23600605143721634, | |
| "grad_norm": 1.2152412782879942, | |
| "learning_rate": 1.785316930880745e-05, | |
| "loss": 0.6571, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.23903177004538578, | |
| "grad_norm": 1.25082583010917, | |
| "learning_rate": 1.779201242640517e-05, | |
| "loss": 0.6668, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.24205748865355523, | |
| "grad_norm": 1.1813278651969805, | |
| "learning_rate": 1.773010453362737e-05, | |
| "loss": 0.6445, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.24508320726172467, | |
| "grad_norm": 1.2629648431875875, | |
| "learning_rate": 1.7667451597285617e-05, | |
| "loss": 0.6669, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.2481089258698941, | |
| "grad_norm": 1.2511431799532198, | |
| "learning_rate": 1.7604059656000313e-05, | |
| "loss": 0.6517, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.25113464447806355, | |
| "grad_norm": 1.3278472678063655, | |
| "learning_rate": 1.7539934819618696e-05, | |
| "loss": 0.7062, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.25113464447806355, | |
| "eval_loss": 0.6539373397827148, | |
| "eval_runtime": 517.9727, | |
| "eval_samples_per_second": 8.166, | |
| "eval_steps_per_second": 0.257, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.25416036308623297, | |
| "grad_norm": 1.195691216149565, | |
| "learning_rate": 1.747508326862597e-05, | |
| "loss": 0.6549, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.25718608169440244, | |
| "grad_norm": 1.3106381000281655, | |
| "learning_rate": 1.7409511253549592e-05, | |
| "loss": 0.6566, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.26021180030257185, | |
| "grad_norm": 1.4465663259887058, | |
| "learning_rate": 1.7343225094356857e-05, | |
| "loss": 0.6312, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.2632375189107413, | |
| "grad_norm": 1.261517340493337, | |
| "learning_rate": 1.727623117984575e-05, | |
| "loss": 0.6614, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.26626323751891073, | |
| "grad_norm": 1.3621146600403318, | |
| "learning_rate": 1.720853596702919e-05, | |
| "loss": 0.6505, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.2692889561270802, | |
| "grad_norm": 1.2845918854537075, | |
| "learning_rate": 1.7140145980512684e-05, | |
| "loss": 0.6496, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.2723146747352496, | |
| "grad_norm": 1.2903185873747185, | |
| "learning_rate": 1.7071067811865477e-05, | |
| "loss": 0.6549, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2753403933434191, | |
| "grad_norm": 1.3203037534971733, | |
| "learning_rate": 1.7001308118985237e-05, | |
| "loss": 0.6795, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.2783661119515885, | |
| "grad_norm": 1.271934504746521, | |
| "learning_rate": 1.6930873625456362e-05, | |
| "loss": 0.6679, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.2813918305597579, | |
| "grad_norm": 1.19935850222408, | |
| "learning_rate": 1.685977111990193e-05, | |
| "loss": 0.6562, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.2844175491679274, | |
| "grad_norm": 1.1463148559918699, | |
| "learning_rate": 1.678800745532942e-05, | |
| "loss": 0.6306, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.2874432677760968, | |
| "grad_norm": 1.2342597403887718, | |
| "learning_rate": 1.6715589548470187e-05, | |
| "loss": 0.6441, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.29046898638426627, | |
| "grad_norm": 1.192973031987555, | |
| "learning_rate": 1.664252437911282e-05, | |
| "loss": 0.6434, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.2934947049924357, | |
| "grad_norm": 1.2392218132779826, | |
| "learning_rate": 1.6568818989430416e-05, | |
| "loss": 0.6709, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.29652042360060515, | |
| "grad_norm": 1.1940577422773688, | |
| "learning_rate": 1.6494480483301836e-05, | |
| "loss": 0.6472, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.29954614220877457, | |
| "grad_norm": 1.2035237616611898, | |
| "learning_rate": 1.641951602562703e-05, | |
| "loss": 0.638, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.30257186081694404, | |
| "grad_norm": 1.1957243470765997, | |
| "learning_rate": 1.6343932841636455e-05, | |
| "loss": 0.6522, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.30559757942511345, | |
| "grad_norm": 1.182041345122576, | |
| "learning_rate": 1.6267738216194698e-05, | |
| "loss": 0.6372, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.3086232980332829, | |
| "grad_norm": 1.2275892324526625, | |
| "learning_rate": 1.6190939493098344e-05, | |
| "loss": 0.6399, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.31164901664145234, | |
| "grad_norm": 1.2474415165677772, | |
| "learning_rate": 1.6113544074368166e-05, | |
| "loss": 0.6535, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.3146747352496218, | |
| "grad_norm": 1.2175516565864553, | |
| "learning_rate": 1.6035559419535714e-05, | |
| "loss": 0.6489, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.3177004538577912, | |
| "grad_norm": 1.2628229502278452, | |
| "learning_rate": 1.5956993044924334e-05, | |
| "loss": 0.6705, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.3207261724659607, | |
| "grad_norm": 1.1348825334058605, | |
| "learning_rate": 1.5877852522924733e-05, | |
| "loss": 0.6083, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.3237518910741301, | |
| "grad_norm": 1.1495314181861485, | |
| "learning_rate": 1.579814548126514e-05, | |
| "loss": 0.6239, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.3267776096822995, | |
| "grad_norm": 1.1964854618221463, | |
| "learning_rate": 1.5717879602276123e-05, | |
| "loss": 0.6363, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.329803328290469, | |
| "grad_norm": 1.1792532141608787, | |
| "learning_rate": 1.5637062622150168e-05, | |
| "loss": 0.6625, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.3328290468986384, | |
| "grad_norm": 1.178622078286246, | |
| "learning_rate": 1.5555702330196024e-05, | |
| "loss": 0.6384, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3358547655068079, | |
| "grad_norm": 1.3750465045364273, | |
| "learning_rate": 1.547380656808797e-05, | |
| "loss": 0.6415, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.3388804841149773, | |
| "grad_norm": 1.1893414197245047, | |
| "learning_rate": 1.5391383229110005e-05, | |
| "loss": 0.6385, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.34190620272314676, | |
| "grad_norm": 1.2575199704949114, | |
| "learning_rate": 1.5308440257395095e-05, | |
| "loss": 0.6213, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.34493192133131617, | |
| "grad_norm": 1.2684705470221596, | |
| "learning_rate": 1.5224985647159489e-05, | |
| "loss": 0.6417, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.34795763993948564, | |
| "grad_norm": 1.2779635417166455, | |
| "learning_rate": 1.5141027441932217e-05, | |
| "loss": 0.6629, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.35098335854765506, | |
| "grad_norm": 1.2106527731125027, | |
| "learning_rate": 1.5056573733779848e-05, | |
| "loss": 0.6145, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.3540090771558245, | |
| "grad_norm": 1.3096453071684298, | |
| "learning_rate": 1.4971632662526545e-05, | |
| "loss": 0.6438, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.35703479576399394, | |
| "grad_norm": 1.2680171691145932, | |
| "learning_rate": 1.4886212414969551e-05, | |
| "loss": 0.6577, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.3600605143721634, | |
| "grad_norm": 1.2422923332660138, | |
| "learning_rate": 1.4800321224090114e-05, | |
| "loss": 0.6454, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.3630862329803328, | |
| "grad_norm": 1.1699490116012057, | |
| "learning_rate": 1.4713967368259981e-05, | |
| "loss": 0.6187, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3661119515885023, | |
| "grad_norm": 1.2632268107037634, | |
| "learning_rate": 1.4627159170443504e-05, | |
| "loss": 0.6518, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.3691376701966717, | |
| "grad_norm": 1.2445701989291957, | |
| "learning_rate": 1.4539904997395468e-05, | |
| "loss": 0.6353, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.3721633888048411, | |
| "grad_norm": 1.1721137453844679, | |
| "learning_rate": 1.4452213258854684e-05, | |
| "loss": 0.6481, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.3751891074130106, | |
| "grad_norm": 1.1710154709386082, | |
| "learning_rate": 1.436409240673342e-05, | |
| "loss": 0.6226, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.37821482602118, | |
| "grad_norm": 1.230481237987904, | |
| "learning_rate": 1.4275550934302822e-05, | |
| "loss": 0.6301, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.3812405446293495, | |
| "grad_norm": 1.1045551599001833, | |
| "learning_rate": 1.4186597375374283e-05, | |
| "loss": 0.6082, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.3842662632375189, | |
| "grad_norm": 1.270890856731403, | |
| "learning_rate": 1.4097240303476955e-05, | |
| "loss": 0.6239, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.38729198184568836, | |
| "grad_norm": 1.209681465291139, | |
| "learning_rate": 1.4007488331031409e-05, | |
| "loss": 0.6215, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.3903177004538578, | |
| "grad_norm": 1.2010343473701133, | |
| "learning_rate": 1.391735010851956e-05, | |
| "loss": 0.6631, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.39334341906202724, | |
| "grad_norm": 1.2020888955592754, | |
| "learning_rate": 1.3826834323650899e-05, | |
| "loss": 0.6392, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.39636913767019666, | |
| "grad_norm": 1.1487663280027558, | |
| "learning_rate": 1.3735949700525164e-05, | |
| "loss": 0.6231, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.39939485627836613, | |
| "grad_norm": 1.2084902235391288, | |
| "learning_rate": 1.3644704998791501e-05, | |
| "loss": 0.6701, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.40242057488653554, | |
| "grad_norm": 1.1195154082552468, | |
| "learning_rate": 1.3553109012804162e-05, | |
| "loss": 0.6468, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.405446293494705, | |
| "grad_norm": 1.1738326477965182, | |
| "learning_rate": 1.346117057077493e-05, | |
| "loss": 0.642, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.4084720121028744, | |
| "grad_norm": 1.2105895076313393, | |
| "learning_rate": 1.3368898533922202e-05, | |
| "loss": 0.6439, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.4114977307110439, | |
| "grad_norm": 1.5606864803103329, | |
| "learning_rate": 1.3276301795616937e-05, | |
| "loss": 0.6451, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.4145234493192133, | |
| "grad_norm": 1.1442624872092446, | |
| "learning_rate": 1.3183389280525497e-05, | |
| "loss": 0.6247, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.4175491679273828, | |
| "grad_norm": 1.1536291106956869, | |
| "learning_rate": 1.3090169943749475e-05, | |
| "loss": 0.6316, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.4205748865355522, | |
| "grad_norm": 1.1831101262833283, | |
| "learning_rate": 1.2996652769962567e-05, | |
| "loss": 0.6335, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.4236006051437216, | |
| "grad_norm": 1.1912437237028088, | |
| "learning_rate": 1.2902846772544625e-05, | |
| "loss": 0.6399, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.4266263237518911, | |
| "grad_norm": 1.1251643146637977, | |
| "learning_rate": 1.2808760992712923e-05, | |
| "loss": 0.6306, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.4296520423600605, | |
| "grad_norm": 1.2001309518394503, | |
| "learning_rate": 1.2714404498650743e-05, | |
| "loss": 0.65, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.43267776096822996, | |
| "grad_norm": 1.12760607919054, | |
| "learning_rate": 1.2619786384633374e-05, | |
| "loss": 0.5942, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.4357034795763994, | |
| "grad_norm": 1.1893189807548705, | |
| "learning_rate": 1.252491577015158e-05, | |
| "loss": 0.6379, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.43872919818456885, | |
| "grad_norm": 1.1816565122038842, | |
| "learning_rate": 1.242980179903264e-05, | |
| "loss": 0.6499, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.44175491679273826, | |
| "grad_norm": 1.1272239154012946, | |
| "learning_rate": 1.2334453638559057e-05, | |
| "loss": 0.6109, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.44478063540090773, | |
| "grad_norm": 1.2377211531913517, | |
| "learning_rate": 1.2238880478584987e-05, | |
| "loss": 0.6335, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.44780635400907715, | |
| "grad_norm": 1.18055747817248, | |
| "learning_rate": 1.2143091530650508e-05, | |
| "loss": 0.6351, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.4508320726172466, | |
| "grad_norm": 1.147951061684639, | |
| "learning_rate": 1.2047096027093798e-05, | |
| "loss": 0.6307, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.45385779122541603, | |
| "grad_norm": 1.1253862800883228, | |
| "learning_rate": 1.1950903220161286e-05, | |
| "loss": 0.6615, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.4568835098335855, | |
| "grad_norm": 1.1922102982062615, | |
| "learning_rate": 1.185452238111591e-05, | |
| "loss": 0.6186, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.4599092284417549, | |
| "grad_norm": 1.1413878459701803, | |
| "learning_rate": 1.1757962799343548e-05, | |
| "loss": 0.6174, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.4629349470499244, | |
| "grad_norm": 1.1274239305205713, | |
| "learning_rate": 1.1661233781457655e-05, | |
| "loss": 0.6039, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.4659606656580938, | |
| "grad_norm": 1.1577293786563616, | |
| "learning_rate": 1.156434465040231e-05, | |
| "loss": 0.6175, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.4689863842662632, | |
| "grad_norm": 1.1193128885203618, | |
| "learning_rate": 1.1467304744553618e-05, | |
| "loss": 0.6217, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.4720121028744327, | |
| "grad_norm": 1.173935492953021, | |
| "learning_rate": 1.1370123416819683e-05, | |
| "loss": 0.6166, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.4750378214826021, | |
| "grad_norm": 1.1018685228300982, | |
| "learning_rate": 1.1272810033739134e-05, | |
| "loss": 0.6275, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.47806354009077157, | |
| "grad_norm": 1.1143389993315551, | |
| "learning_rate": 1.1175373974578378e-05, | |
| "loss": 0.6147, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.481089258698941, | |
| "grad_norm": 1.1183090058031513, | |
| "learning_rate": 1.1077824630427593e-05, | |
| "loss": 0.6023, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.48411497730711045, | |
| "grad_norm": 1.0760482878636355, | |
| "learning_rate": 1.098017140329561e-05, | |
| "loss": 0.6332, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.48714069591527986, | |
| "grad_norm": 1.1457835297378434, | |
| "learning_rate": 1.0882423705203698e-05, | |
| "loss": 0.6296, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.49016641452344933, | |
| "grad_norm": 1.0719482322284357, | |
| "learning_rate": 1.0784590957278452e-05, | |
| "loss": 0.5907, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.49319213313161875, | |
| "grad_norm": 1.1559971453594726, | |
| "learning_rate": 1.0686682588843737e-05, | |
| "loss": 0.6145, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.4962178517397882, | |
| "grad_norm": 1.1128953951058218, | |
| "learning_rate": 1.058870803651189e-05, | |
| "loss": 0.6278, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.49924357034795763, | |
| "grad_norm": 1.1950923510690505, | |
| "learning_rate": 1.0490676743274181e-05, | |
| "loss": 0.6445, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5022692889561271, | |
| "grad_norm": 1.2068875099806577, | |
| "learning_rate": 1.0392598157590687e-05, | |
| "loss": 0.5958, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.5022692889561271, | |
| "eval_loss": 0.6210768818855286, | |
| "eval_runtime": 509.1551, | |
| "eval_samples_per_second": 8.308, | |
| "eval_steps_per_second": 0.261, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.5052950075642966, | |
| "grad_norm": 1.1188327526562485, | |
| "learning_rate": 1.0294481732479635e-05, | |
| "loss": 0.6284, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.5083207261724659, | |
| "grad_norm": 1.0755209583049905, | |
| "learning_rate": 1.0196336924606282e-05, | |
| "loss": 0.6237, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.5113464447806354, | |
| "grad_norm": 1.1965795137847735, | |
| "learning_rate": 1.0098173193371498e-05, | |
| "loss": 0.5947, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.5143721633888049, | |
| "grad_norm": 1.1602878236650798, | |
| "learning_rate": 1e-05, | |
| "loss": 0.622, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5173978819969742, | |
| "grad_norm": 1.1465484727449606, | |
| "learning_rate": 9.901826806628505e-06, | |
| "loss": 0.6392, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.5204236006051437, | |
| "grad_norm": 1.165809981442701, | |
| "learning_rate": 9.80366307539372e-06, | |
| "loss": 0.6346, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.5234493192133132, | |
| "grad_norm": 1.1913219568057807, | |
| "learning_rate": 9.705518267520369e-06, | |
| "loss": 0.6328, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.5264750378214826, | |
| "grad_norm": 1.0613532585644816, | |
| "learning_rate": 9.607401842409318e-06, | |
| "loss": 0.5804, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.529500756429652, | |
| "grad_norm": 1.1552650155426534, | |
| "learning_rate": 9.50932325672582e-06, | |
| "loss": 0.626, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.5325264750378215, | |
| "grad_norm": 1.1350285592003395, | |
| "learning_rate": 9.41129196348811e-06, | |
| "loss": 0.6014, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.5355521936459909, | |
| "grad_norm": 1.216950617218585, | |
| "learning_rate": 9.313317411156265e-06, | |
| "loss": 0.6213, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.5385779122541604, | |
| "grad_norm": 1.118609826885826, | |
| "learning_rate": 9.215409042721553e-06, | |
| "loss": 0.6057, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.5416036308623298, | |
| "grad_norm": 1.188155151455158, | |
| "learning_rate": 9.117576294796307e-06, | |
| "loss": 0.6145, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.5446293494704992, | |
| "grad_norm": 1.104116332061839, | |
| "learning_rate": 9.019828596704394e-06, | |
| "loss": 0.5826, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.5476550680786687, | |
| "grad_norm": 1.2042406267126835, | |
| "learning_rate": 8.922175369572407e-06, | |
| "loss": 0.6413, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.5506807866868382, | |
| "grad_norm": 1.0973076609897112, | |
| "learning_rate": 8.824626025421625e-06, | |
| "loss": 0.6166, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.5537065052950075, | |
| "grad_norm": 1.3471325546622461, | |
| "learning_rate": 8.72718996626087e-06, | |
| "loss": 0.6289, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.556732223903177, | |
| "grad_norm": 1.1120256137494327, | |
| "learning_rate": 8.629876583180322e-06, | |
| "loss": 0.6231, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.5597579425113465, | |
| "grad_norm": 1.061138780482171, | |
| "learning_rate": 8.532695255446384e-06, | |
| "loss": 0.5869, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.5627836611195158, | |
| "grad_norm": 1.1109792746650928, | |
| "learning_rate": 8.43565534959769e-06, | |
| "loss": 0.5999, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.5658093797276853, | |
| "grad_norm": 1.1359210835859026, | |
| "learning_rate": 8.338766218542348e-06, | |
| "loss": 0.5973, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.5688350983358548, | |
| "grad_norm": 1.070519194256954, | |
| "learning_rate": 8.242037200656455e-06, | |
| "loss": 0.5914, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.5718608169440242, | |
| "grad_norm": 1.1376721044325269, | |
| "learning_rate": 8.145477618884092e-06, | |
| "loss": 0.5974, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.5748865355521936, | |
| "grad_norm": 1.063249642986741, | |
| "learning_rate": 8.04909677983872e-06, | |
| "loss": 0.6151, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.5779122541603631, | |
| "grad_norm": 1.1163908474355613, | |
| "learning_rate": 7.952903972906205e-06, | |
| "loss": 0.6271, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.5809379727685325, | |
| "grad_norm": 1.10402989463971, | |
| "learning_rate": 7.856908469349495e-06, | |
| "loss": 0.6032, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.583963691376702, | |
| "grad_norm": 1.1268103929858282, | |
| "learning_rate": 7.761119521415017e-06, | |
| "loss": 0.6361, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.5869894099848714, | |
| "grad_norm": 1.117624515150089, | |
| "learning_rate": 7.66554636144095e-06, | |
| "loss": 0.6243, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.5900151285930408, | |
| "grad_norm": 1.2109925184652597, | |
| "learning_rate": 7.570198200967363e-06, | |
| "loss": 0.6261, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.5930408472012103, | |
| "grad_norm": 1.160096122552475, | |
| "learning_rate": 7.4750842298484205e-06, | |
| "loss": 0.6119, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.5960665658093798, | |
| "grad_norm": 1.078517815120857, | |
| "learning_rate": 7.380213615366627e-06, | |
| "loss": 0.6218, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.5990922844175491, | |
| "grad_norm": 1.1304014873685944, | |
| "learning_rate": 7.285595501349259e-06, | |
| "loss": 0.6393, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.6021180030257186, | |
| "grad_norm": 1.152813536667268, | |
| "learning_rate": 7.191239007287082e-06, | |
| "loss": 0.6355, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.6051437216338881, | |
| "grad_norm": 1.2404709127337052, | |
| "learning_rate": 7.097153227455379e-06, | |
| "loss": 0.6077, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6081694402420574, | |
| "grad_norm": 1.1109088360168653, | |
| "learning_rate": 7.003347230037434e-06, | |
| "loss": 0.5981, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.6111951588502269, | |
| "grad_norm": 1.1091223872438236, | |
| "learning_rate": 6.909830056250527e-06, | |
| "loss": 0.6105, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.6142208774583964, | |
| "grad_norm": 1.0646921278347719, | |
| "learning_rate": 6.816610719474503e-06, | |
| "loss": 0.5825, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.6172465960665658, | |
| "grad_norm": 1.0573083724617258, | |
| "learning_rate": 6.723698204383067e-06, | |
| "loss": 0.6025, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.6202723146747352, | |
| "grad_norm": 1.058391745894815, | |
| "learning_rate": 6.631101466077801e-06, | |
| "loss": 0.5882, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.6232980332829047, | |
| "grad_norm": 1.0860673187757808, | |
| "learning_rate": 6.538829429225068e-06, | |
| "loss": 0.5855, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.6263237518910741, | |
| "grad_norm": 1.1028815842721016, | |
| "learning_rate": 6.446890987195842e-06, | |
| "loss": 0.6009, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.6293494704992436, | |
| "grad_norm": 1.190831334333711, | |
| "learning_rate": 6.355295001208504e-06, | |
| "loss": 0.6073, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.632375189107413, | |
| "grad_norm": 1.1040886621274888, | |
| "learning_rate": 6.2640502994748375e-06, | |
| "loss": 0.5754, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.6354009077155824, | |
| "grad_norm": 1.1484682306046927, | |
| "learning_rate": 6.173165676349103e-06, | |
| "loss": 0.6251, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.6384266263237519, | |
| "grad_norm": 1.1093112202274276, | |
| "learning_rate": 6.082649891480441e-06, | |
| "loss": 0.6131, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.6414523449319214, | |
| "grad_norm": 1.2377771411781646, | |
| "learning_rate": 5.9925116689685925e-06, | |
| "loss": 0.6218, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.6444780635400907, | |
| "grad_norm": 1.0644605371844993, | |
| "learning_rate": 5.902759696523046e-06, | |
| "loss": 0.5958, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.6475037821482602, | |
| "grad_norm": 1.1423519631175298, | |
| "learning_rate": 5.813402624625722e-06, | |
| "loss": 0.5949, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.6505295007564297, | |
| "grad_norm": 1.1906467124974514, | |
| "learning_rate": 5.724449065697182e-06, | |
| "loss": 0.631, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.653555219364599, | |
| "grad_norm": 1.1651203942266402, | |
| "learning_rate": 5.635907593266578e-06, | |
| "loss": 0.643, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.6565809379727685, | |
| "grad_norm": 1.0906584200307796, | |
| "learning_rate": 5.54778674114532e-06, | |
| "loss": 0.598, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.659606656580938, | |
| "grad_norm": 1.1679679858048087, | |
| "learning_rate": 5.460095002604533e-06, | |
| "loss": 0.6017, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.6626323751891074, | |
| "grad_norm": 1.0817989627430684, | |
| "learning_rate": 5.3728408295565e-06, | |
| "loss": 0.5846, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.6656580937972768, | |
| "grad_norm": 1.0652519984067952, | |
| "learning_rate": 5.286032631740023e-06, | |
| "loss": 0.5988, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.6686838124054463, | |
| "grad_norm": 1.1308685991861618, | |
| "learning_rate": 5.199678775909889e-06, | |
| "loss": 0.6254, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.6717095310136157, | |
| "grad_norm": 1.1014848980473453, | |
| "learning_rate": 5.1137875850304545e-06, | |
| "loss": 0.6062, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.6747352496217852, | |
| "grad_norm": 1.1040686588501816, | |
| "learning_rate": 5.0283673374734546e-06, | |
| "loss": 0.6, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.6777609682299546, | |
| "grad_norm": 1.1100475808626904, | |
| "learning_rate": 4.943426266220156e-06, | |
| "loss": 0.5965, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.680786686838124, | |
| "grad_norm": 1.135886262904592, | |
| "learning_rate": 4.858972558067784e-06, | |
| "loss": 0.6162, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.6838124054462935, | |
| "grad_norm": 1.156656991004886, | |
| "learning_rate": 4.775014352840512e-06, | |
| "loss": 0.6092, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.686838124054463, | |
| "grad_norm": 1.1362751840271528, | |
| "learning_rate": 4.691559742604906e-06, | |
| "loss": 0.6199, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.6898638426626323, | |
| "grad_norm": 1.3413766650130845, | |
| "learning_rate": 4.608616770889998e-06, | |
| "loss": 0.5994, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.6928895612708018, | |
| "grad_norm": 1.1054887530044246, | |
| "learning_rate": 4.526193431912038e-06, | |
| "loss": 0.601, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.6959152798789713, | |
| "grad_norm": 1.1434835536105747, | |
| "learning_rate": 4.444297669803981e-06, | |
| "loss": 0.6147, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.6989409984871406, | |
| "grad_norm": 1.0939569921226582, | |
| "learning_rate": 4.362937377849832e-06, | |
| "loss": 0.5832, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.7019667170953101, | |
| "grad_norm": 1.082711369071378, | |
| "learning_rate": 4.282120397723879e-06, | |
| "loss": 0.6098, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.7049924357034796, | |
| "grad_norm": 1.1879264179303997, | |
| "learning_rate": 4.2018545187348645e-06, | |
| "loss": 0.6324, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.708018154311649, | |
| "grad_norm": 1.148322035402214, | |
| "learning_rate": 4.12214747707527e-06, | |
| "loss": 0.6231, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.7110438729198184, | |
| "grad_norm": 1.114138492320442, | |
| "learning_rate": 4.043006955075667e-06, | |
| "loss": 0.5985, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.7140695915279879, | |
| "grad_norm": 1.149912019430139, | |
| "learning_rate": 3.964440580464286e-06, | |
| "loss": 0.5873, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.7170953101361573, | |
| "grad_norm": 1.0773097040934607, | |
| "learning_rate": 3.8864559256318375e-06, | |
| "loss": 0.6061, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.7201210287443268, | |
| "grad_norm": 1.1287631308059118, | |
| "learning_rate": 3.8090605069016596e-06, | |
| "loss": 0.6185, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.7231467473524962, | |
| "grad_norm": 1.0887097046608571, | |
| "learning_rate": 3.7322617838053066e-06, | |
| "loss": 0.5916, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.7261724659606656, | |
| "grad_norm": 1.1507226185265305, | |
| "learning_rate": 3.6560671583635467e-06, | |
| "loss": 0.6013, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.7291981845688351, | |
| "grad_norm": 1.1132742488152783, | |
| "learning_rate": 3.58048397437297e-06, | |
| "loss": 0.594, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.7322239031770046, | |
| "grad_norm": 1.14757173958559, | |
| "learning_rate": 3.505519516698165e-06, | |
| "loss": 0.6131, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.735249621785174, | |
| "grad_norm": 1.076757142610778, | |
| "learning_rate": 3.4311810105695875e-06, | |
| "loss": 0.5957, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.7382753403933434, | |
| "grad_norm": 1.0833721025321175, | |
| "learning_rate": 3.3574756208871862e-06, | |
| "loss": 0.5957, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.7413010590015129, | |
| "grad_norm": 1.1782563183368346, | |
| "learning_rate": 3.284410451529816e-06, | |
| "loss": 0.6264, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.7443267776096822, | |
| "grad_norm": 1.1962519652472652, | |
| "learning_rate": 3.2119925446705824e-06, | |
| "loss": 0.5734, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.7473524962178517, | |
| "grad_norm": 1.1419543389549884, | |
| "learning_rate": 3.140228880098074e-06, | |
| "loss": 0.6049, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.7503782148260212, | |
| "grad_norm": 1.112382576402018, | |
| "learning_rate": 3.069126374543643e-06, | |
| "loss": 0.5939, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.7534039334341907, | |
| "grad_norm": 1.1138275367863544, | |
| "learning_rate": 2.998691881014765e-06, | |
| "loss": 0.6102, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.7534039334341907, | |
| "eval_loss": 0.6005925536155701, | |
| "eval_runtime": 508.9696, | |
| "eval_samples_per_second": 8.311, | |
| "eval_steps_per_second": 0.261, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.75642965204236, | |
| "grad_norm": 1.112838008847437, | |
| "learning_rate": 2.9289321881345257e-06, | |
| "loss": 0.6292, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.7594553706505295, | |
| "grad_norm": 1.1565006243041378, | |
| "learning_rate": 2.859854019487318e-06, | |
| "loss": 0.6142, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.762481089258699, | |
| "grad_norm": 1.2801042349950706, | |
| "learning_rate": 2.791464032970812e-06, | |
| "loss": 0.6238, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.7655068078668684, | |
| "grad_norm": 1.0489320492279628, | |
| "learning_rate": 2.723768820154251e-06, | |
| "loss": 0.5646, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.7685325264750378, | |
| "grad_norm": 1.084220370401727, | |
| "learning_rate": 2.656774905643147e-06, | |
| "loss": 0.5772, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.7715582450832073, | |
| "grad_norm": 1.0781673091449353, | |
| "learning_rate": 2.5904887464504115e-06, | |
| "loss": 0.6088, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.7745839636913767, | |
| "grad_norm": 1.1401732584364122, | |
| "learning_rate": 2.5249167313740307e-06, | |
| "loss": 0.5889, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.7776096822995462, | |
| "grad_norm": 1.1065290368457845, | |
| "learning_rate": 2.4600651803813057e-06, | |
| "loss": 0.5956, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.7806354009077155, | |
| "grad_norm": 1.1083030548851147, | |
| "learning_rate": 2.395940343999691e-06, | |
| "loss": 0.5945, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.783661119515885, | |
| "grad_norm": 1.0465024217478145, | |
| "learning_rate": 2.332548402714385e-06, | |
| "loss": 0.5683, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.7866868381240545, | |
| "grad_norm": 1.0447251914422466, | |
| "learning_rate": 2.26989546637263e-06, | |
| "loss": 0.5594, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.789712556732224, | |
| "grad_norm": 1.0714188640502096, | |
| "learning_rate": 2.207987573594833e-06, | |
| "loss": 0.6147, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.7927382753403933, | |
| "grad_norm": 1.0819256620522593, | |
| "learning_rate": 2.146830691192553e-06, | |
| "loss": 0.608, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.7957639939485628, | |
| "grad_norm": 1.1103552345943686, | |
| "learning_rate": 2.086430713593397e-06, | |
| "loss": 0.5998, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.7987897125567323, | |
| "grad_norm": 1.1115854411103623, | |
| "learning_rate": 2.02679346227293e-06, | |
| "loss": 0.6249, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.8018154311649016, | |
| "grad_norm": 1.0997785192243315, | |
| "learning_rate": 1.967924685193552e-06, | |
| "loss": 0.6121, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.8048411497730711, | |
| "grad_norm": 1.1115287397946079, | |
| "learning_rate": 1.9098300562505266e-06, | |
| "loss": 0.5951, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.8078668683812406, | |
| "grad_norm": 1.094826560847298, | |
| "learning_rate": 1.8525151747251058e-06, | |
| "loss": 0.5913, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.81089258698941, | |
| "grad_norm": 1.0830430435230047, | |
| "learning_rate": 1.7959855647448642e-06, | |
| "loss": 0.5881, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.8139183055975794, | |
| "grad_norm": 1.1125565398099833, | |
| "learning_rate": 1.7402466747512704e-06, | |
| "loss": 0.5858, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.8169440242057489, | |
| "grad_norm": 1.0872500758994135, | |
| "learning_rate": 1.6853038769745466e-06, | |
| "loss": 0.6078, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.8199697428139183, | |
| "grad_norm": 1.0357455255529537, | |
| "learning_rate": 1.6311624669159064e-06, | |
| "loss": 0.6008, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.8229954614220878, | |
| "grad_norm": 1.0898946760954742, | |
| "learning_rate": 1.577827662837136e-06, | |
| "loss": 0.5906, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.8260211800302572, | |
| "grad_norm": 1.0858383840210986, | |
| "learning_rate": 1.5253046052576559e-06, | |
| "loss": 0.5937, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.8290468986384266, | |
| "grad_norm": 1.1250562661744625, | |
| "learning_rate": 1.4735983564590784e-06, | |
| "loss": 0.6102, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.8320726172465961, | |
| "grad_norm": 1.115804430633625, | |
| "learning_rate": 1.4227138999972801e-06, | |
| "loss": 0.6204, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.8350983358547656, | |
| "grad_norm": 1.0987165616536443, | |
| "learning_rate": 1.3726561402220818e-06, | |
| "loss": 0.5629, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.8381240544629349, | |
| "grad_norm": 1.0870859009064557, | |
| "learning_rate": 1.3234299018045615e-06, | |
| "loss": 0.5875, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.8411497730711044, | |
| "grad_norm": 1.1503941572696326, | |
| "learning_rate": 1.2750399292720284e-06, | |
| "loss": 0.5635, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.8441754916792739, | |
| "grad_norm": 1.0860425500738413, | |
| "learning_rate": 1.2274908865507595e-06, | |
| "loss": 0.5707, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.8472012102874432, | |
| "grad_norm": 1.0304768605145838, | |
| "learning_rate": 1.1807873565164507e-06, | |
| "loss": 0.545, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.8502269288956127, | |
| "grad_norm": 1.131625318456932, | |
| "learning_rate": 1.1349338405525368e-06, | |
| "loss": 0.6175, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.8532526475037822, | |
| "grad_norm": 1.1418856722653956, | |
| "learning_rate": 1.0899347581163222e-06, | |
| "loss": 0.6141, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.8562783661119516, | |
| "grad_norm": 1.0977924723698635, | |
| "learning_rate": 1.045794446313031e-06, | |
| "loss": 0.6036, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.859304084720121, | |
| "grad_norm": 1.062874492585984, | |
| "learning_rate": 1.0025171594777872e-06, | |
| "loss": 0.5852, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.8623298033282905, | |
| "grad_norm": 1.1393199500566773, | |
| "learning_rate": 9.601070687655667e-07, | |
| "loss": 0.6129, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.8653555219364599, | |
| "grad_norm": 1.0841346943250296, | |
| "learning_rate": 9.185682617491865e-07, | |
| "loss": 0.6005, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.8683812405446294, | |
| "grad_norm": 1.0768617351076946, | |
| "learning_rate": 8.779047420253239e-07, | |
| "loss": 0.5819, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.8714069591527988, | |
| "grad_norm": 1.0499503423511858, | |
| "learning_rate": 8.381204288286415e-07, | |
| "loss": 0.5776, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.8744326777609682, | |
| "grad_norm": 1.0532903742876403, | |
| "learning_rate": 7.992191566540519e-07, | |
| "loss": 0.5681, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.8774583963691377, | |
| "grad_norm": 1.050926284473097, | |
| "learning_rate": 7.612046748871327e-07, | |
| "loss": 0.5747, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.8804841149773072, | |
| "grad_norm": 1.1049748051812167, | |
| "learning_rate": 7.240806474427598e-07, | |
| "loss": 0.613, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.8835098335854765, | |
| "grad_norm": 1.0916132558143439, | |
| "learning_rate": 6.878506524119644e-07, | |
| "loss": 0.6287, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.886535552193646, | |
| "grad_norm": 1.0799738463224524, | |
| "learning_rate": 6.525181817170756e-07, | |
| "loss": 0.577, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.8895612708018155, | |
| "grad_norm": 1.0833236918014248, | |
| "learning_rate": 6.180866407751595e-07, | |
| "loss": 0.595, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.8925869894099848, | |
| "grad_norm": 1.150147934485175, | |
| "learning_rate": 5.845593481697931e-07, | |
| "loss": 0.61, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.8956127080181543, | |
| "grad_norm": 1.22574194046837, | |
| "learning_rate": 5.519395353312195e-07, | |
| "loss": 0.5907, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.8986384266263238, | |
| "grad_norm": 1.0755706729431784, | |
| "learning_rate": 5.20230346224897e-07, | |
| "loss": 0.5567, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.9016641452344932, | |
| "grad_norm": 1.0522750804167624, | |
| "learning_rate": 4.894348370484648e-07, | |
| "loss": 0.5703, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.9046898638426626, | |
| "grad_norm": 1.0735148344794117, | |
| "learning_rate": 4.5955597593719593e-07, | |
| "loss": 0.6051, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.9077155824508321, | |
| "grad_norm": 1.0466743069418312, | |
| "learning_rate": 4.305966426779118e-07, | |
| "loss": 0.5811, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.9107413010590015, | |
| "grad_norm": 1.102638050289901, | |
| "learning_rate": 4.025596284314259e-07, | |
| "loss": 0.6205, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.913767019667171, | |
| "grad_norm": 1.0622405448996148, | |
| "learning_rate": 3.7544763546352834e-07, | |
| "loss": 0.5963, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.9167927382753404, | |
| "grad_norm": 1.0808909726872267, | |
| "learning_rate": 3.492632768845261e-07, | |
| "loss": 0.589, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.9198184568835098, | |
| "grad_norm": 1.0738124055012128, | |
| "learning_rate": 3.2400907639740243e-07, | |
| "loss": 0.5979, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.9228441754916793, | |
| "grad_norm": 1.1384061540707928, | |
| "learning_rate": 2.996874680545603e-07, | |
| "loss": 0.6013, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.9258698940998488, | |
| "grad_norm": 1.0886226051949452, | |
| "learning_rate": 2.7630079602323447e-07, | |
| "loss": 0.5832, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.9288956127080181, | |
| "grad_norm": 1.1188724454503705, | |
| "learning_rate": 2.5385131435955e-07, | |
| "loss": 0.5994, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.9319213313161876, | |
| "grad_norm": 1.139537885155441, | |
| "learning_rate": 2.3234118679127615e-07, | |
| "loss": 0.6168, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.9349470499243571, | |
| "grad_norm": 1.083778037406682, | |
| "learning_rate": 2.117724865092774e-07, | |
| "loss": 0.6113, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.9379727685325264, | |
| "grad_norm": 1.1828617403532906, | |
| "learning_rate": 1.921471959676957e-07, | |
| "loss": 0.6032, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.9409984871406959, | |
| "grad_norm": 1.0671859867441293, | |
| "learning_rate": 1.734672066928822e-07, | |
| "loss": 0.5709, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.9440242057488654, | |
| "grad_norm": 1.120674735822889, | |
| "learning_rate": 1.5573431910108404e-07, | |
| "loss": 0.6162, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.9470499243570348, | |
| "grad_norm": 1.1181661651804178, | |
| "learning_rate": 1.3895024232491338e-07, | |
| "loss": 0.6075, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.9500756429652042, | |
| "grad_norm": 1.0672311239089352, | |
| "learning_rate": 1.231165940486234e-07, | |
| "loss": 0.5741, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.9531013615733737, | |
| "grad_norm": 1.0820611496488675, | |
| "learning_rate": 1.0823490035218986e-07, | |
| "loss": 0.5796, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.9561270801815431, | |
| "grad_norm": 1.0979366103230697, | |
| "learning_rate": 9.43065955642275e-08, | |
| "loss": 0.607, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.9591527987897126, | |
| "grad_norm": 1.0391957214147853, | |
| "learning_rate": 8.133302212373961e-08, | |
| "loss": 0.5727, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.962178517397882, | |
| "grad_norm": 1.0825268730703392, | |
| "learning_rate": 6.931543045073708e-08, | |
| "loss": 0.6091, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.9652042360060514, | |
| "grad_norm": 1.1206320365705718, | |
| "learning_rate": 5.8254978825718065e-08, | |
| "loss": 0.5804, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.9682299546142209, | |
| "grad_norm": 1.0965045707696892, | |
| "learning_rate": 4.815273327803183e-08, | |
| "loss": 0.6004, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.9712556732223904, | |
| "grad_norm": 1.0830160972024074, | |
| "learning_rate": 3.900966748312862e-08, | |
| "loss": 0.5963, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.9742813918305597, | |
| "grad_norm": 1.1595612074546133, | |
| "learning_rate": 3.082666266872036e-08, | |
| "loss": 0.6178, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.9773071104387292, | |
| "grad_norm": 1.084299584807782, | |
| "learning_rate": 2.3604507529843e-08, | |
| "loss": 0.5844, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.9803328290468987, | |
| "grad_norm": 1.1123976947126495, | |
| "learning_rate": 1.7343898152841765e-08, | |
| "loss": 0.6222, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.983358547655068, | |
| "grad_norm": 1.077207200028075, | |
| "learning_rate": 1.2045437948275952e-08, | |
| "loss": 0.5762, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.9863842662632375, | |
| "grad_norm": 1.0623564397423757, | |
| "learning_rate": 7.70963759277099e-09, | |
| "loss": 0.5905, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.989409984871407, | |
| "grad_norm": 1.1937120449472454, | |
| "learning_rate": 4.336914979787832e-09, | |
| "loss": 0.6185, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.9924357034795764, | |
| "grad_norm": 1.0739987927513992, | |
| "learning_rate": 1.9275951793518154e-09, | |
| "loss": 0.5927, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.9954614220877458, | |
| "grad_norm": 1.03081894078653, | |
| "learning_rate": 4.819104067199653e-10, | |
| "loss": 0.5778, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.9984871406959153, | |
| "grad_norm": 1.0805627061619258, | |
| "learning_rate": 0.0, | |
| "loss": 0.5809, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.9984871406959153, | |
| "step": 330, | |
| "total_flos": 276171765841920.0, | |
| "train_loss": 0.6502883394559225, | |
| "train_runtime": 13177.1704, | |
| "train_samples_per_second": 3.21, | |
| "train_steps_per_second": 0.025 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 330, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 83, | |
| "total_flos": 276171765841920.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |