| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.964622641509434, | |
| "eval_steps": 500, | |
| "global_step": 424, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.009433962264150943, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 2e-05, | |
| "loss": 1.6292, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.018867924528301886, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 4e-05, | |
| "loss": 1.8541, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.02830188679245283, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 6e-05, | |
| "loss": 1.8592, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.03773584905660377, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 8e-05, | |
| "loss": 1.8298, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.04716981132075472, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 0.0001, | |
| "loss": 1.801, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.05660377358490566, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 0.00012, | |
| "loss": 1.7895, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0660377358490566, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 0.00014, | |
| "loss": 1.7302, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.07547169811320754, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 0.00016, | |
| "loss": 1.7972, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.08490566037735849, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.00018, | |
| "loss": 1.5813, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.09433962264150944, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.0002, | |
| "loss": 1.6154, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.10377358490566038, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00019999712083215463, | |
| "loss": 1.7572, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.11320754716981132, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 0.00019998848349441062, | |
| "loss": 1.7039, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.12264150943396226, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 0.00019997408848413493, | |
| "loss": 1.541, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1320754716981132, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 0.00019995393663024054, | |
| "loss": 1.6942, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.14150943396226415, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.0001999280290931388, | |
| "loss": 1.6445, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1509433962264151, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 0.00019989636736467278, | |
| "loss": 1.6148, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.16037735849056603, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 0.00019985895326803097, | |
| "loss": 1.5653, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.16981132075471697, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 0.00019981578895764273, | |
| "loss": 1.5683, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1792452830188679, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 0.00019976687691905393, | |
| "loss": 1.4737, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.18867924528301888, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 0.00019971221996878394, | |
| "loss": 1.595, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.19811320754716982, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 0.0001996518212541634, | |
| "loss": 1.5053, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.20754716981132076, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 0.00019958568425315314, | |
| "loss": 1.5134, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2169811320754717, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 0.0001995138127741436, | |
| "loss": 1.5173, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.22641509433962265, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 0.00019943621095573586, | |
| "loss": 1.5506, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.2358490566037736, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 0.00019935288326650312, | |
| "loss": 1.5697, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.24528301886792453, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 0.00019926383450473344, | |
| "loss": 1.4924, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.25471698113207547, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 0.00019916906979815347, | |
| "loss": 1.5674, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.2641509433962264, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 0.00019906859460363307, | |
| "loss": 1.5699, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.27358490566037735, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.0001989624147068713, | |
| "loss": 1.5094, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.2830188679245283, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 0.00019885053622206304, | |
| "loss": 1.5293, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.29245283018867924, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 0.00019873296559154698, | |
| "loss": 1.5266, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.3018867924528302, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 0.0001986097095854347, | |
| "loss": 1.5474, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.3113207547169811, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 0.00019848077530122083, | |
| "loss": 1.5738, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.32075471698113206, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 0.0001983461701633742, | |
| "loss": 1.5296, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.330188679245283, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 0.0001982059019229106, | |
| "loss": 1.5545, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.33962264150943394, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 0.00019805997865694614, | |
| "loss": 1.5179, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3490566037735849, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 0.00019790840876823232, | |
| "loss": 1.4534, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3584905660377358, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 0.0001977512009846721, | |
| "loss": 1.4827, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.36792452830188677, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 0.00019758836435881746, | |
| "loss": 1.4441, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.37735849056603776, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 0.00019741990826734794, | |
| "loss": 1.5114, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3867924528301887, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 0.0001972458424105307, | |
| "loss": 1.5024, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.39622641509433965, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 0.00019706617681166218, | |
| "loss": 1.5126, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.4056603773584906, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 0.00019688092181649065, | |
| "loss": 1.5009, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.41509433962264153, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 0.00019669008809262062, | |
| "loss": 1.5124, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.42452830188679247, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 0.00019649368662889855, | |
| "loss": 1.4302, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4339622641509434, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 0.00019629172873477995, | |
| "loss": 1.4321, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.44339622641509435, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 0.00019608422603967836, | |
| "loss": 1.4764, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.4528301886792453, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 0.00019587119049229557, | |
| "loss": 1.5262, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.46226415094339623, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 0.0001956526343599335, | |
| "loss": 1.3968, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.4716981132075472, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 0.0001954285702277879, | |
| "loss": 1.4958, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4811320754716981, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 0.00019519901099822372, | |
| "loss": 1.472, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.49056603773584906, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 0.00019496396989003193, | |
| "loss": 1.3241, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 0.00019472346043766865, | |
| "loss": 1.4498, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5094339622641509, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 0.00019447749649047542, | |
| "loss": 1.4044, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5188679245283019, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 0.00019422609221188207, | |
| "loss": 1.4725, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5283018867924528, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 0.00019396926207859084, | |
| "loss": 1.3035, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5377358490566038, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 0.00019370702087974302, | |
| "loss": 1.4331, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5471698113207547, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.00019343938371606712, | |
| "loss": 1.4268, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5566037735849056, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.00019316636599900946, | |
| "loss": 1.4898, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.5660377358490566, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.00019288798344984672, | |
| "loss": 1.4587, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5754716981132075, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.00019260425209878052, | |
| "loss": 1.2718, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5849056603773585, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 0.00019231518828401458, | |
| "loss": 1.4943, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.5943396226415094, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 0.00019202080865081368, | |
| "loss": 1.4563, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6037735849056604, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 0.00019172113015054532, | |
| "loss": 1.5265, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.6132075471698113, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 0.0001914161700397035, | |
| "loss": 1.3905, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6226415094339622, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 0.00019110594587891519, | |
| "loss": 1.4768, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6320754716981132, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 0.0001907904755319289, | |
| "loss": 1.3191, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6415094339622641, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 0.00019046977716458626, | |
| "loss": 1.3206, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6509433962264151, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.00019014386924377582, | |
| "loss": 1.331, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.660377358490566, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 0.0001898127705363696, | |
| "loss": 1.373, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6698113207547169, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.0001894765001081428, | |
| "loss": 1.4195, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6792452830188679, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 0.0001891350773226754, | |
| "loss": 1.3853, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.6886792452830188, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.0001887885218402375, | |
| "loss": 1.496, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.6981132075471698, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 0.00018843685361665723, | |
| "loss": 1.4916, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.7075471698113207, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 0.00018808009290217136, | |
| "loss": 1.2506, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7169811320754716, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 0.00018771826024025946, | |
| "loss": 1.3273, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.7264150943396226, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 0.00018735137646646078, | |
| "loss": 1.343, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.7358490566037735, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 0.00018697946270717467, | |
| "loss": 1.4744, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7452830188679245, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.00018660254037844388, | |
| "loss": 1.3758, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.7547169811320755, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 0.00018622063118472134, | |
| "loss": 1.3836, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7641509433962265, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 0.00018583375711762052, | |
| "loss": 1.347, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7735849056603774, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 0.00018544194045464886, | |
| "loss": 1.4363, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7830188679245284, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 0.0001850452037579251, | |
| "loss": 1.4564, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.7924528301886793, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 0.00018464356987288013, | |
| "loss": 1.4342, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.8018867924528302, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 0.00018423706192694116, | |
| "loss": 1.4234, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8113207547169812, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 0.00018382570332820043, | |
| "loss": 1.3817, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.8207547169811321, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 0.00018340951776406694, | |
| "loss": 1.4327, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.8301886792452831, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 0.00018298852919990252, | |
| "loss": 1.3911, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.839622641509434, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.00018256276187764197, | |
| "loss": 1.4534, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.8490566037735849, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 0.0001821322403143969, | |
| "loss": 1.3534, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8584905660377359, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.0001816969893010442, | |
| "loss": 1.4886, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8679245283018868, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 0.0001812570339007983, | |
| "loss": 1.3435, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8773584905660378, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.00018081239944776805, | |
| "loss": 1.2901, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.8867924528301887, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 0.00018036311154549784, | |
| "loss": 1.4074, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.8962264150943396, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 0.00017990919606549328, | |
| "loss": 1.3902, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9056603773584906, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 0.00017945067914573146, | |
| "loss": 1.4234, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.9150943396226415, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 0.00017898758718915586, | |
| "loss": 1.2415, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.9245283018867925, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 0.0001785199468621559, | |
| "loss": 1.4503, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.9339622641509434, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 0.00017804778509303138, | |
| "loss": 1.415, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.9433962264150944, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 0.000177571129070442, | |
| "loss": 1.3419, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9528301886792453, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 0.00017709000624184162, | |
| "loss": 1.4011, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9622641509433962, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.0001766044443118978, | |
| "loss": 1.3546, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.9716981132075472, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 0.00017611447124089649, | |
| "loss": 1.3989, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9811320754716981, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 0.00017562011524313185, | |
| "loss": 1.3564, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.9905660377358491, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 0.0001751214047852818, | |
| "loss": 1.3605, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 0.00017461836858476856, | |
| "loss": 1.4459, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.009433962264151, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 0.00017411103560810526, | |
| "loss": 1.3842, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.0070754716981132, | |
| "grad_norm": 0.2255859375, | |
| "learning_rate": 0.00017359943506922774, | |
| "loss": 1.3091, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.0165094339622642, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 0.00017308359642781242, | |
| "loss": 1.3549, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.025943396226415, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 0.0001725635493875799, | |
| "loss": 1.2693, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0353773584905661, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 0.00017203932389458454, | |
| "loss": 1.2991, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.044811320754717, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 0.00017151095013548994, | |
| "loss": 1.379, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.054245283018868, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 0.0001709784585358309, | |
| "loss": 1.2804, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0636792452830188, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 0.00017044187975826124, | |
| "loss": 1.2645, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.0731132075471699, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 0.00016990124470078822, | |
| "loss": 1.1872, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.0825471698113207, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 0.0001693565844949933, | |
| "loss": 1.2613, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.0919811320754718, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 0.0001688079305042395, | |
| "loss": 1.2765, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.1014150943396226, | |
| "grad_norm": 0.216796875, | |
| "learning_rate": 0.00016825531432186543, | |
| "loss": 1.2478, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.1108490566037736, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 0.0001676987677693659, | |
| "loss": 1.3338, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.1202830188679245, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 0.0001671383228945597, | |
| "loss": 1.2246, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.1297169811320755, | |
| "grad_norm": 0.2197265625, | |
| "learning_rate": 0.00016657401196974405, | |
| "loss": 1.2763, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.1391509433962264, | |
| "grad_norm": 0.220703125, | |
| "learning_rate": 0.00016600586748983641, | |
| "loss": 1.1468, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.1485849056603774, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 0.00016543392217050314, | |
| "loss": 1.3413, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.1580188679245282, | |
| "grad_norm": 0.2314453125, | |
| "learning_rate": 0.0001648582089462756, | |
| "loss": 1.418, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.1674528301886793, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 0.00016427876096865394, | |
| "loss": 1.2445, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.1768867924528301, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 0.00016369561160419784, | |
| "loss": 1.23, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.1863207547169812, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 0.00016310879443260528, | |
| "loss": 1.3247, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.195754716981132, | |
| "grad_norm": 0.2265625, | |
| "learning_rate": 0.0001625183432447789, | |
| "loss": 1.2278, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.205188679245283, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 0.0001619242920408802, | |
| "loss": 1.1869, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.2146226415094339, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 0.00016132667502837165, | |
| "loss": 1.2633, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.224056603773585, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 0.00016072552662004696, | |
| "loss": 1.2912, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.2334905660377358, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 0.00016012088143204953, | |
| "loss": 1.3428, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.2429245283018868, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 0.00015951277428187898, | |
| "loss": 1.2032, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.2523584905660377, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.00015890124018638638, | |
| "loss": 1.2246, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.2617924528301887, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 0.00015828631435975784, | |
| "loss": 1.2724, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.2712264150943398, | |
| "grad_norm": 0.2314453125, | |
| "learning_rate": 0.00015766803221148673, | |
| "loss": 1.2384, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.2806603773584906, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 0.0001570464293443346, | |
| "loss": 1.2844, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.2900943396226414, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.00015642154155228122, | |
| "loss": 1.2727, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.2995283018867925, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 0.00015579340481846336, | |
| "loss": 1.2884, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.3089622641509435, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 0.00015516205531310273, | |
| "loss": 1.3399, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.3183962264150944, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 0.00015452752939142328, | |
| "loss": 1.3155, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.3278301886792452, | |
| "grad_norm": 0.23046875, | |
| "learning_rate": 0.00015388986359155758, | |
| "loss": 1.1984, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.3372641509433962, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 0.00015324909463244296, | |
| "loss": 1.1836, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.3466981132075473, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 0.00015260525941170712, | |
| "loss": 1.27, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.3561320754716981, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.00015195839500354335, | |
| "loss": 1.3095, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.365566037735849, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.0001513085386565758, | |
| "loss": 1.168, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.375, | |
| "grad_norm": 0.2333984375, | |
| "learning_rate": 0.00015065572779171432, | |
| "loss": 1.2784, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.384433962264151, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 1.2809, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.3938679245283019, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 0.00014934139304044033, | |
| "loss": 1.3085, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.4033018867924527, | |
| "grad_norm": 0.228515625, | |
| "learning_rate": 0.00014867994483783485, | |
| "loss": 1.214, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4127358490566038, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.00014801569348059157, | |
| "loss": 1.2022, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.4221698113207548, | |
| "grad_norm": 0.2314453125, | |
| "learning_rate": 0.0001473486772185334, | |
| "loss": 1.2022, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.4316037735849056, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 0.00014667893446069588, | |
| "loss": 1.3266, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.4410377358490565, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 0.00014600650377311522, | |
| "loss": 1.2686, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.4504716981132075, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.00014533142387660773, | |
| "loss": 1.2602, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.4599056603773586, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 0.00014465373364454001, | |
| "loss": 1.2541, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.4693396226415094, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 0.00014397347210059057, | |
| "loss": 1.1867, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.4787735849056602, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 0.00014329067841650274, | |
| "loss": 1.3164, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.4882075471698113, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 0.00014260539190982886, | |
| "loss": 1.3074, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.4976415094339623, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 0.00014191765204166643, | |
| "loss": 1.2422, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5070754716981132, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 0.00014122749841438575, | |
| "loss": 1.2188, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.516509433962264, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 0.00014053497076934948, | |
| "loss": 1.2709, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.525943396226415, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.00013984010898462416, | |
| "loss": 1.2864, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.5353773584905661, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 0.00013914295307268396, | |
| "loss": 1.2356, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.544811320754717, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 0.0001384435431781065, | |
| "loss": 1.2602, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.5542452830188678, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 0.00013774191957526143, | |
| "loss": 1.266, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.5636792452830188, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 0.00013703812266599113, | |
| "loss": 1.2911, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.5731132075471699, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 0.00013633219297728416, | |
| "loss": 1.3344, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.5825471698113207, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 0.00013562417115894172, | |
| "loss": 1.1809, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.5919811320754715, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00013491409798123687, | |
| "loss": 1.2117, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6014150943396226, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.00013420201433256689, | |
| "loss": 1.1622, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.6108490566037736, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 0.00013348796121709862, | |
| "loss": 1.1991, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.6202830188679245, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.0001327719797524075, | |
| "loss": 1.0869, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.6297169811320755, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 0.00013205411116710972, | |
| "loss": 1.2271, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.6391509433962264, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.00013133439679848823, | |
| "loss": 1.3048, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.6485849056603774, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.00013061287809011242, | |
| "loss": 1.1293, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.6580188679245285, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.0001298895965894516, | |
| "loss": 1.2416, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.6674528301886793, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 0.0001291645939454825, | |
| "loss": 1.2559, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.6768867924528301, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 0.0001284379119062912, | |
| "loss": 1.2819, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.6863207547169812, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 0.0001277095923166689, | |
| "loss": 1.2085, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.6957547169811322, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.00012697967711570242, | |
| "loss": 1.2698, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.705188679245283, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.00012624820833435937, | |
| "loss": 1.2076, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.7146226415094339, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.0001255152280930676, | |
| "loss": 1.2315, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.724056603773585, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 0.00012478077859929, | |
| "loss": 1.2389, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.733490566037736, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00012404490214509386, | |
| "loss": 1.2874, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.7429245283018868, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.00012330764110471566, | |
| "loss": 1.1518, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.7523584905660377, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.00012256903793212107, | |
| "loss": 1.2779, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.7617924528301887, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.00012182913515856015, | |
| "loss": 1.1925, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.7712264150943398, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00012108797539011847, | |
| "loss": 1.2957, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.7806603773584906, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.0001203456013052634, | |
| "loss": 1.2427, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.7900943396226414, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 0.00011960205565238684, | |
| "loss": 1.0626, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.7995283018867925, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00011885738124734358, | |
| "loss": 1.2465, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.8089622641509435, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 0.00011811162097098558, | |
| "loss": 1.1583, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.8183962264150944, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 1.2873, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.8278301886792452, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.00011661701463790142, | |
| "loss": 1.2776, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.8372641509433962, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.00011586825464562514, | |
| "loss": 1.1212, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.8466981132075473, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 0.0001151185809059781, | |
| "loss": 1.1373, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.8561320754716981, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 0.00011436803658769082, | |
| "loss": 1.2672, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.865566037735849, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.00011361666490962468, | |
| "loss": 1.2042, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.875, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 0.00011286450913828312, | |
| "loss": 1.2579, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.884433962264151, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 0.00011211161258532041, | |
| "loss": 1.2116, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.8938679245283019, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.00011135801860504749, | |
| "loss": 1.2045, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.9033018867924527, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 0.00011060377059193547, | |
| "loss": 1.1928, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.9127358490566038, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00010984891197811687, | |
| "loss": 1.3457, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.9221698113207548, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.0001090934862308847, | |
| "loss": 1.2525, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.9316037735849056, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 0.00010833753685018935, | |
| "loss": 1.2322, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.9410377358490565, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 0.00010758110736613385, | |
| "loss": 1.2157, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.9504716981132075, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.0001068242413364671, | |
| "loss": 1.1082, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.9599056603773586, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00010606698234407586, | |
| "loss": 1.2336, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.9693396226415094, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 0.00010530937399447496, | |
| "loss": 1.1619, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.9787735849056602, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.00010455145991329638, | |
| "loss": 1.2344, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.9882075471698113, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.00010379328374377715, | |
| "loss": 1.2143, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.9976415094339623, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.00010303488914424624, | |
| "loss": 1.231, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.007075471698113, | |
| "grad_norm": 0.2392578125, | |
| "learning_rate": 0.00010227631978561056, | |
| "loss": 1.1562, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.0023584905660377, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00010151761934884028, | |
| "loss": 1.1339, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.0117924528301887, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00010075883152245334, | |
| "loss": 1.1717, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.0212264150943398, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.0001, | |
| "loss": 1.1575, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.0306603773584904, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 9.92411684775467e-05, | |
| "loss": 1.0906, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.0400943396226414, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 9.848238065115975e-05, | |
| "loss": 1.1255, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.0495283018867925, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 9.772368021438943e-05, | |
| "loss": 1.0733, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.0589622641509435, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 9.696511085575377e-05, | |
| "loss": 1.1415, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.068396226415094, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 9.620671625622288e-05, | |
| "loss": 1.0607, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.077830188679245, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 9.544854008670367e-05, | |
| "loss": 1.0438, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.0872641509433962, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 9.469062600552509e-05, | |
| "loss": 1.1176, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.0966981132075473, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 9.393301765592415e-05, | |
| "loss": 1.0588, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.106132075471698, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 9.317575866353292e-05, | |
| "loss": 1.0319, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.115566037735849, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 9.241889263386618e-05, | |
| "loss": 1.1098, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.125, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 9.166246314981066e-05, | |
| "loss": 1.1213, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.134433962264151, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 9.09065137691153e-05, | |
| "loss": 1.0372, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.143867924528302, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 9.015108802188313e-05, | |
| "loss": 1.1706, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.1533018867924527, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 8.939622940806455e-05, | |
| "loss": 1.0834, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.1627358490566038, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 8.86419813949525e-05, | |
| "loss": 1.044, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.172169811320755, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 8.788838741467962e-05, | |
| "loss": 1.0498, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.1816037735849054, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 8.713549086171691e-05, | |
| "loss": 1.0347, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.1910377358490565, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 8.638333509037536e-05, | |
| "loss": 1.0902, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.2004716981132075, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 8.563196341230919e-05, | |
| "loss": 0.9978, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.2099056603773586, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 8.488141909402191e-05, | |
| "loss": 1.1299, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.2193396226415096, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 8.413174535437487e-05, | |
| "loss": 0.9632, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.2287735849056602, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 8.33829853620986e-05, | |
| "loss": 1.029, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.2382075471698113, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 8.263518223330697e-05, | |
| "loss": 1.0729, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.2476415094339623, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 8.188837902901442e-05, | |
| "loss": 1.0296, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.2570754716981134, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 8.114261875265643e-05, | |
| "loss": 1.0439, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.266509433962264, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 8.039794434761318e-05, | |
| "loss": 1.0191, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.275943396226415, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 7.965439869473664e-05, | |
| "loss": 1.1443, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.285377358490566, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 7.891202460988158e-05, | |
| "loss": 1.1637, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.294811320754717, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 7.817086484143986e-05, | |
| "loss": 1.1807, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.3042452830188678, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 7.743096206787894e-05, | |
| "loss": 1.0508, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.313679245283019, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 7.669235889528436e-05, | |
| "loss": 1.0764, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.32311320754717, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 7.595509785490617e-05, | |
| "loss": 1.0247, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.332547169811321, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 7.521922140071002e-05, | |
| "loss": 1.0341, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.3419811320754715, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 7.448477190693238e-05, | |
| "loss": 1.0841, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.3514150943396226, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 7.375179166564063e-05, | |
| "loss": 1.001, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.3608490566037736, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 7.302032288429756e-05, | |
| "loss": 1.073, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.3702830188679247, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 7.229040768333115e-05, | |
| "loss": 1.139, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.3797169811320753, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 7.156208809370883e-05, | |
| "loss": 1.035, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.3891509433962264, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 7.08354060545175e-05, | |
| "loss": 1.0059, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.3985849056603774, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 7.011040341054845e-05, | |
| "loss": 1.0855, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.4080188679245285, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 6.93871219098876e-05, | |
| "loss": 1.0351, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.417452830188679, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 6.866560320151179e-05, | |
| "loss": 1.0143, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.42688679245283, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 6.79458888328903e-05, | |
| "loss": 1.106, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.436320754716981, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 6.722802024759252e-05, | |
| "loss": 1.0725, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.4457547169811322, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 6.651203878290139e-05, | |
| "loss": 1.0353, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.455188679245283, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 6.579798566743314e-05, | |
| "loss": 1.0543, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.464622641509434, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 6.508590201876317e-05, | |
| "loss": 1.135, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.474056603773585, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 6.437582884105835e-05, | |
| "loss": 1.0466, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.483490566037736, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 6.366780702271589e-05, | |
| "loss": 1.0234, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.4929245283018866, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 6.29618773340089e-05, | |
| "loss": 1.0979, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.5023584905660377, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 6.225808042473858e-05, | |
| "loss": 1.066, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.5117924528301887, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 6.155645682189351e-05, | |
| "loss": 1.1694, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.5212264150943398, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 6.085704692731609e-05, | |
| "loss": 0.9722, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.5306603773584904, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 6.015989101537586e-05, | |
| "loss": 1.0155, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.5400943396226414, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 5.9465029230650534e-05, | |
| "loss": 1.063, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.5495283018867925, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 5.877250158561425e-05, | |
| "loss": 1.0669, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.5589622641509435, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 5.8082347958333625e-05, | |
| "loss": 1.0505, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.5683962264150946, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 5.73946080901712e-05, | |
| "loss": 1.177, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.577830188679245, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 5.670932158349731e-05, | |
| "loss": 1.0335, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.5872641509433962, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 5.602652789940941e-05, | |
| "loss": 1.0383, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.5966981132075473, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 5.5346266355459995e-05, | |
| "loss": 0.9778, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.606132075471698, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 5.466857612339229e-05, | |
| "loss": 0.9718, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.615566037735849, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 5.399349622688479e-05, | |
| "loss": 1.1602, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.625, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 5.332106553930414e-05, | |
| "loss": 1.0645, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.634433962264151, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 5.26513227814666e-05, | |
| "loss": 0.9681, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.643867924528302, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 5.1984306519408456e-05, | |
| "loss": 1.0741, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.6533018867924527, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 5.1320055162165115e-05, | |
| "loss": 1.0404, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.6627358490566038, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 5.065860695955971e-05, | |
| "loss": 1.114, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.672169811320755, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 5.000000000000002e-05, | |
| "loss": 1.0966, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.6816037735849054, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 4.934427220828571e-05, | |
| "loss": 1.0322, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.6910377358490565, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 4.869146134342426e-05, | |
| "loss": 1.0087, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.7004716981132075, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 4.804160499645667e-05, | |
| "loss": 1.1284, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.7099056603773586, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 4.739474058829289e-05, | |
| "loss": 0.9876, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.7193396226415096, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 4.675090536755705e-05, | |
| "loss": 1.1121, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.7287735849056602, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 4.611013640844245e-05, | |
| "loss": 1.0433, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.7382075471698113, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 4.547247060857675e-05, | |
| "loss": 0.9529, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.7476415094339623, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 4.483794468689728e-05, | |
| "loss": 1.0347, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.757075471698113, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 4.420659518153667e-05, | |
| "loss": 1.0604, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.766509433962264, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 4.357845844771881e-05, | |
| "loss": 1.1131, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.775943396226415, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 4.295357065566543e-05, | |
| "loss": 1.0776, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.785377358490566, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 4.2331967788513295e-05, | |
| "loss": 1.0204, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.794811320754717, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 4.1713685640242165e-05, | |
| "loss": 1.0161, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.8042452830188678, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 4.109875981361363e-05, | |
| "loss": 1.1091, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.813679245283019, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 4.048722571812105e-05, | |
| "loss": 1.0495, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.82311320754717, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 3.987911856795047e-05, | |
| "loss": 1.0086, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.8325471698113205, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 3.927447337995304e-05, | |
| "loss": 0.9878, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.8419811320754715, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 3.8673324971628357e-05, | |
| "loss": 0.9856, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.8514150943396226, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 3.8075707959119846e-05, | |
| "loss": 1.0848, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.8608490566037736, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 3.7481656755221125e-05, | |
| "loss": 1.0856, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.8702830188679247, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 3.689120556739475e-05, | |
| "loss": 0.9755, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.8797169811320753, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 3.630438839580217e-05, | |
| "loss": 1.0983, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.8891509433962264, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 3.5721239031346066e-05, | |
| "loss": 1.0877, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.8985849056603774, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 3.5141791053724405e-05, | |
| "loss": 1.1485, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.9080188679245285, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 3.456607782949689e-05, | |
| "loss": 1.0692, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.9174528301886795, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 3.399413251016359e-05, | |
| "loss": 1.0783, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.92688679245283, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 3.342598803025595e-05, | |
| "loss": 1.1049, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 2.936320754716981, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 3.2861677105440336e-05, | |
| "loss": 1.143, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 2.9457547169811322, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 3.2301232230634104e-05, | |
| "loss": 1.0474, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.955188679245283, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 3.174468567813461e-05, | |
| "loss": 1.078, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 2.964622641509434, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 3.119206949576052e-05, | |
| "loss": 1.0219, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 2.974056603773585, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 3.0643415505006735e-05, | |
| "loss": 1.061, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 2.983490566037736, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 3.009875529921181e-05, | |
| "loss": 1.0573, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 2.992924528301887, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 2.9558120241738784e-05, | |
| "loss": 1.0131, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.0023584905660377, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 2.90215414641691e-05, | |
| "loss": 1.0359, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 3.0023584905660377, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 2.8489049864510054e-05, | |
| "loss": 1.07, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 3.0117924528301887, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 2.7960676105415472e-05, | |
| "loss": 1.0113, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 3.0212264150943398, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 2.7436450612420095e-05, | |
| "loss": 0.9147, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 3.0306603773584904, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 2.691640357218759e-05, | |
| "loss": 0.8318, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 3.0400943396226414, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 2.640056493077231e-05, | |
| "loss": 0.9546, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 3.0495283018867925, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 2.5888964391894766e-05, | |
| "loss": 1.0179, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 3.0589622641509435, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 2.5381631415231454e-05, | |
| "loss": 0.8373, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 3.068396226415094, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 2.4878595214718236e-05, | |
| "loss": 0.8973, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 3.077830188679245, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 2.4379884756868167e-05, | |
| "loss": 0.9058, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 3.0872641509433962, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 2.3885528759103538e-05, | |
| "loss": 0.9879, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 3.0966981132075473, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 2.339555568810221e-05, | |
| "loss": 0.9923, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 3.106132075471698, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 2.2909993758158412e-05, | |
| "loss": 0.9279, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 3.115566037735849, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 2.242887092955801e-05, | |
| "loss": 0.9026, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 3.125, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 2.1952214906968627e-05, | |
| "loss": 0.9476, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 3.134433962264151, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 2.1480053137844115e-05, | |
| "loss": 0.9352, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 3.143867924528302, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 2.101241281084416e-05, | |
| "loss": 0.9964, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 3.1533018867924527, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 2.054932085426856e-05, | |
| "loss": 1.0295, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 3.1627358490566038, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 2.0090803934506764e-05, | |
| "loss": 0.9968, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 3.172169811320755, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 1.9636888454502178e-05, | |
| "loss": 0.8935, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.1816037735849054, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 1.9187600552231955e-05, | |
| "loss": 0.9172, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 3.1910377358490565, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 1.8742966099201697e-05, | |
| "loss": 0.9573, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 3.2004716981132075, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 1.8303010698955804e-05, | |
| "loss": 0.9661, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 3.2099056603773586, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 1.7867759685603114e-05, | |
| "loss": 0.9576, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 3.2193396226415096, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.7437238122358057e-05, | |
| "loss": 1.0133, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 3.2287735849056602, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 1.7011470800097496e-05, | |
| "loss": 0.9558, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 3.2382075471698113, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 1.659048223593308e-05, | |
| "loss": 0.8893, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 3.2476415094339623, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 1.6174296671799572e-05, | |
| "loss": 1.0161, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 3.2570754716981134, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 1.5762938073058853e-05, | |
| "loss": 1.0262, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 3.266509433962264, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 1.5356430127119913e-05, | |
| "loss": 0.9567, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.275943396226415, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 1.4954796242074898e-05, | |
| "loss": 0.9701, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 3.285377358490566, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 1.4558059545351143e-05, | |
| "loss": 0.9455, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 3.294811320754717, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 1.4166242882379476e-05, | |
| "loss": 0.9922, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 3.3042452830188678, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 1.3779368815278647e-05, | |
| "loss": 0.9102, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 3.313679245283019, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 1.339745962155613e-05, | |
| "loss": 0.9873, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 3.32311320754717, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 1.302053729282533e-05, | |
| "loss": 0.9758, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 3.332547169811321, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 1.2648623533539261e-05, | |
| "loss": 0.851, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 3.3419811320754715, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 1.2281739759740574e-05, | |
| "loss": 1.0054, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 3.3514150943396226, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 1.1919907097828653e-05, | |
| "loss": 0.8892, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 3.3608490566037736, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 1.1563146383342772e-05, | |
| "loss": 0.9823, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.3702830188679247, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 1.1211478159762478e-05, | |
| "loss": 0.8435, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 3.3797169811320753, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 1.0864922677324618e-05, | |
| "loss": 1.0188, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 3.3891509433962264, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 1.0523499891857225e-05, | |
| "loss": 1.0157, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 3.3985849056603774, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 1.01872294636304e-05, | |
| "loss": 0.9674, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 3.4080188679245285, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.856130756224213e-06, | |
| "loss": 0.9332, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 3.417452830188679, | |
| "grad_norm": 0.375, | |
| "learning_rate": 9.530222835413738e-06, | |
| "loss": 0.9148, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.42688679245283, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 9.209524468071096e-06, | |
| "loss": 1.0542, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 3.436320754716981, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 8.894054121084838e-06, | |
| "loss": 0.9474, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 3.4457547169811322, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 8.58382996029652e-06, | |
| "loss": 0.9664, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 3.455188679245283, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 8.278869849454718e-06, | |
| "loss": 0.9389, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.464622641509434, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 7.97919134918632e-06, | |
| "loss": 0.9574, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 3.474056603773585, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 7.684811715985429e-06, | |
| "loss": 0.9318, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 3.483490566037736, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 7.395747901219474e-06, | |
| "loss": 0.8918, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 3.4929245283018866, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 7.1120165501533e-06, | |
| "loss": 1.0351, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 3.5023584905660377, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 6.833634000990541e-06, | |
| "loss": 1.0396, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 3.5117924528301887, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 6.560616283932897e-06, | |
| "loss": 0.9884, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 3.5212264150943398, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 6.292979120256992e-06, | |
| "loss": 0.8938, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 3.5306603773584904, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 6.030737921409169e-06, | |
| "loss": 0.9728, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 3.5400943396226414, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 5.77390778811796e-06, | |
| "loss": 1.0086, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 3.5495283018867925, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 5.52250350952459e-06, | |
| "loss": 1.0073, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.5589622641509435, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 5.276539562331384e-06, | |
| "loss": 0.9953, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 3.5683962264150946, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 5.036030109968082e-06, | |
| "loss": 0.9377, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 3.577830188679245, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 4.800989001776324e-06, | |
| "loss": 0.9637, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 3.5872641509433962, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 4.5714297722121106e-06, | |
| "loss": 0.8614, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 3.5966981132075473, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 4.347365640066525e-06, | |
| "loss": 0.9701, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 3.606132075471698, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 4.128809507704445e-06, | |
| "loss": 0.9451, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 3.615566037735849, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 3.915773960321634e-06, | |
| "loss": 0.9175, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 3.625, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 3.7082712652200867e-06, | |
| "loss": 0.8858, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 3.634433962264151, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 3.5063133711014882e-06, | |
| "loss": 0.9181, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 3.643867924528302, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 3.3099119073793928e-06, | |
| "loss": 0.9641, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.6533018867924527, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 3.119078183509372e-06, | |
| "loss": 1.0059, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 3.6627358490566038, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 2.9338231883378366e-06, | |
| "loss": 0.9977, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 3.672169811320755, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 2.7541575894693194e-06, | |
| "loss": 0.9899, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 3.6816037735849054, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 2.580091732652101e-06, | |
| "loss": 0.9088, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 3.6910377358490565, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 2.4116356411825525e-06, | |
| "loss": 0.9484, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 3.7004716981132075, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 2.248799015327907e-06, | |
| "loss": 0.9853, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 3.7099056603773586, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 2.091591231767709e-06, | |
| "loss": 0.9616, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 3.7193396226415096, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 1.9400213430538773e-06, | |
| "loss": 0.9615, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 3.7287735849056602, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.7940980770894122e-06, | |
| "loss": 1.003, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 3.7382075471698113, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 1.6538298366257976e-06, | |
| "loss": 0.9496, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.7476415094339623, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 1.5192246987791981e-06, | |
| "loss": 0.9563, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 3.757075471698113, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 1.3902904145653096e-06, | |
| "loss": 0.9478, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 3.766509433962264, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 1.2670344084530383e-06, | |
| "loss": 0.9733, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 3.775943396226415, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 1.1494637779369766e-06, | |
| "loss": 0.8937, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 3.785377358490566, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.0375852931286956e-06, | |
| "loss": 0.9241, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 3.794811320754717, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 9.314053963669245e-07, | |
| "loss": 0.9335, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 3.8042452830188678, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 8.309302018465581e-07, | |
| "loss": 0.9593, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 3.813679245283019, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 7.361654952665609e-07, | |
| "loss": 0.9529, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 3.82311320754717, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 6.471167334968886e-07, | |
| "loss": 0.9608, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 3.8325471698113205, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 5.637890442641402e-07, | |
| "loss": 0.9223, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.8419811320754715, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 4.861872258564049e-07, | |
| "loss": 1.0127, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 3.8514150943396226, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 4.143157468468717e-07, | |
| "loss": 0.9592, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 3.8608490566037736, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 3.481787458365915e-07, | |
| "loss": 0.9963, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 3.8702830188679247, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 2.877800312160783e-07, | |
| "loss": 0.9346, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 3.8797169811320753, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 2.3312308094607382e-07, | |
| "loss": 0.8629, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 3.8891509433962264, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 1.8421104235727405e-07, | |
| "loss": 0.9351, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 3.8985849056603774, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.4104673196903005e-07, | |
| "loss": 1.0274, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 3.9080188679245285, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 1.0363263532724432e-07, | |
| "loss": 0.9699, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 3.9174528301886795, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 7.197090686119623e-08, | |
| "loss": 0.9988, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 3.92688679245283, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 4.606336975948589e-08, | |
| "loss": 0.9501, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.936320754716981, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 2.5911515865084667e-08, | |
| "loss": 0.9581, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 3.9457547169811322, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.1516505589381776e-08, | |
| "loss": 0.9865, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 3.955188679245283, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 2.8791678453821135e-09, | |
| "loss": 0.9402, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 3.964622641509434, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.0, | |
| "loss": 0.9966, | |
| "step": 424 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 424, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 106, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.322438582095053e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |