sgptv / checkpoint-150 /trainer_state.json
jinghan23's picture
Upload to Hugging Face
7066eea verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.4778972520908005,
"eval_steps": 500,
"global_step": 150,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"grad_norm": 10.092560768127441,
"learning_rate": 1.0000000000000002e-06,
"loss": 1.9497,
"step": 1
},
{
"epoch": 0.01,
"grad_norm": 8.963287353515625,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.951,
"step": 2
},
{
"epoch": 0.01,
"grad_norm": 9.27600383758545,
"learning_rate": 3e-06,
"loss": 1.9345,
"step": 3
},
{
"epoch": 0.01,
"grad_norm": 5.632405757904053,
"learning_rate": 4.000000000000001e-06,
"loss": 1.9157,
"step": 4
},
{
"epoch": 0.02,
"grad_norm": 5.668580055236816,
"learning_rate": 5e-06,
"loss": 1.9026,
"step": 5
},
{
"epoch": 0.02,
"grad_norm": 3.8046553134918213,
"learning_rate": 6e-06,
"loss": 1.8923,
"step": 6
},
{
"epoch": 0.02,
"grad_norm": 4.357985973358154,
"learning_rate": 7e-06,
"loss": 1.8241,
"step": 7
},
{
"epoch": 0.03,
"grad_norm": 4.685062885284424,
"learning_rate": 8.000000000000001e-06,
"loss": 1.8467,
"step": 8
},
{
"epoch": 0.03,
"grad_norm": 4.768229961395264,
"learning_rate": 9e-06,
"loss": 1.8199,
"step": 9
},
{
"epoch": 0.03,
"grad_norm": 4.796407699584961,
"learning_rate": 1e-05,
"loss": 1.8374,
"step": 10
},
{
"epoch": 0.04,
"grad_norm": 5.7536139488220215,
"learning_rate": 9.999731248679734e-06,
"loss": 1.779,
"step": 11
},
{
"epoch": 0.04,
"grad_norm": 4.202663898468018,
"learning_rate": 9.99892502360984e-06,
"loss": 1.7579,
"step": 12
},
{
"epoch": 0.04,
"grad_norm": 2.9114131927490234,
"learning_rate": 9.99758141145994e-06,
"loss": 1.7433,
"step": 13
},
{
"epoch": 0.04,
"grad_norm": 2.3823723793029785,
"learning_rate": 9.995700556669052e-06,
"loss": 1.7212,
"step": 14
},
{
"epoch": 0.05,
"grad_norm": 2.3254876136779785,
"learning_rate": 9.993282661430058e-06,
"loss": 1.7218,
"step": 15
},
{
"epoch": 0.05,
"grad_norm": 2.053166151046753,
"learning_rate": 9.990327985667972e-06,
"loss": 1.7256,
"step": 16
},
{
"epoch": 0.05,
"grad_norm": 2.3782012462615967,
"learning_rate": 9.986836847012001e-06,
"loss": 1.713,
"step": 17
},
{
"epoch": 0.06,
"grad_norm": 2.1311683654785156,
"learning_rate": 9.98280962076139e-06,
"loss": 1.6785,
"step": 18
},
{
"epoch": 0.06,
"grad_norm": 2.325747489929199,
"learning_rate": 9.978246739845095e-06,
"loss": 1.7167,
"step": 19
},
{
"epoch": 0.06,
"grad_norm": 2.1330366134643555,
"learning_rate": 9.973148694775217e-06,
"loss": 1.676,
"step": 20
},
{
"epoch": 0.07,
"grad_norm": 2.2632806301116943,
"learning_rate": 9.967516033594295e-06,
"loss": 1.7033,
"step": 21
},
{
"epoch": 0.07,
"grad_norm": 2.6582744121551514,
"learning_rate": 9.961349361816384e-06,
"loss": 1.6957,
"step": 22
},
{
"epoch": 0.07,
"grad_norm": 1.9663337469100952,
"learning_rate": 9.954649342361952e-06,
"loss": 1.6729,
"step": 23
},
{
"epoch": 0.08,
"grad_norm": 2.2557435035705566,
"learning_rate": 9.947416695486633e-06,
"loss": 1.6399,
"step": 24
},
{
"epoch": 0.08,
"grad_norm": 2.090054512023926,
"learning_rate": 9.939652198703785e-06,
"loss": 1.6792,
"step": 25
},
{
"epoch": 0.08,
"grad_norm": 2.1648776531219482,
"learning_rate": 9.93135668670091e-06,
"loss": 1.6625,
"step": 26
},
{
"epoch": 0.09,
"grad_norm": 2.1209850311279297,
"learning_rate": 9.92253105124993e-06,
"loss": 1.7057,
"step": 27
},
{
"epoch": 0.09,
"grad_norm": 2.2777392864227295,
"learning_rate": 9.91317624111132e-06,
"loss": 1.6052,
"step": 28
},
{
"epoch": 0.09,
"grad_norm": 2.1739561557769775,
"learning_rate": 9.903293261932106e-06,
"loss": 1.6139,
"step": 29
},
{
"epoch": 0.1,
"grad_norm": 2.4518871307373047,
"learning_rate": 9.89288317613777e-06,
"loss": 1.6759,
"step": 30
},
{
"epoch": 0.1,
"grad_norm": 1.795519471168518,
"learning_rate": 9.881947102818036e-06,
"loss": 1.7036,
"step": 31
},
{
"epoch": 0.1,
"grad_norm": 2.150223731994629,
"learning_rate": 9.870486217606557e-06,
"loss": 1.6469,
"step": 32
},
{
"epoch": 0.11,
"grad_norm": 1.91805899143219,
"learning_rate": 9.858501752554548e-06,
"loss": 1.6353,
"step": 33
},
{
"epoch": 0.11,
"grad_norm": 1.8129810094833374,
"learning_rate": 9.845994995998332e-06,
"loss": 1.6551,
"step": 34
},
{
"epoch": 0.11,
"grad_norm": 2.1308679580688477,
"learning_rate": 9.83296729242084e-06,
"loss": 1.617,
"step": 35
},
{
"epoch": 0.11,
"grad_norm": 1.8321012258529663,
"learning_rate": 9.819420042307091e-06,
"loss": 1.6213,
"step": 36
},
{
"epoch": 0.12,
"grad_norm": 1.9112164974212646,
"learning_rate": 9.805354701993624e-06,
"loss": 1.6245,
"step": 37
},
{
"epoch": 0.12,
"grad_norm": 2.1160471439361572,
"learning_rate": 9.79077278351195e-06,
"loss": 1.6405,
"step": 38
},
{
"epoch": 0.12,
"grad_norm": 2.6318371295928955,
"learning_rate": 9.77567585442601e-06,
"loss": 1.6234,
"step": 39
},
{
"epoch": 0.13,
"grad_norm": 3.238373279571533,
"learning_rate": 9.76006553766365e-06,
"loss": 1.6452,
"step": 40
},
{
"epoch": 0.13,
"grad_norm": 2.056736469268799,
"learning_rate": 9.743943511342168e-06,
"loss": 1.679,
"step": 41
},
{
"epoch": 0.13,
"grad_norm": 2.000826358795166,
"learning_rate": 9.727311508587907e-06,
"loss": 1.6904,
"step": 42
},
{
"epoch": 0.14,
"grad_norm": 2.089482069015503,
"learning_rate": 9.710171317349946e-06,
"loss": 1.62,
"step": 43
},
{
"epoch": 0.14,
"grad_norm": 2.5748705863952637,
"learning_rate": 9.692524780207897e-06,
"loss": 1.6669,
"step": 44
},
{
"epoch": 0.14,
"grad_norm": 1.9039987325668335,
"learning_rate": 9.674373794173818e-06,
"loss": 1.6489,
"step": 45
},
{
"epoch": 0.15,
"grad_norm": 1.8047181367874146,
"learning_rate": 9.655720310488298e-06,
"loss": 1.6227,
"step": 46
},
{
"epoch": 0.15,
"grad_norm": 2.0552868843078613,
"learning_rate": 9.636566334410682e-06,
"loss": 1.5898,
"step": 47
},
{
"epoch": 0.15,
"grad_norm": 2.252218008041382,
"learning_rate": 9.616913925003514e-06,
"loss": 1.6667,
"step": 48
},
{
"epoch": 0.16,
"grad_norm": 2.015887498855591,
"learning_rate": 9.596765194911182e-06,
"loss": 1.6668,
"step": 49
},
{
"epoch": 0.16,
"grad_norm": 2.582007884979248,
"learning_rate": 9.576122310132814e-06,
"loss": 1.6542,
"step": 50
},
{
"epoch": 0.16,
"grad_norm": 1.9070011377334595,
"learning_rate": 9.554987489789426e-06,
"loss": 1.691,
"step": 51
},
{
"epoch": 0.17,
"grad_norm": 1.9671483039855957,
"learning_rate": 9.533363005885362e-06,
"loss": 1.6116,
"step": 52
},
{
"epoch": 0.17,
"grad_norm": 2.4808287620544434,
"learning_rate": 9.511251183064068e-06,
"loss": 1.6653,
"step": 53
},
{
"epoch": 0.17,
"grad_norm": 2.237518548965454,
"learning_rate": 9.48865439835817e-06,
"loss": 1.6571,
"step": 54
},
{
"epoch": 0.18,
"grad_norm": 2.142627477645874,
"learning_rate": 9.465575080933959e-06,
"loss": 1.6247,
"step": 55
},
{
"epoch": 0.18,
"grad_norm": 2.0365779399871826,
"learning_rate": 9.442015711830246e-06,
"loss": 1.6323,
"step": 56
},
{
"epoch": 0.18,
"grad_norm": 2.427443742752075,
"learning_rate": 9.417978823691652e-06,
"loss": 1.6646,
"step": 57
},
{
"epoch": 0.18,
"grad_norm": 2.014340877532959,
"learning_rate": 9.393467000496345e-06,
"loss": 1.5754,
"step": 58
},
{
"epoch": 0.19,
"grad_norm": 2.6110949516296387,
"learning_rate": 9.368482877278264e-06,
"loss": 1.657,
"step": 59
},
{
"epoch": 0.19,
"grad_norm": 2.048231363296509,
"learning_rate": 9.34302913984385e-06,
"loss": 1.61,
"step": 60
},
{
"epoch": 0.19,
"grad_norm": 2.672330379486084,
"learning_rate": 9.317108524483319e-06,
"loss": 1.5894,
"step": 61
},
{
"epoch": 0.2,
"grad_norm": 2.6745121479034424,
"learning_rate": 9.29072381767651e-06,
"loss": 1.5623,
"step": 62
},
{
"epoch": 0.2,
"grad_norm": 2.1490442752838135,
"learning_rate": 9.26387785579334e-06,
"loss": 1.6668,
"step": 63
},
{
"epoch": 0.2,
"grad_norm": 3.805147171020508,
"learning_rate": 9.236573524788888e-06,
"loss": 1.6274,
"step": 64
},
{
"epoch": 0.21,
"grad_norm": 2.243023157119751,
"learning_rate": 9.208813759893158e-06,
"loss": 1.6496,
"step": 65
},
{
"epoch": 0.21,
"grad_norm": 2.5260114669799805,
"learning_rate": 9.180601545295535e-06,
"loss": 1.623,
"step": 66
},
{
"epoch": 0.21,
"grad_norm": 2.0329606533050537,
"learning_rate": 9.151939913823988e-06,
"loss": 1.5873,
"step": 67
},
{
"epoch": 0.22,
"grad_norm": 3.3269598484039307,
"learning_rate": 9.122831946619038e-06,
"loss": 1.6327,
"step": 68
},
{
"epoch": 0.22,
"grad_norm": 2.1953978538513184,
"learning_rate": 9.093280772802527e-06,
"loss": 1.6321,
"step": 69
},
{
"epoch": 0.22,
"grad_norm": 3.2001328468322754,
"learning_rate": 9.063289569141251e-06,
"loss": 1.6338,
"step": 70
},
{
"epoch": 0.23,
"grad_norm": 2.076502799987793,
"learning_rate": 9.032861559705442e-06,
"loss": 1.6648,
"step": 71
},
{
"epoch": 0.23,
"grad_norm": 2.0800352096557617,
"learning_rate": 9.002000015522182e-06,
"loss": 1.5694,
"step": 72
},
{
"epoch": 0.23,
"grad_norm": 2.0901432037353516,
"learning_rate": 8.970708254223768e-06,
"loss": 1.6571,
"step": 73
},
{
"epoch": 0.24,
"grad_norm": 3.1389923095703125,
"learning_rate": 8.938989639691068e-06,
"loss": 1.6151,
"step": 74
},
{
"epoch": 0.24,
"grad_norm": 2.0514352321624756,
"learning_rate": 8.90684758169189e-06,
"loss": 1.618,
"step": 75
},
{
"epoch": 0.24,
"grad_norm": 1.9462100267410278,
"learning_rate": 8.87428553551445e-06,
"loss": 1.6367,
"step": 76
},
{
"epoch": 0.25,
"grad_norm": 2.2343838214874268,
"learning_rate": 8.841307001595904e-06,
"loss": 1.6321,
"step": 77
},
{
"epoch": 0.25,
"grad_norm": 2.6064207553863525,
"learning_rate": 8.807915525146065e-06,
"loss": 1.5913,
"step": 78
},
{
"epoch": 0.25,
"grad_norm": 2.589186191558838,
"learning_rate": 8.774114695766286e-06,
"loss": 1.6,
"step": 79
},
{
"epoch": 0.25,
"grad_norm": 2.0082924365997314,
"learning_rate": 8.739908147063576e-06,
"loss": 1.6266,
"step": 80
},
{
"epoch": 0.26,
"grad_norm": 1.9739817380905151,
"learning_rate": 8.705299556259986e-06,
"loss": 1.5926,
"step": 81
},
{
"epoch": 0.26,
"grad_norm": 2.9566433429718018,
"learning_rate": 8.670292643797302e-06,
"loss": 1.624,
"step": 82
},
{
"epoch": 0.26,
"grad_norm": 2.10595703125,
"learning_rate": 8.634891172937102e-06,
"loss": 1.6437,
"step": 83
},
{
"epoch": 0.27,
"grad_norm": 1.7080891132354736,
"learning_rate": 8.599098949356201e-06,
"loss": 1.6131,
"step": 84
},
{
"epoch": 0.27,
"grad_norm": 2.358856678009033,
"learning_rate": 8.562919820737537e-06,
"loss": 1.5748,
"step": 85
},
{
"epoch": 0.27,
"grad_norm": 2.030383825302124,
"learning_rate": 8.526357676356538e-06,
"loss": 1.6389,
"step": 86
},
{
"epoch": 0.28,
"grad_norm": 2.3567280769348145,
"learning_rate": 8.489416446663037e-06,
"loss": 1.6288,
"step": 87
},
{
"epoch": 0.28,
"grad_norm": 2.1183040142059326,
"learning_rate": 8.452100102858734e-06,
"loss": 1.5929,
"step": 88
},
{
"epoch": 0.28,
"grad_norm": 2.467944860458374,
"learning_rate": 8.414412656470297e-06,
"loss": 1.5579,
"step": 89
},
{
"epoch": 0.29,
"grad_norm": 2.200274705886841,
"learning_rate": 8.376358158918114e-06,
"loss": 1.587,
"step": 90
},
{
"epoch": 0.29,
"grad_norm": 2.3279902935028076,
"learning_rate": 8.33794070108077e-06,
"loss": 1.6496,
"step": 91
},
{
"epoch": 0.29,
"grad_norm": 2.7403910160064697,
"learning_rate": 8.299164412855268e-06,
"loss": 1.5665,
"step": 92
},
{
"epoch": 0.3,
"grad_norm": 2.5097343921661377,
"learning_rate": 8.260033462713073e-06,
"loss": 1.6245,
"step": 93
},
{
"epoch": 0.3,
"grad_norm": 2.2657604217529297,
"learning_rate": 8.22055205725199e-06,
"loss": 1.5785,
"step": 94
},
{
"epoch": 0.3,
"grad_norm": 2.4038476943969727,
"learning_rate": 8.180724440743957e-06,
"loss": 1.6912,
"step": 95
},
{
"epoch": 0.31,
"grad_norm": 2.811049461364746,
"learning_rate": 8.14055489467878e-06,
"loss": 1.6059,
"step": 96
},
{
"epoch": 0.31,
"grad_norm": 2.474153518676758,
"learning_rate": 8.100047737303877e-06,
"loss": 1.6423,
"step": 97
},
{
"epoch": 0.31,
"grad_norm": 2.351736307144165,
"learning_rate": 8.059207323160057e-06,
"loss": 1.6199,
"step": 98
},
{
"epoch": 0.32,
"grad_norm": 2.3526997566223145,
"learning_rate": 8.018038042613407e-06,
"loss": 1.5856,
"step": 99
},
{
"epoch": 0.32,
"grad_norm": 2.075406789779663,
"learning_rate": 7.97654432138333e-06,
"loss": 1.5901,
"step": 100
},
{
"epoch": 0.32,
"grad_norm": 2.6973438262939453,
"learning_rate": 7.93473062006677e-06,
"loss": 1.5696,
"step": 101
},
{
"epoch": 0.32,
"grad_norm": 2.6449167728424072,
"learning_rate": 7.892601433658705e-06,
"loss": 1.5939,
"step": 102
},
{
"epoch": 0.33,
"grad_norm": 2.360849380493164,
"learning_rate": 7.850161291068915e-06,
"loss": 1.5449,
"step": 103
},
{
"epoch": 0.33,
"grad_norm": 2.467226028442383,
"learning_rate": 7.807414754635145e-06,
"loss": 1.5926,
"step": 104
},
{
"epoch": 0.33,
"grad_norm": 2.449989080429077,
"learning_rate": 7.764366419632636e-06,
"loss": 1.5591,
"step": 105
},
{
"epoch": 0.34,
"grad_norm": 5.518803119659424,
"learning_rate": 7.721020913780137e-06,
"loss": 1.5406,
"step": 106
},
{
"epoch": 0.34,
"grad_norm": 2.6841485500335693,
"learning_rate": 7.677382896742417e-06,
"loss": 1.5797,
"step": 107
},
{
"epoch": 0.34,
"grad_norm": 2.475085496902466,
"learning_rate": 7.63345705962935e-06,
"loss": 1.5653,
"step": 108
},
{
"epoch": 0.35,
"grad_norm": 2.3905832767486572,
"learning_rate": 7.589248124491627e-06,
"loss": 1.5641,
"step": 109
},
{
"epoch": 0.35,
"grad_norm": 2.7013638019561768,
"learning_rate": 7.544760843813122e-06,
"loss": 1.5837,
"step": 110
},
{
"epoch": 0.35,
"grad_norm": 2.500148296356201,
"learning_rate": 7.500000000000001e-06,
"loss": 1.6255,
"step": 111
},
{
"epoch": 0.36,
"grad_norm": 2.106887102127075,
"learning_rate": 7.454970404866612e-06,
"loss": 1.6069,
"step": 112
},
{
"epoch": 0.36,
"grad_norm": 2.1840200424194336,
"learning_rate": 7.409676899118213e-06,
"loss": 1.5959,
"step": 113
},
{
"epoch": 0.36,
"grad_norm": 5.758022308349609,
"learning_rate": 7.3641243518305915e-06,
"loss": 1.616,
"step": 114
},
{
"epoch": 0.37,
"grad_norm": 2.24625301361084,
"learning_rate": 7.318317659926637e-06,
"loss": 1.639,
"step": 115
},
{
"epoch": 0.37,
"grad_norm": 3.8651440143585205,
"learning_rate": 7.272261747649922e-06,
"loss": 1.6287,
"step": 116
},
{
"epoch": 0.37,
"grad_norm": 1.9262839555740356,
"learning_rate": 7.225961566035335e-06,
"loss": 1.5851,
"step": 117
},
{
"epoch": 0.38,
"grad_norm": 2.4312078952789307,
"learning_rate": 7.179422092376856e-06,
"loss": 1.5934,
"step": 118
},
{
"epoch": 0.38,
"grad_norm": 2.391693592071533,
"learning_rate": 7.132648329692478e-06,
"loss": 1.5719,
"step": 119
},
{
"epoch": 0.38,
"grad_norm": 2.10438871383667,
"learning_rate": 7.085645306186391e-06,
"loss": 1.5876,
"step": 120
},
{
"epoch": 0.39,
"grad_norm": 1.984710931777954,
"learning_rate": 7.038418074708444e-06,
"loss": 1.6506,
"step": 121
},
{
"epoch": 0.39,
"grad_norm": 2.4563817977905273,
"learning_rate": 6.990971712210966e-06,
"loss": 1.5578,
"step": 122
},
{
"epoch": 0.39,
"grad_norm": 1.856941819190979,
"learning_rate": 6.943311319202976e-06,
"loss": 1.5806,
"step": 123
},
{
"epoch": 0.4,
"grad_norm": 2.228983163833618,
"learning_rate": 6.895442019201898e-06,
"loss": 1.5639,
"step": 124
},
{
"epoch": 0.4,
"grad_norm": 2.9688351154327393,
"learning_rate": 6.8473689581827585e-06,
"loss": 1.5939,
"step": 125
},
{
"epoch": 0.4,
"grad_norm": 2.028876304626465,
"learning_rate": 6.7990973040250055e-06,
"loss": 1.6096,
"step": 126
},
{
"epoch": 0.4,
"grad_norm": 2.385794162750244,
"learning_rate": 6.750632245956954e-06,
"loss": 1.5617,
"step": 127
},
{
"epoch": 0.41,
"grad_norm": 2.0037996768951416,
"learning_rate": 6.701978993997942e-06,
"loss": 1.6028,
"step": 128
},
{
"epoch": 0.41,
"grad_norm": 2.0149664878845215,
"learning_rate": 6.653142778398247e-06,
"loss": 1.583,
"step": 129
},
{
"epoch": 0.41,
"grad_norm": 2.466585874557495,
"learning_rate": 6.6041288490768385e-06,
"loss": 1.6368,
"step": 130
},
{
"epoch": 0.42,
"grad_norm": 2.181319236755371,
"learning_rate": 6.554942475057003e-06,
"loss": 1.5819,
"step": 131
},
{
"epoch": 0.42,
"grad_norm": 1.9432387351989746,
"learning_rate": 6.505588943899923e-06,
"loss": 1.5551,
"step": 132
},
{
"epoch": 0.42,
"grad_norm": 2.277068614959717,
"learning_rate": 6.456073561136261e-06,
"loss": 1.5788,
"step": 133
},
{
"epoch": 0.43,
"grad_norm": 1.9018107652664185,
"learning_rate": 6.406401649695814e-06,
"loss": 1.5639,
"step": 134
},
{
"epoch": 0.43,
"grad_norm": 2.755958318710327,
"learning_rate": 6.356578549335295e-06,
"loss": 1.6015,
"step": 135
},
{
"epoch": 0.43,
"grad_norm": 2.2761142253875732,
"learning_rate": 6.306609616064304e-06,
"loss": 1.6054,
"step": 136
},
{
"epoch": 0.44,
"grad_norm": 3.212486505508423,
"learning_rate": 6.256500221569556e-06,
"loss": 1.5953,
"step": 137
},
{
"epoch": 0.44,
"grad_norm": 2.2175703048706055,
"learning_rate": 6.2062557526374226e-06,
"loss": 1.5745,
"step": 138
},
{
"epoch": 0.44,
"grad_norm": 1.9958215951919556,
"learning_rate": 6.15588161057485e-06,
"loss": 1.5981,
"step": 139
},
{
"epoch": 0.45,
"grad_norm": 2.4180896282196045,
"learning_rate": 6.10538321062871e-06,
"loss": 1.5836,
"step": 140
},
{
"epoch": 0.45,
"grad_norm": 2.0187716484069824,
"learning_rate": 6.0547659814036664e-06,
"loss": 1.6062,
"step": 141
},
{
"epoch": 0.45,
"grad_norm": 2.3086986541748047,
"learning_rate": 6.004035364278593e-06,
"loss": 1.5764,
"step": 142
},
{
"epoch": 0.46,
"grad_norm": 1.9949414730072021,
"learning_rate": 5.953196812821622e-06,
"loss": 1.6037,
"step": 143
},
{
"epoch": 0.46,
"grad_norm": 1.8632932901382446,
"learning_rate": 5.902255792203882e-06,
"loss": 1.6051,
"step": 144
},
{
"epoch": 0.46,
"grad_norm": 2.034745931625366,
"learning_rate": 5.851217778611994e-06,
"loss": 1.6386,
"step": 145
},
{
"epoch": 0.47,
"grad_norm": 2.1993603706359863,
"learning_rate": 5.800088258659371e-06,
"loss": 1.6081,
"step": 146
},
{
"epoch": 0.47,
"grad_norm": 2.0585718154907227,
"learning_rate": 5.748872728796409e-06,
"loss": 1.5955,
"step": 147
},
{
"epoch": 0.47,
"grad_norm": 2.2753782272338867,
"learning_rate": 5.697576694719616e-06,
"loss": 1.5655,
"step": 148
},
{
"epoch": 0.47,
"grad_norm": 3.237483024597168,
"learning_rate": 5.646205670779745e-06,
"loss": 1.6004,
"step": 149
},
{
"epoch": 0.48,
"grad_norm": 3.019814968109131,
"learning_rate": 5.594765179389003e-06,
"loss": 1.5721,
"step": 150
}
],
"logging_steps": 1.0,
"max_steps": 313,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 50,
"total_flos": 164924379693056.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}