eeve-2.8B_B / trainer_state.json
c10's picture
Model save
1a8a9b9 verified
raw
history blame contribute delete
No virus
51.5 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 291,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.003436426116838488,
"grad_norm": 6.773601055145264,
"learning_rate": 3.3333333333333335e-07,
"loss": 2.3294,
"step": 1
},
{
"epoch": 0.006872852233676976,
"grad_norm": 7.0779709815979,
"learning_rate": 6.666666666666667e-07,
"loss": 2.343,
"step": 2
},
{
"epoch": 0.010309278350515464,
"grad_norm": 6.805663108825684,
"learning_rate": 1.0000000000000002e-06,
"loss": 2.3024,
"step": 3
},
{
"epoch": 0.013745704467353952,
"grad_norm": 6.982776641845703,
"learning_rate": 1.3333333333333334e-06,
"loss": 2.3704,
"step": 4
},
{
"epoch": 0.01718213058419244,
"grad_norm": 6.750400543212891,
"learning_rate": 1.6666666666666667e-06,
"loss": 2.3224,
"step": 5
},
{
"epoch": 0.020618556701030927,
"grad_norm": 6.902153968811035,
"learning_rate": 2.0000000000000003e-06,
"loss": 2.3783,
"step": 6
},
{
"epoch": 0.024054982817869417,
"grad_norm": 6.6329731941223145,
"learning_rate": 2.3333333333333336e-06,
"loss": 2.2336,
"step": 7
},
{
"epoch": 0.027491408934707903,
"grad_norm": 6.543385982513428,
"learning_rate": 2.666666666666667e-06,
"loss": 2.2793,
"step": 8
},
{
"epoch": 0.030927835051546393,
"grad_norm": 5.860514163970947,
"learning_rate": 3e-06,
"loss": 2.1503,
"step": 9
},
{
"epoch": 0.03436426116838488,
"grad_norm": 5.754838466644287,
"learning_rate": 3.3333333333333333e-06,
"loss": 2.0999,
"step": 10
},
{
"epoch": 0.037800687285223365,
"grad_norm": 5.671762466430664,
"learning_rate": 3.6666666666666666e-06,
"loss": 2.101,
"step": 11
},
{
"epoch": 0.041237113402061855,
"grad_norm": 4.444242477416992,
"learning_rate": 4.000000000000001e-06,
"loss": 1.7642,
"step": 12
},
{
"epoch": 0.044673539518900345,
"grad_norm": 4.336024284362793,
"learning_rate": 4.333333333333334e-06,
"loss": 1.6966,
"step": 13
},
{
"epoch": 0.048109965635738834,
"grad_norm": 4.255687713623047,
"learning_rate": 4.666666666666667e-06,
"loss": 1.6585,
"step": 14
},
{
"epoch": 0.05154639175257732,
"grad_norm": 3.6368279457092285,
"learning_rate": 5e-06,
"loss": 1.3429,
"step": 15
},
{
"epoch": 0.054982817869415807,
"grad_norm": 3.6595640182495117,
"learning_rate": 5.333333333333334e-06,
"loss": 1.1194,
"step": 16
},
{
"epoch": 0.058419243986254296,
"grad_norm": 3.600109338760376,
"learning_rate": 5.666666666666667e-06,
"loss": 1.05,
"step": 17
},
{
"epoch": 0.061855670103092786,
"grad_norm": 3.134242296218872,
"learning_rate": 6e-06,
"loss": 0.9555,
"step": 18
},
{
"epoch": 0.06529209621993128,
"grad_norm": 2.7783243656158447,
"learning_rate": 6.333333333333333e-06,
"loss": 0.8856,
"step": 19
},
{
"epoch": 0.06872852233676977,
"grad_norm": 2.519134283065796,
"learning_rate": 6.666666666666667e-06,
"loss": 0.8207,
"step": 20
},
{
"epoch": 0.07216494845360824,
"grad_norm": 2.0445821285247803,
"learning_rate": 7e-06,
"loss": 0.4981,
"step": 21
},
{
"epoch": 0.07560137457044673,
"grad_norm": 1.3868145942687988,
"learning_rate": 7.333333333333333e-06,
"loss": 0.4683,
"step": 22
},
{
"epoch": 0.07903780068728522,
"grad_norm": 1.092483639717102,
"learning_rate": 7.666666666666667e-06,
"loss": 0.5258,
"step": 23
},
{
"epoch": 0.08247422680412371,
"grad_norm": 0.9057236313819885,
"learning_rate": 8.000000000000001e-06,
"loss": 0.4232,
"step": 24
},
{
"epoch": 0.0859106529209622,
"grad_norm": 0.7798469662666321,
"learning_rate": 8.333333333333334e-06,
"loss": 0.4377,
"step": 25
},
{
"epoch": 0.08934707903780069,
"grad_norm": 0.7577869892120361,
"learning_rate": 8.666666666666668e-06,
"loss": 0.3982,
"step": 26
},
{
"epoch": 0.09278350515463918,
"grad_norm": 0.7064369320869446,
"learning_rate": 9e-06,
"loss": 0.4299,
"step": 27
},
{
"epoch": 0.09621993127147767,
"grad_norm": 0.6497183442115784,
"learning_rate": 9.333333333333334e-06,
"loss": 0.4115,
"step": 28
},
{
"epoch": 0.09965635738831616,
"grad_norm": 0.6399904489517212,
"learning_rate": 9.666666666666667e-06,
"loss": 0.4078,
"step": 29
},
{
"epoch": 0.10309278350515463,
"grad_norm": 0.6032602190971375,
"learning_rate": 1e-05,
"loss": 0.3716,
"step": 30
},
{
"epoch": 0.10652920962199312,
"grad_norm": 0.6270999908447266,
"learning_rate": 9.999637795788383e-06,
"loss": 0.3949,
"step": 31
},
{
"epoch": 0.10996563573883161,
"grad_norm": 0.5787239074707031,
"learning_rate": 9.99855123563029e-06,
"loss": 0.3646,
"step": 32
},
{
"epoch": 0.1134020618556701,
"grad_norm": 0.5683437585830688,
"learning_rate": 9.996740476948386e-06,
"loss": 0.3752,
"step": 33
},
{
"epoch": 0.11683848797250859,
"grad_norm": 0.5568259954452515,
"learning_rate": 9.994205782088438e-06,
"loss": 0.3973,
"step": 34
},
{
"epoch": 0.12027491408934708,
"grad_norm": 0.5675911903381348,
"learning_rate": 9.990947518281312e-06,
"loss": 0.3204,
"step": 35
},
{
"epoch": 0.12371134020618557,
"grad_norm": 0.5186944603919983,
"learning_rate": 9.986966157589751e-06,
"loss": 0.3472,
"step": 36
},
{
"epoch": 0.12714776632302405,
"grad_norm": 0.5074464082717896,
"learning_rate": 9.982262276840002e-06,
"loss": 0.3487,
"step": 37
},
{
"epoch": 0.13058419243986255,
"grad_norm": 0.5508702397346497,
"learning_rate": 9.976836557538234e-06,
"loss": 0.3412,
"step": 38
},
{
"epoch": 0.13402061855670103,
"grad_norm": 0.513957679271698,
"learning_rate": 9.970689785771798e-06,
"loss": 0.368,
"step": 39
},
{
"epoch": 0.13745704467353953,
"grad_norm": 0.4993857145309448,
"learning_rate": 9.963822852095344e-06,
"loss": 0.365,
"step": 40
},
{
"epoch": 0.140893470790378,
"grad_norm": 0.503234326839447,
"learning_rate": 9.95623675140179e-06,
"loss": 0.3657,
"step": 41
},
{
"epoch": 0.14432989690721648,
"grad_norm": 0.47215989232063293,
"learning_rate": 9.947932582778188e-06,
"loss": 0.354,
"step": 42
},
{
"epoch": 0.14776632302405499,
"grad_norm": 0.4891766607761383,
"learning_rate": 9.938911549346473e-06,
"loss": 0.3539,
"step": 43
},
{
"epoch": 0.15120274914089346,
"grad_norm": 0.5077570080757141,
"learning_rate": 9.929174958089167e-06,
"loss": 0.36,
"step": 44
},
{
"epoch": 0.15463917525773196,
"grad_norm": 0.4428570866584778,
"learning_rate": 9.918724219660013e-06,
"loss": 0.3325,
"step": 45
},
{
"epoch": 0.15807560137457044,
"grad_norm": 0.46060407161712646,
"learning_rate": 9.907560848179607e-06,
"loss": 0.3391,
"step": 46
},
{
"epoch": 0.16151202749140894,
"grad_norm": 0.46653836965560913,
"learning_rate": 9.895686461016007e-06,
"loss": 0.3469,
"step": 47
},
{
"epoch": 0.16494845360824742,
"grad_norm": 0.6124569177627563,
"learning_rate": 9.883102778550434e-06,
"loss": 0.3625,
"step": 48
},
{
"epoch": 0.16838487972508592,
"grad_norm": 0.47772711515426636,
"learning_rate": 9.869811623928001e-06,
"loss": 0.3559,
"step": 49
},
{
"epoch": 0.1718213058419244,
"grad_norm": 0.4554615914821625,
"learning_rate": 9.855814922793583e-06,
"loss": 0.3236,
"step": 50
},
{
"epoch": 0.17525773195876287,
"grad_norm": 0.4689699113368988,
"learning_rate": 9.841114703012817e-06,
"loss": 0.3441,
"step": 51
},
{
"epoch": 0.17869415807560138,
"grad_norm": 0.4912395477294922,
"learning_rate": 9.82571309437831e-06,
"loss": 0.3505,
"step": 52
},
{
"epoch": 0.18213058419243985,
"grad_norm": 0.48207777738571167,
"learning_rate": 9.809612328301071e-06,
"loss": 0.3444,
"step": 53
},
{
"epoch": 0.18556701030927836,
"grad_norm": 0.4854188859462738,
"learning_rate": 9.792814737487207e-06,
"loss": 0.3574,
"step": 54
},
{
"epoch": 0.18900343642611683,
"grad_norm": 0.4824926257133484,
"learning_rate": 9.775322755599979e-06,
"loss": 0.3577,
"step": 55
},
{
"epoch": 0.19243986254295534,
"grad_norm": 0.467854768037796,
"learning_rate": 9.757138916907184e-06,
"loss": 0.3144,
"step": 56
},
{
"epoch": 0.1958762886597938,
"grad_norm": 0.4348817765712738,
"learning_rate": 9.738265855914014e-06,
"loss": 0.3057,
"step": 57
},
{
"epoch": 0.19931271477663232,
"grad_norm": 0.48006704449653625,
"learning_rate": 9.718706306981332e-06,
"loss": 0.3294,
"step": 58
},
{
"epoch": 0.2027491408934708,
"grad_norm": 0.4489571750164032,
"learning_rate": 9.698463103929542e-06,
"loss": 0.3263,
"step": 59
},
{
"epoch": 0.20618556701030927,
"grad_norm": 0.42884257435798645,
"learning_rate": 9.677539179628005e-06,
"loss": 0.3015,
"step": 60
},
{
"epoch": 0.20962199312714777,
"grad_norm": 0.4882860779762268,
"learning_rate": 9.655937565570124e-06,
"loss": 0.3531,
"step": 61
},
{
"epoch": 0.21305841924398625,
"grad_norm": 0.45341619849205017,
"learning_rate": 9.63366139143413e-06,
"loss": 0.3245,
"step": 62
},
{
"epoch": 0.21649484536082475,
"grad_norm": 0.47687625885009766,
"learning_rate": 9.610713884629667e-06,
"loss": 0.3019,
"step": 63
},
{
"epoch": 0.21993127147766323,
"grad_norm": 0.4295818507671356,
"learning_rate": 9.587098369830171e-06,
"loss": 0.2952,
"step": 64
},
{
"epoch": 0.22336769759450173,
"grad_norm": 0.4635133743286133,
"learning_rate": 9.562818268491216e-06,
"loss": 0.3243,
"step": 65
},
{
"epoch": 0.2268041237113402,
"grad_norm": 0.41768181324005127,
"learning_rate": 9.537877098354787e-06,
"loss": 0.2814,
"step": 66
},
{
"epoch": 0.23024054982817868,
"grad_norm": 0.47494152188301086,
"learning_rate": 9.512278472939627e-06,
"loss": 0.3122,
"step": 67
},
{
"epoch": 0.23367697594501718,
"grad_norm": 0.48896145820617676,
"learning_rate": 9.486026101017711e-06,
"loss": 0.3254,
"step": 68
},
{
"epoch": 0.23711340206185566,
"grad_norm": 0.444621741771698,
"learning_rate": 9.459123786076911e-06,
"loss": 0.3092,
"step": 69
},
{
"epoch": 0.24054982817869416,
"grad_norm": 0.49557366967201233,
"learning_rate": 9.431575425769938e-06,
"loss": 0.3571,
"step": 70
},
{
"epoch": 0.24398625429553264,
"grad_norm": 0.4832090139389038,
"learning_rate": 9.40338501134964e-06,
"loss": 0.3053,
"step": 71
},
{
"epoch": 0.24742268041237114,
"grad_norm": 0.49034735560417175,
"learning_rate": 9.374556627090749e-06,
"loss": 0.3264,
"step": 72
},
{
"epoch": 0.2508591065292096,
"grad_norm": 0.4763636589050293,
"learning_rate": 9.345094449698143e-06,
"loss": 0.3601,
"step": 73
},
{
"epoch": 0.2542955326460481,
"grad_norm": 0.4429002106189728,
"learning_rate": 9.315002747701716e-06,
"loss": 0.2926,
"step": 74
},
{
"epoch": 0.25773195876288657,
"grad_norm": 0.44965827465057373,
"learning_rate": 9.284285880837947e-06,
"loss": 0.3413,
"step": 75
},
{
"epoch": 0.2611683848797251,
"grad_norm": 0.43706703186035156,
"learning_rate": 9.252948299418255e-06,
"loss": 0.3132,
"step": 76
},
{
"epoch": 0.2646048109965636,
"grad_norm": 0.446598082780838,
"learning_rate": 9.220994543684225e-06,
"loss": 0.3136,
"step": 77
},
{
"epoch": 0.26804123711340205,
"grad_norm": 0.4524715542793274,
"learning_rate": 9.188429243149824e-06,
"loss": 0.3436,
"step": 78
},
{
"epoch": 0.27147766323024053,
"grad_norm": 0.46307334303855896,
"learning_rate": 9.155257115930651e-06,
"loss": 0.3319,
"step": 79
},
{
"epoch": 0.27491408934707906,
"grad_norm": 0.41813838481903076,
"learning_rate": 9.121482968060384e-06,
"loss": 0.2707,
"step": 80
},
{
"epoch": 0.27835051546391754,
"grad_norm": 0.461912602186203,
"learning_rate": 9.08711169279446e-06,
"loss": 0.3477,
"step": 81
},
{
"epoch": 0.281786941580756,
"grad_norm": 0.4281875789165497,
"learning_rate": 9.052148269901145e-06,
"loss": 0.2952,
"step": 82
},
{
"epoch": 0.2852233676975945,
"grad_norm": 0.44423964619636536,
"learning_rate": 9.01659776494005e-06,
"loss": 0.3284,
"step": 83
},
{
"epoch": 0.28865979381443296,
"grad_norm": 0.4334423243999481,
"learning_rate": 8.98046532852822e-06,
"loss": 0.3024,
"step": 84
},
{
"epoch": 0.2920962199312715,
"grad_norm": 0.4503464698791504,
"learning_rate": 8.943756195593916e-06,
"loss": 0.3053,
"step": 85
},
{
"epoch": 0.29553264604810997,
"grad_norm": 0.45096880197525024,
"learning_rate": 8.90647568461816e-06,
"loss": 0.2971,
"step": 86
},
{
"epoch": 0.29896907216494845,
"grad_norm": 0.4469160735607147,
"learning_rate": 8.868629196864182e-06,
"loss": 0.3323,
"step": 87
},
{
"epoch": 0.3024054982817869,
"grad_norm": 0.45453527569770813,
"learning_rate": 8.83022221559489e-06,
"loss": 0.3152,
"step": 88
},
{
"epoch": 0.30584192439862545,
"grad_norm": 0.42542845010757446,
"learning_rate": 8.791260305278434e-06,
"loss": 0.2697,
"step": 89
},
{
"epoch": 0.30927835051546393,
"grad_norm": 0.3987956643104553,
"learning_rate": 8.751749110782013e-06,
"loss": 0.2691,
"step": 90
},
{
"epoch": 0.3127147766323024,
"grad_norm": 0.43505221605300903,
"learning_rate": 8.71169435655405e-06,
"loss": 0.3071,
"step": 91
},
{
"epoch": 0.3161512027491409,
"grad_norm": 0.42406681180000305,
"learning_rate": 8.671101845794816e-06,
"loss": 0.3137,
"step": 92
},
{
"epoch": 0.31958762886597936,
"grad_norm": 0.4488978981971741,
"learning_rate": 8.629977459615655e-06,
"loss": 0.3211,
"step": 93
},
{
"epoch": 0.3230240549828179,
"grad_norm": 0.4613654613494873,
"learning_rate": 8.588327156186915e-06,
"loss": 0.3398,
"step": 94
},
{
"epoch": 0.32646048109965636,
"grad_norm": 0.40654006600379944,
"learning_rate": 8.546156969874723e-06,
"loss": 0.3183,
"step": 95
},
{
"epoch": 0.32989690721649484,
"grad_norm": 0.45063990354537964,
"learning_rate": 8.503473010366713e-06,
"loss": 0.3039,
"step": 96
},
{
"epoch": 0.3333333333333333,
"grad_norm": 0.4447152316570282,
"learning_rate": 8.460281461786848e-06,
"loss": 0.283,
"step": 97
},
{
"epoch": 0.33676975945017185,
"grad_norm": 0.4078894555568695,
"learning_rate": 8.416588581799447e-06,
"loss": 0.2948,
"step": 98
},
{
"epoch": 0.3402061855670103,
"grad_norm": 0.42324382066726685,
"learning_rate": 8.372400700702569e-06,
"loss": 0.2622,
"step": 99
},
{
"epoch": 0.3436426116838488,
"grad_norm": 0.46570709347724915,
"learning_rate": 8.327724220510873e-06,
"loss": 0.3214,
"step": 100
},
{
"epoch": 0.3470790378006873,
"grad_norm": 0.46272405982017517,
"learning_rate": 8.282565614028068e-06,
"loss": 0.3264,
"step": 101
},
{
"epoch": 0.35051546391752575,
"grad_norm": 0.4732251763343811,
"learning_rate": 8.23693142390914e-06,
"loss": 0.3502,
"step": 102
},
{
"epoch": 0.3539518900343643,
"grad_norm": 0.4501849114894867,
"learning_rate": 8.19082826171243e-06,
"loss": 0.3211,
"step": 103
},
{
"epoch": 0.35738831615120276,
"grad_norm": 0.44297897815704346,
"learning_rate": 8.144262806941743e-06,
"loss": 0.3162,
"step": 104
},
{
"epoch": 0.36082474226804123,
"grad_norm": 0.4206938147544861,
"learning_rate": 8.097241806078616e-06,
"loss": 0.3029,
"step": 105
},
{
"epoch": 0.3642611683848797,
"grad_norm": 0.4197562634944916,
"learning_rate": 8.049772071604864e-06,
"loss": 0.2688,
"step": 106
},
{
"epoch": 0.36769759450171824,
"grad_norm": 0.44810447096824646,
"learning_rate": 8.001860481015594e-06,
"loss": 0.3005,
"step": 107
},
{
"epoch": 0.3711340206185567,
"grad_norm": 0.4516235291957855,
"learning_rate": 7.953513975822755e-06,
"loss": 0.2865,
"step": 108
},
{
"epoch": 0.3745704467353952,
"grad_norm": 0.4506089985370636,
"learning_rate": 7.904739560549475e-06,
"loss": 0.3204,
"step": 109
},
{
"epoch": 0.37800687285223367,
"grad_norm": 0.45452532172203064,
"learning_rate": 7.855544301715203e-06,
"loss": 0.2961,
"step": 110
},
{
"epoch": 0.38144329896907214,
"grad_norm": 0.5450659990310669,
"learning_rate": 7.805935326811913e-06,
"loss": 0.3216,
"step": 111
},
{
"epoch": 0.3848797250859107,
"grad_norm": 0.4857536554336548,
"learning_rate": 7.755919823271466e-06,
"loss": 0.3348,
"step": 112
},
{
"epoch": 0.38831615120274915,
"grad_norm": 0.4399779438972473,
"learning_rate": 7.70550503742427e-06,
"loss": 0.2682,
"step": 113
},
{
"epoch": 0.3917525773195876,
"grad_norm": 0.4279418885707855,
"learning_rate": 7.654698273449435e-06,
"loss": 0.2986,
"step": 114
},
{
"epoch": 0.3951890034364261,
"grad_norm": 0.45266300439834595,
"learning_rate": 7.603506892316513e-06,
"loss": 0.3116,
"step": 115
},
{
"epoch": 0.39862542955326463,
"grad_norm": 0.4144250750541687,
"learning_rate": 7.551938310719043e-06,
"loss": 0.2864,
"step": 116
},
{
"epoch": 0.4020618556701031,
"grad_norm": 0.5105399489402771,
"learning_rate": 7.500000000000001e-06,
"loss": 0.3088,
"step": 117
},
{
"epoch": 0.4054982817869416,
"grad_norm": 0.4652271568775177,
"learning_rate": 7.447699485069342e-06,
"loss": 0.2821,
"step": 118
},
{
"epoch": 0.40893470790378006,
"grad_norm": 0.4278913140296936,
"learning_rate": 7.395044343313777e-06,
"loss": 0.2981,
"step": 119
},
{
"epoch": 0.41237113402061853,
"grad_norm": 0.41499921679496765,
"learning_rate": 7.342042203498952e-06,
"loss": 0.283,
"step": 120
},
{
"epoch": 0.41580756013745707,
"grad_norm": 0.4199381172657013,
"learning_rate": 7.288700744664167e-06,
"loss": 0.2977,
"step": 121
},
{
"epoch": 0.41924398625429554,
"grad_norm": 0.4549289047718048,
"learning_rate": 7.235027695009846e-06,
"loss": 0.329,
"step": 122
},
{
"epoch": 0.422680412371134,
"grad_norm": 0.43108847737312317,
"learning_rate": 7.181030830777838e-06,
"loss": 0.2919,
"step": 123
},
{
"epoch": 0.4261168384879725,
"grad_norm": 0.42637932300567627,
"learning_rate": 7.1267179751248005e-06,
"loss": 0.304,
"step": 124
},
{
"epoch": 0.42955326460481097,
"grad_norm": 0.4469113349914551,
"learning_rate": 7.0720969969887595e-06,
"loss": 0.2966,
"step": 125
},
{
"epoch": 0.4329896907216495,
"grad_norm": 0.41228166222572327,
"learning_rate": 7.017175809949044e-06,
"loss": 0.2874,
"step": 126
},
{
"epoch": 0.436426116838488,
"grad_norm": 0.45446354150772095,
"learning_rate": 6.961962371079752e-06,
"loss": 0.3129,
"step": 127
},
{
"epoch": 0.43986254295532645,
"grad_norm": 0.4572490155696869,
"learning_rate": 6.906464679796927e-06,
"loss": 0.3398,
"step": 128
},
{
"epoch": 0.44329896907216493,
"grad_norm": 0.4100750982761383,
"learning_rate": 6.850690776699574e-06,
"loss": 0.2885,
"step": 129
},
{
"epoch": 0.44673539518900346,
"grad_norm": 0.4656309187412262,
"learning_rate": 6.79464874240473e-06,
"loss": 0.3347,
"step": 130
},
{
"epoch": 0.45017182130584193,
"grad_norm": 0.4740864038467407,
"learning_rate": 6.7383466963767386e-06,
"loss": 0.3141,
"step": 131
},
{
"epoch": 0.4536082474226804,
"grad_norm": 0.4347783029079437,
"learning_rate": 6.681792795750876e-06,
"loss": 0.2836,
"step": 132
},
{
"epoch": 0.4570446735395189,
"grad_norm": 0.46490252017974854,
"learning_rate": 6.624995234151539e-06,
"loss": 0.3036,
"step": 133
},
{
"epoch": 0.46048109965635736,
"grad_norm": 0.4635082185268402,
"learning_rate": 6.567962240505136e-06,
"loss": 0.2989,
"step": 134
},
{
"epoch": 0.4639175257731959,
"grad_norm": 0.43528881669044495,
"learning_rate": 6.510702077847864e-06,
"loss": 0.3137,
"step": 135
},
{
"epoch": 0.46735395189003437,
"grad_norm": 0.4689432978630066,
"learning_rate": 6.453223042128556e-06,
"loss": 0.2838,
"step": 136
},
{
"epoch": 0.47079037800687284,
"grad_norm": 0.45231714844703674,
"learning_rate": 6.395533461006736e-06,
"loss": 0.2994,
"step": 137
},
{
"epoch": 0.4742268041237113,
"grad_norm": 0.507282018661499,
"learning_rate": 6.337641692646106e-06,
"loss": 0.32,
"step": 138
},
{
"epoch": 0.47766323024054985,
"grad_norm": 0.47324004769325256,
"learning_rate": 6.2795561245035895e-06,
"loss": 0.3178,
"step": 139
},
{
"epoch": 0.48109965635738833,
"grad_norm": 0.4751383364200592,
"learning_rate": 6.221285172114156e-06,
"loss": 0.3112,
"step": 140
},
{
"epoch": 0.4845360824742268,
"grad_norm": 0.47212111949920654,
"learning_rate": 6.162837277871553e-06,
"loss": 0.2964,
"step": 141
},
{
"epoch": 0.4879725085910653,
"grad_norm": 0.48457249999046326,
"learning_rate": 6.104220909805162e-06,
"loss": 0.2869,
"step": 142
},
{
"epoch": 0.49140893470790376,
"grad_norm": 0.4023948609828949,
"learning_rate": 6.045444560353136e-06,
"loss": 0.2437,
"step": 143
},
{
"epoch": 0.4948453608247423,
"grad_norm": 0.433889776468277,
"learning_rate": 5.986516745132e-06,
"loss": 0.2714,
"step": 144
},
{
"epoch": 0.49828178694158076,
"grad_norm": 0.47038504481315613,
"learning_rate": 5.927446001702899e-06,
"loss": 0.3033,
"step": 145
},
{
"epoch": 0.5017182130584192,
"grad_norm": 0.4430581331253052,
"learning_rate": 5.8682408883346535e-06,
"loss": 0.2951,
"step": 146
},
{
"epoch": 0.5051546391752577,
"grad_norm": 0.4938088357448578,
"learning_rate": 5.808909982763825e-06,
"loss": 0.3056,
"step": 147
},
{
"epoch": 0.5085910652920962,
"grad_norm": 0.4448769688606262,
"learning_rate": 5.749461880951966e-06,
"loss": 0.3015,
"step": 148
},
{
"epoch": 0.5120274914089347,
"grad_norm": 0.44447487592697144,
"learning_rate": 5.689905195840216e-06,
"loss": 0.2846,
"step": 149
},
{
"epoch": 0.5154639175257731,
"grad_norm": 0.411819189786911,
"learning_rate": 5.630248556101448e-06,
"loss": 0.272,
"step": 150
},
{
"epoch": 0.5189003436426117,
"grad_norm": 0.46941104531288147,
"learning_rate": 5.570500604890124e-06,
"loss": 0.3198,
"step": 151
},
{
"epoch": 0.5223367697594502,
"grad_norm": 0.49389877915382385,
"learning_rate": 5.510669998590074e-06,
"loss": 0.291,
"step": 152
},
{
"epoch": 0.5257731958762887,
"grad_norm": 0.4679196774959564,
"learning_rate": 5.450765405560328e-06,
"loss": 0.2919,
"step": 153
},
{
"epoch": 0.5292096219931272,
"grad_norm": 0.48511338233947754,
"learning_rate": 5.390795504879243e-06,
"loss": 0.3203,
"step": 154
},
{
"epoch": 0.5326460481099656,
"grad_norm": 0.49272918701171875,
"learning_rate": 5.330768985087059e-06,
"loss": 0.3433,
"step": 155
},
{
"epoch": 0.5360824742268041,
"grad_norm": 0.45313969254493713,
"learning_rate": 5.270694542927089e-06,
"loss": 0.3076,
"step": 156
},
{
"epoch": 0.5395189003436426,
"grad_norm": 0.45044413208961487,
"learning_rate": 5.2105808820857126e-06,
"loss": 0.3174,
"step": 157
},
{
"epoch": 0.5429553264604811,
"grad_norm": 0.4462970495223999,
"learning_rate": 5.150436711931387e-06,
"loss": 0.3114,
"step": 158
},
{
"epoch": 0.5463917525773195,
"grad_norm": 0.40542709827423096,
"learning_rate": 5.090270746252803e-06,
"loss": 0.2906,
"step": 159
},
{
"epoch": 0.5498281786941581,
"grad_norm": 0.44431909918785095,
"learning_rate": 5.030091701996428e-06,
"loss": 0.2672,
"step": 160
},
{
"epoch": 0.5532646048109966,
"grad_norm": 0.4110300540924072,
"learning_rate": 4.9699082980035735e-06,
"loss": 0.2853,
"step": 161
},
{
"epoch": 0.5567010309278351,
"grad_norm": 0.5007030367851257,
"learning_rate": 4.909729253747197e-06,
"loss": 0.3131,
"step": 162
},
{
"epoch": 0.5601374570446735,
"grad_norm": 0.5697958469390869,
"learning_rate": 4.8495632880686155e-06,
"loss": 0.2935,
"step": 163
},
{
"epoch": 0.563573883161512,
"grad_norm": 0.43557995557785034,
"learning_rate": 4.789419117914288e-06,
"loss": 0.3175,
"step": 164
},
{
"epoch": 0.5670103092783505,
"grad_norm": 0.4249924123287201,
"learning_rate": 4.729305457072913e-06,
"loss": 0.3049,
"step": 165
},
{
"epoch": 0.570446735395189,
"grad_norm": 0.4352003335952759,
"learning_rate": 4.6692310149129425e-06,
"loss": 0.2796,
"step": 166
},
{
"epoch": 0.5738831615120275,
"grad_norm": 0.4522886872291565,
"learning_rate": 4.609204495120759e-06,
"loss": 0.2519,
"step": 167
},
{
"epoch": 0.5773195876288659,
"grad_norm": 0.4559308588504791,
"learning_rate": 4.549234594439674e-06,
"loss": 0.3026,
"step": 168
},
{
"epoch": 0.5807560137457045,
"grad_norm": 0.5845784544944763,
"learning_rate": 4.489330001409929e-06,
"loss": 0.2776,
"step": 169
},
{
"epoch": 0.584192439862543,
"grad_norm": 0.46254462003707886,
"learning_rate": 4.429499395109877e-06,
"loss": 0.3038,
"step": 170
},
{
"epoch": 0.5876288659793815,
"grad_norm": 0.42879557609558105,
"learning_rate": 4.369751443898554e-06,
"loss": 0.2798,
"step": 171
},
{
"epoch": 0.5910652920962199,
"grad_norm": 0.46193256974220276,
"learning_rate": 4.310094804159784e-06,
"loss": 0.2779,
"step": 172
},
{
"epoch": 0.5945017182130584,
"grad_norm": 0.5215131640434265,
"learning_rate": 4.250538119048036e-06,
"loss": 0.3455,
"step": 173
},
{
"epoch": 0.5979381443298969,
"grad_norm": 0.44782477617263794,
"learning_rate": 4.191090017236177e-06,
"loss": 0.3083,
"step": 174
},
{
"epoch": 0.6013745704467354,
"grad_norm": 0.48400431871414185,
"learning_rate": 4.131759111665349e-06,
"loss": 0.2975,
"step": 175
},
{
"epoch": 0.6048109965635738,
"grad_norm": 0.41204163432121277,
"learning_rate": 4.072553998297103e-06,
"loss": 0.2757,
"step": 176
},
{
"epoch": 0.6082474226804123,
"grad_norm": 0.4403284192085266,
"learning_rate": 4.013483254868001e-06,
"loss": 0.2673,
"step": 177
},
{
"epoch": 0.6116838487972509,
"grad_norm": 0.44256702065467834,
"learning_rate": 3.9545554396468655e-06,
"loss": 0.2941,
"step": 178
},
{
"epoch": 0.6151202749140894,
"grad_norm": 0.44513365626335144,
"learning_rate": 3.89577909019484e-06,
"loss": 0.2773,
"step": 179
},
{
"epoch": 0.6185567010309279,
"grad_norm": 0.4568428695201874,
"learning_rate": 3.8371627221284495e-06,
"loss": 0.2946,
"step": 180
},
{
"epoch": 0.6219931271477663,
"grad_norm": 0.4727785289287567,
"learning_rate": 3.7787148278858453e-06,
"loss": 0.3204,
"step": 181
},
{
"epoch": 0.6254295532646048,
"grad_norm": 0.42963531613349915,
"learning_rate": 3.7204438754964113e-06,
"loss": 0.2498,
"step": 182
},
{
"epoch": 0.6288659793814433,
"grad_norm": 0.42374187707901,
"learning_rate": 3.662358307353897e-06,
"loss": 0.2989,
"step": 183
},
{
"epoch": 0.6323024054982818,
"grad_norm": 0.495601624250412,
"learning_rate": 3.6044665389932663e-06,
"loss": 0.3289,
"step": 184
},
{
"epoch": 0.6357388316151202,
"grad_norm": 0.4799273610115051,
"learning_rate": 3.5467769578714455e-06,
"loss": 0.2627,
"step": 185
},
{
"epoch": 0.6391752577319587,
"grad_norm": 0.4173251986503601,
"learning_rate": 3.489297922152136e-06,
"loss": 0.2576,
"step": 186
},
{
"epoch": 0.6426116838487973,
"grad_norm": 0.4326072335243225,
"learning_rate": 3.432037759494867e-06,
"loss": 0.2963,
"step": 187
},
{
"epoch": 0.6460481099656358,
"grad_norm": 0.4364977180957794,
"learning_rate": 3.375004765848463e-06,
"loss": 0.2613,
"step": 188
},
{
"epoch": 0.6494845360824743,
"grad_norm": 0.4443832337856293,
"learning_rate": 3.3182072042491244e-06,
"loss": 0.3029,
"step": 189
},
{
"epoch": 0.6529209621993127,
"grad_norm": 0.4171838164329529,
"learning_rate": 3.2616533036232635e-06,
"loss": 0.2758,
"step": 190
},
{
"epoch": 0.6563573883161512,
"grad_norm": 0.44542646408081055,
"learning_rate": 3.205351257595272e-06,
"loss": 0.2947,
"step": 191
},
{
"epoch": 0.6597938144329897,
"grad_norm": 0.42361024022102356,
"learning_rate": 3.149309223300428e-06,
"loss": 0.2644,
"step": 192
},
{
"epoch": 0.6632302405498282,
"grad_norm": 0.42630866169929504,
"learning_rate": 3.093535320203074e-06,
"loss": 0.2739,
"step": 193
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.45428791642189026,
"learning_rate": 3.0380376289202497e-06,
"loss": 0.3163,
"step": 194
},
{
"epoch": 0.6701030927835051,
"grad_norm": 0.3921767771244049,
"learning_rate": 2.982824190050958e-06,
"loss": 0.2675,
"step": 195
},
{
"epoch": 0.6735395189003437,
"grad_norm": 0.41902685165405273,
"learning_rate": 2.927903003011241e-06,
"loss": 0.2682,
"step": 196
},
{
"epoch": 0.6769759450171822,
"grad_norm": 0.4761234521865845,
"learning_rate": 2.8732820248752016e-06,
"loss": 0.326,
"step": 197
},
{
"epoch": 0.6804123711340206,
"grad_norm": 0.4254363775253296,
"learning_rate": 2.8189691692221627e-06,
"loss": 0.2915,
"step": 198
},
{
"epoch": 0.6838487972508591,
"grad_norm": 0.41192832589149475,
"learning_rate": 2.7649723049901554e-06,
"loss": 0.2926,
"step": 199
},
{
"epoch": 0.6872852233676976,
"grad_norm": 0.47728779911994934,
"learning_rate": 2.711299255335833e-06,
"loss": 0.2738,
"step": 200
},
{
"epoch": 0.6907216494845361,
"grad_norm": 0.5057693123817444,
"learning_rate": 2.65795779650105e-06,
"loss": 0.2806,
"step": 201
},
{
"epoch": 0.6941580756013745,
"grad_norm": 0.4783018231391907,
"learning_rate": 2.6049556566862234e-06,
"loss": 0.2891,
"step": 202
},
{
"epoch": 0.697594501718213,
"grad_norm": 0.5034655928611755,
"learning_rate": 2.552300514930657e-06,
"loss": 0.2977,
"step": 203
},
{
"epoch": 0.7010309278350515,
"grad_norm": 0.41531136631965637,
"learning_rate": 2.5000000000000015e-06,
"loss": 0.275,
"step": 204
},
{
"epoch": 0.7044673539518901,
"grad_norm": 0.4129270315170288,
"learning_rate": 2.4480616892809593e-06,
"loss": 0.2381,
"step": 205
},
{
"epoch": 0.7079037800687286,
"grad_norm": 0.46732890605926514,
"learning_rate": 2.396493107683488e-06,
"loss": 0.2779,
"step": 206
},
{
"epoch": 0.711340206185567,
"grad_norm": 0.4486265480518341,
"learning_rate": 2.345301726550567e-06,
"loss": 0.267,
"step": 207
},
{
"epoch": 0.7147766323024055,
"grad_norm": 0.45522594451904297,
"learning_rate": 2.2944949625757295e-06,
"loss": 0.3015,
"step": 208
},
{
"epoch": 0.718213058419244,
"grad_norm": 0.6432047486305237,
"learning_rate": 2.244080176728536e-06,
"loss": 0.2662,
"step": 209
},
{
"epoch": 0.7216494845360825,
"grad_norm": 0.4285944104194641,
"learning_rate": 2.1940646731880887e-06,
"loss": 0.3036,
"step": 210
},
{
"epoch": 0.7250859106529209,
"grad_norm": 0.4231526255607605,
"learning_rate": 2.1444556982847996e-06,
"loss": 0.2896,
"step": 211
},
{
"epoch": 0.7285223367697594,
"grad_norm": 0.42347821593284607,
"learning_rate": 2.095260439450526e-06,
"loss": 0.2684,
"step": 212
},
{
"epoch": 0.7319587628865979,
"grad_norm": 0.43296700716018677,
"learning_rate": 2.0464860241772454e-06,
"loss": 0.2788,
"step": 213
},
{
"epoch": 0.7353951890034365,
"grad_norm": 0.42099738121032715,
"learning_rate": 1.998139518984409e-06,
"loss": 0.2873,
"step": 214
},
{
"epoch": 0.738831615120275,
"grad_norm": 0.467330664396286,
"learning_rate": 1.9502279283951363e-06,
"loss": 0.3004,
"step": 215
},
{
"epoch": 0.7422680412371134,
"grad_norm": 0.4013458788394928,
"learning_rate": 1.9027581939213852e-06,
"loss": 0.2443,
"step": 216
},
{
"epoch": 0.7457044673539519,
"grad_norm": 0.4329524040222168,
"learning_rate": 1.8557371930582579e-06,
"loss": 0.289,
"step": 217
},
{
"epoch": 0.7491408934707904,
"grad_norm": 0.4044208824634552,
"learning_rate": 1.8091717382875723e-06,
"loss": 0.2895,
"step": 218
},
{
"epoch": 0.7525773195876289,
"grad_norm": 0.4422338008880615,
"learning_rate": 1.7630685760908623e-06,
"loss": 0.2801,
"step": 219
},
{
"epoch": 0.7560137457044673,
"grad_norm": 0.4364713728427887,
"learning_rate": 1.7174343859719334e-06,
"loss": 0.279,
"step": 220
},
{
"epoch": 0.7594501718213058,
"grad_norm": 0.4441254734992981,
"learning_rate": 1.6722757794891287e-06,
"loss": 0.271,
"step": 221
},
{
"epoch": 0.7628865979381443,
"grad_norm": 0.446011483669281,
"learning_rate": 1.627599299297431e-06,
"loss": 0.2627,
"step": 222
},
{
"epoch": 0.7663230240549829,
"grad_norm": 0.4614294469356537,
"learning_rate": 1.5834114182005544e-06,
"loss": 0.29,
"step": 223
},
{
"epoch": 0.7697594501718213,
"grad_norm": 0.419196754693985,
"learning_rate": 1.5397185382131524e-06,
"loss": 0.2672,
"step": 224
},
{
"epoch": 0.7731958762886598,
"grad_norm": 0.47108274698257446,
"learning_rate": 1.4965269896332884e-06,
"loss": 0.2684,
"step": 225
},
{
"epoch": 0.7766323024054983,
"grad_norm": 0.48111191391944885,
"learning_rate": 1.4538430301252783e-06,
"loss": 0.327,
"step": 226
},
{
"epoch": 0.7800687285223368,
"grad_norm": 0.4369780421257019,
"learning_rate": 1.411672843813086e-06,
"loss": 0.283,
"step": 227
},
{
"epoch": 0.7835051546391752,
"grad_norm": 0.4390357732772827,
"learning_rate": 1.370022540384347e-06,
"loss": 0.3011,
"step": 228
},
{
"epoch": 0.7869415807560137,
"grad_norm": 0.40961310267448425,
"learning_rate": 1.3288981542051844e-06,
"loss": 0.2714,
"step": 229
},
{
"epoch": 0.7903780068728522,
"grad_norm": 0.4285667836666107,
"learning_rate": 1.2883056434459506e-06,
"loss": 0.2577,
"step": 230
},
{
"epoch": 0.7938144329896907,
"grad_norm": 0.4189035892486572,
"learning_rate": 1.2482508892179884e-06,
"loss": 0.27,
"step": 231
},
{
"epoch": 0.7972508591065293,
"grad_norm": 0.3942403495311737,
"learning_rate": 1.2087396947215678e-06,
"loss": 0.2605,
"step": 232
},
{
"epoch": 0.8006872852233677,
"grad_norm": 0.49882808327674866,
"learning_rate": 1.1697777844051105e-06,
"loss": 0.3014,
"step": 233
},
{
"epoch": 0.8041237113402062,
"grad_norm": 0.3978383243083954,
"learning_rate": 1.1313708031358183e-06,
"loss": 0.2566,
"step": 234
},
{
"epoch": 0.8075601374570447,
"grad_norm": 0.40835458040237427,
"learning_rate": 1.0935243153818437e-06,
"loss": 0.2851,
"step": 235
},
{
"epoch": 0.8109965635738832,
"grad_norm": 0.37506747245788574,
"learning_rate": 1.0562438044060846e-06,
"loss": 0.2456,
"step": 236
},
{
"epoch": 0.8144329896907216,
"grad_norm": 0.38689589500427246,
"learning_rate": 1.0195346714717813e-06,
"loss": 0.2514,
"step": 237
},
{
"epoch": 0.8178694158075601,
"grad_norm": 0.4160013794898987,
"learning_rate": 9.834022350599538e-07,
"loss": 0.297,
"step": 238
},
{
"epoch": 0.8213058419243986,
"grad_norm": 0.42283937335014343,
"learning_rate": 9.47851730098856e-07,
"loss": 0.2556,
"step": 239
},
{
"epoch": 0.8247422680412371,
"grad_norm": 0.464019775390625,
"learning_rate": 9.128883072055411e-07,
"loss": 0.283,
"step": 240
},
{
"epoch": 0.8281786941580757,
"grad_norm": 0.44965192675590515,
"learning_rate": 8.785170319396174e-07,
"loss": 0.2795,
"step": 241
},
{
"epoch": 0.8316151202749141,
"grad_norm": 0.4575161039829254,
"learning_rate": 8.447428840693489e-07,
"loss": 0.2793,
"step": 242
},
{
"epoch": 0.8350515463917526,
"grad_norm": 0.4319188594818115,
"learning_rate": 8.115707568501768e-07,
"loss": 0.2569,
"step": 243
},
{
"epoch": 0.8384879725085911,
"grad_norm": 0.43016260862350464,
"learning_rate": 7.790054563157745e-07,
"loss": 0.3022,
"step": 244
},
{
"epoch": 0.8419243986254296,
"grad_norm": 0.4093724489212036,
"learning_rate": 7.470517005817473e-07,
"loss": 0.251,
"step": 245
},
{
"epoch": 0.845360824742268,
"grad_norm": 0.4726552963256836,
"learning_rate": 7.157141191620548e-07,
"loss": 0.3445,
"step": 246
},
{
"epoch": 0.8487972508591065,
"grad_norm": 0.4052773118019104,
"learning_rate": 6.849972522982845e-07,
"loss": 0.2703,
"step": 247
},
{
"epoch": 0.852233676975945,
"grad_norm": 0.4114190936088562,
"learning_rate": 6.549055503018575e-07,
"loss": 0.2632,
"step": 248
},
{
"epoch": 0.8556701030927835,
"grad_norm": 0.43223080039024353,
"learning_rate": 6.254433729092518e-07,
"loss": 0.303,
"step": 249
},
{
"epoch": 0.8591065292096219,
"grad_norm": 0.42186257243156433,
"learning_rate": 5.966149886503614e-07,
"loss": 0.2673,
"step": 250
},
{
"epoch": 0.8625429553264605,
"grad_norm": 0.4302360713481903,
"learning_rate": 5.684245742300625e-07,
"loss": 0.2864,
"step": 251
},
{
"epoch": 0.865979381443299,
"grad_norm": 0.4030810594558716,
"learning_rate": 5.408762139230889e-07,
"loss": 0.2608,
"step": 252
},
{
"epoch": 0.8694158075601375,
"grad_norm": 0.4175474941730499,
"learning_rate": 5.139738989822901e-07,
"loss": 0.2626,
"step": 253
},
{
"epoch": 0.872852233676976,
"grad_norm": 0.4360395669937134,
"learning_rate": 4.877215270603752e-07,
"loss": 0.2955,
"step": 254
},
{
"epoch": 0.8762886597938144,
"grad_norm": 0.417015939950943,
"learning_rate": 4.6212290164521554e-07,
"loss": 0.2755,
"step": 255
},
{
"epoch": 0.8797250859106529,
"grad_norm": 0.42125383019447327,
"learning_rate": 4.371817315087845e-07,
"loss": 0.2799,
"step": 256
},
{
"epoch": 0.8831615120274914,
"grad_norm": 0.41367045044898987,
"learning_rate": 4.1290163016982855e-07,
"loss": 0.2639,
"step": 257
},
{
"epoch": 0.8865979381443299,
"grad_norm": 0.44137805700302124,
"learning_rate": 3.8928611537033424e-07,
"loss": 0.2726,
"step": 258
},
{
"epoch": 0.8900343642611683,
"grad_norm": 0.4337911307811737,
"learning_rate": 3.663386085658693e-07,
"loss": 0.2758,
"step": 259
},
{
"epoch": 0.8934707903780069,
"grad_norm": 0.3822688162326813,
"learning_rate": 3.4406243442987765e-07,
"loss": 0.2581,
"step": 260
},
{
"epoch": 0.8969072164948454,
"grad_norm": 0.4298803508281708,
"learning_rate": 3.224608203719953e-07,
"loss": 0.281,
"step": 261
},
{
"epoch": 0.9003436426116839,
"grad_norm": 0.46638035774230957,
"learning_rate": 3.015368960704584e-07,
"loss": 0.3077,
"step": 262
},
{
"epoch": 0.9037800687285223,
"grad_norm": 0.4740549325942993,
"learning_rate": 2.812936930186688e-07,
"loss": 0.3094,
"step": 263
},
{
"epoch": 0.9072164948453608,
"grad_norm": 0.4159590005874634,
"learning_rate": 2.617341440859883e-07,
"loss": 0.2729,
"step": 264
},
{
"epoch": 0.9106529209621993,
"grad_norm": 0.4018743932247162,
"learning_rate": 2.428610830928152e-07,
"loss": 0.2579,
"step": 265
},
{
"epoch": 0.9140893470790378,
"grad_norm": 0.4325768053531647,
"learning_rate": 2.2467724440002336e-07,
"loss": 0.3084,
"step": 266
},
{
"epoch": 0.9175257731958762,
"grad_norm": 0.4185415208339691,
"learning_rate": 2.0718526251279346e-07,
"loss": 0.269,
"step": 267
},
{
"epoch": 0.9209621993127147,
"grad_norm": 0.41172951459884644,
"learning_rate": 1.9038767169893058e-07,
"loss": 0.2688,
"step": 268
},
{
"epoch": 0.9243986254295533,
"grad_norm": 0.4149523675441742,
"learning_rate": 1.7428690562169003e-07,
"loss": 0.2528,
"step": 269
},
{
"epoch": 0.9278350515463918,
"grad_norm": 0.4421743154525757,
"learning_rate": 1.5888529698718347e-07,
"loss": 0.2684,
"step": 270
},
{
"epoch": 0.9312714776632303,
"grad_norm": 0.4239802360534668,
"learning_rate": 1.4418507720641794e-07,
"loss": 0.2632,
"step": 271
},
{
"epoch": 0.9347079037800687,
"grad_norm": 0.4376448690891266,
"learning_rate": 1.3018837607199909e-07,
"loss": 0.2861,
"step": 272
},
{
"epoch": 0.9381443298969072,
"grad_norm": 0.432050496339798,
"learning_rate": 1.1689722144956672e-07,
"loss": 0.287,
"step": 273
},
{
"epoch": 0.9415807560137457,
"grad_norm": 0.5766061544418335,
"learning_rate": 1.0431353898399388e-07,
"loss": 0.3096,
"step": 274
},
{
"epoch": 0.9450171821305842,
"grad_norm": 0.40074169635772705,
"learning_rate": 9.243915182039431e-08,
"loss": 0.2671,
"step": 275
},
{
"epoch": 0.9484536082474226,
"grad_norm": 0.4120236933231354,
"learning_rate": 8.127578033998663e-08,
"loss": 0.2707,
"step": 276
},
{
"epoch": 0.9518900343642611,
"grad_norm": 0.478630006313324,
"learning_rate": 7.082504191083417e-08,
"loss": 0.2886,
"step": 277
},
{
"epoch": 0.9553264604810997,
"grad_norm": 0.4233369529247284,
"learning_rate": 6.108845065352864e-08,
"loss": 0.2715,
"step": 278
},
{
"epoch": 0.9587628865979382,
"grad_norm": 0.43706801533699036,
"learning_rate": 5.206741722181385e-08,
"loss": 0.289,
"step": 279
},
{
"epoch": 0.9621993127147767,
"grad_norm": 0.47116661071777344,
"learning_rate": 4.376324859820924e-08,
"loss": 0.3081,
"step": 280
},
{
"epoch": 0.9656357388316151,
"grad_norm": 0.40870651602745056,
"learning_rate": 3.617714790465576e-08,
"loss": 0.2667,
"step": 281
},
{
"epoch": 0.9690721649484536,
"grad_norm": 0.40682822465896606,
"learning_rate": 2.9310214228202016e-08,
"loss": 0.2728,
"step": 282
},
{
"epoch": 0.9725085910652921,
"grad_norm": 0.4911557734012604,
"learning_rate": 2.3163442461766604e-08,
"loss": 0.3386,
"step": 283
},
{
"epoch": 0.9759450171821306,
"grad_norm": 0.4293031394481659,
"learning_rate": 1.7737723159999e-08,
"loss": 0.2852,
"step": 284
},
{
"epoch": 0.979381443298969,
"grad_norm": 0.39959046244621277,
"learning_rate": 1.3033842410251074e-08,
"loss": 0.2668,
"step": 285
},
{
"epoch": 0.9828178694158075,
"grad_norm": 0.42357000708580017,
"learning_rate": 9.052481718690998e-09,
"loss": 0.2659,
"step": 286
},
{
"epoch": 0.9862542955326461,
"grad_norm": 0.46328771114349365,
"learning_rate": 5.794217911562205e-09,
"loss": 0.2965,
"step": 287
},
{
"epoch": 0.9896907216494846,
"grad_norm": 0.39359450340270996,
"learning_rate": 3.2595230516152543e-09,
"loss": 0.2548,
"step": 288
},
{
"epoch": 0.993127147766323,
"grad_norm": 0.447672963142395,
"learning_rate": 1.4487643697103092e-09,
"loss": 0.2804,
"step": 289
},
{
"epoch": 0.9965635738831615,
"grad_norm": 0.48000165820121765,
"learning_rate": 3.6220421161692333e-10,
"loss": 0.3062,
"step": 290
},
{
"epoch": 1.0,
"grad_norm": 0.4716525673866272,
"learning_rate": 0.0,
"loss": 0.2894,
"step": 291
},
{
"epoch": 1.0,
"eval_loss": 0.39179933071136475,
"eval_runtime": 1.5263,
"eval_samples_per_second": 150.689,
"eval_steps_per_second": 6.552,
"step": 291
},
{
"epoch": 1.0,
"step": 291,
"total_flos": 2.419849045213184e+16,
"train_loss": 0.4077431839151481,
"train_runtime": 379.788,
"train_samples_per_second": 36.668,
"train_steps_per_second": 0.766
}
],
"logging_steps": 1,
"max_steps": 291,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 800,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2.419849045213184e+16,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}