nvidiallama / checkpoint-168 /trainer_state.json
ugaoo's picture
Upload folder using huggingface_hub
a5b5026 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.8023255813953485,
"eval_steps": 500,
"global_step": 168,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.03488372093023256,
"grad_norm": 34.417083740234375,
"learning_rate": 5.0000000000000004e-08,
"loss": 2.2576,
"step": 1
},
{
"epoch": 0.06976744186046512,
"grad_norm": 34.94503402709961,
"learning_rate": 1.0000000000000001e-07,
"loss": 2.2928,
"step": 2
},
{
"epoch": 0.10465116279069768,
"grad_norm": 34.11349105834961,
"learning_rate": 1.5000000000000002e-07,
"loss": 2.2273,
"step": 3
},
{
"epoch": 0.13953488372093023,
"grad_norm": 35.753822326660156,
"learning_rate": 2.0000000000000002e-07,
"loss": 2.3202,
"step": 4
},
{
"epoch": 0.1744186046511628,
"grad_norm": 34.57868957519531,
"learning_rate": 2.5000000000000004e-07,
"loss": 2.2515,
"step": 5
},
{
"epoch": 0.20930232558139536,
"grad_norm": 34.45582580566406,
"learning_rate": 3.0000000000000004e-07,
"loss": 2.241,
"step": 6
},
{
"epoch": 0.2441860465116279,
"grad_norm": 34.65391540527344,
"learning_rate": 3.5000000000000004e-07,
"loss": 2.2543,
"step": 7
},
{
"epoch": 0.27906976744186046,
"grad_norm": 34.55806350708008,
"learning_rate": 4.0000000000000003e-07,
"loss": 2.2219,
"step": 8
},
{
"epoch": 0.313953488372093,
"grad_norm": 34.35453414916992,
"learning_rate": 4.5000000000000003e-07,
"loss": 2.2105,
"step": 9
},
{
"epoch": 0.3488372093023256,
"grad_norm": 34.049232482910156,
"learning_rate": 5.000000000000001e-07,
"loss": 2.1968,
"step": 10
},
{
"epoch": 0.38372093023255816,
"grad_norm": 33.26045608520508,
"learning_rate": 5.5e-07,
"loss": 2.13,
"step": 11
},
{
"epoch": 0.4186046511627907,
"grad_norm": 33.314796447753906,
"learning_rate": 6.000000000000001e-07,
"loss": 2.1306,
"step": 12
},
{
"epoch": 0.45348837209302323,
"grad_norm": 32.00406265258789,
"learning_rate": 6.5e-07,
"loss": 2.0275,
"step": 13
},
{
"epoch": 0.4883720930232558,
"grad_norm": 32.68654251098633,
"learning_rate": 7.000000000000001e-07,
"loss": 2.0442,
"step": 14
},
{
"epoch": 0.5232558139534884,
"grad_norm": 31.643146514892578,
"learning_rate": 7.5e-07,
"loss": 1.9725,
"step": 15
},
{
"epoch": 0.5581395348837209,
"grad_norm": 30.403486251831055,
"learning_rate": 8.000000000000001e-07,
"loss": 1.8724,
"step": 16
},
{
"epoch": 0.5930232558139535,
"grad_norm": 29.716495513916016,
"learning_rate": 8.500000000000001e-07,
"loss": 1.8118,
"step": 17
},
{
"epoch": 0.627906976744186,
"grad_norm": 29.22306251525879,
"learning_rate": 9.000000000000001e-07,
"loss": 1.7551,
"step": 18
},
{
"epoch": 0.6627906976744186,
"grad_norm": 27.425785064697266,
"learning_rate": 9.500000000000001e-07,
"loss": 1.6087,
"step": 19
},
{
"epoch": 0.6976744186046512,
"grad_norm": 26.653409957885742,
"learning_rate": 1.0000000000000002e-06,
"loss": 1.5191,
"step": 20
},
{
"epoch": 0.7325581395348837,
"grad_norm": 25.145450592041016,
"learning_rate": 1.0500000000000001e-06,
"loss": 1.3758,
"step": 21
},
{
"epoch": 0.7674418604651163,
"grad_norm": 24.684560775756836,
"learning_rate": 1.1e-06,
"loss": 1.2955,
"step": 22
},
{
"epoch": 0.8023255813953488,
"grad_norm": 24.005306243896484,
"learning_rate": 1.1500000000000002e-06,
"loss": 1.1665,
"step": 23
},
{
"epoch": 0.8372093023255814,
"grad_norm": 24.253219604492188,
"learning_rate": 1.2000000000000002e-06,
"loss": 1.0143,
"step": 24
},
{
"epoch": 0.872093023255814,
"grad_norm": 24.97549057006836,
"learning_rate": 1.25e-06,
"loss": 0.875,
"step": 25
},
{
"epoch": 0.9069767441860465,
"grad_norm": 22.349998474121094,
"learning_rate": 1.3e-06,
"loss": 0.6222,
"step": 26
},
{
"epoch": 0.9418604651162791,
"grad_norm": 18.814115524291992,
"learning_rate": 1.3500000000000002e-06,
"loss": 0.4715,
"step": 27
},
{
"epoch": 0.9767441860465116,
"grad_norm": 15.833977699279785,
"learning_rate": 1.4000000000000001e-06,
"loss": 0.374,
"step": 28
},
{
"epoch": 1.0,
"grad_norm": 15.833977699279785,
"learning_rate": 1.45e-06,
"loss": 0.2818,
"step": 29
},
{
"epoch": 1.0348837209302326,
"grad_norm": 14.85798168182373,
"learning_rate": 1.5e-06,
"loss": 0.2559,
"step": 30
},
{
"epoch": 1.069767441860465,
"grad_norm": 6.723965167999268,
"learning_rate": 1.5500000000000002e-06,
"loss": 0.2073,
"step": 31
},
{
"epoch": 1.1046511627906976,
"grad_norm": 4.480190753936768,
"learning_rate": 1.6000000000000001e-06,
"loss": 0.1632,
"step": 32
},
{
"epoch": 1.1395348837209303,
"grad_norm": 3.228346824645996,
"learning_rate": 1.6500000000000003e-06,
"loss": 0.1432,
"step": 33
},
{
"epoch": 1.1744186046511629,
"grad_norm": 2.4646286964416504,
"learning_rate": 1.7000000000000002e-06,
"loss": 0.1166,
"step": 34
},
{
"epoch": 1.2093023255813953,
"grad_norm": 1.721303939819336,
"learning_rate": 1.75e-06,
"loss": 0.1034,
"step": 35
},
{
"epoch": 1.244186046511628,
"grad_norm": 1.6059235334396362,
"learning_rate": 1.8000000000000001e-06,
"loss": 0.0943,
"step": 36
},
{
"epoch": 1.2790697674418605,
"grad_norm": 1.390950322151184,
"learning_rate": 1.85e-06,
"loss": 0.0855,
"step": 37
},
{
"epoch": 1.3139534883720931,
"grad_norm": 1.2893034219741821,
"learning_rate": 1.9000000000000002e-06,
"loss": 0.0797,
"step": 38
},
{
"epoch": 1.3488372093023255,
"grad_norm": 1.0295542478561401,
"learning_rate": 1.9500000000000004e-06,
"loss": 0.072,
"step": 39
},
{
"epoch": 1.3837209302325582,
"grad_norm": 1.0248366594314575,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.068,
"step": 40
},
{
"epoch": 1.4186046511627908,
"grad_norm": 0.7489507794380188,
"learning_rate": 2.05e-06,
"loss": 0.0613,
"step": 41
},
{
"epoch": 1.4534883720930232,
"grad_norm": 0.7149010300636292,
"learning_rate": 2.1000000000000002e-06,
"loss": 0.0596,
"step": 42
},
{
"epoch": 1.4883720930232558,
"grad_norm": 0.5656580328941345,
"learning_rate": 2.15e-06,
"loss": 0.0585,
"step": 43
},
{
"epoch": 1.5232558139534884,
"grad_norm": 0.6951081156730652,
"learning_rate": 2.2e-06,
"loss": 0.066,
"step": 44
},
{
"epoch": 1.558139534883721,
"grad_norm": 0.8682136535644531,
"learning_rate": 2.25e-06,
"loss": 0.0623,
"step": 45
},
{
"epoch": 1.5930232558139537,
"grad_norm": 0.6056260466575623,
"learning_rate": 2.3000000000000004e-06,
"loss": 0.054,
"step": 46
},
{
"epoch": 1.627906976744186,
"grad_norm": 0.35378050804138184,
"learning_rate": 2.35e-06,
"loss": 0.0626,
"step": 47
},
{
"epoch": 1.6627906976744184,
"grad_norm": 0.4003937244415283,
"learning_rate": 2.4000000000000003e-06,
"loss": 0.0524,
"step": 48
},
{
"epoch": 1.697674418604651,
"grad_norm": 0.527799665927887,
"learning_rate": 2.4500000000000003e-06,
"loss": 0.056,
"step": 49
},
{
"epoch": 1.7325581395348837,
"grad_norm": 0.5358547568321228,
"learning_rate": 2.5e-06,
"loss": 0.0527,
"step": 50
},
{
"epoch": 1.7674418604651163,
"grad_norm": 0.45014214515686035,
"learning_rate": 2.55e-06,
"loss": 0.0492,
"step": 51
},
{
"epoch": 1.802325581395349,
"grad_norm": 0.4785670340061188,
"learning_rate": 2.6e-06,
"loss": 0.0525,
"step": 52
},
{
"epoch": 1.8372093023255816,
"grad_norm": 0.4931303560733795,
"learning_rate": 2.6500000000000005e-06,
"loss": 0.0485,
"step": 53
},
{
"epoch": 1.872093023255814,
"grad_norm": 0.2308250367641449,
"learning_rate": 2.7000000000000004e-06,
"loss": 0.0499,
"step": 54
},
{
"epoch": 1.9069767441860463,
"grad_norm": 0.5777614712715149,
"learning_rate": 2.7500000000000004e-06,
"loss": 0.0527,
"step": 55
},
{
"epoch": 1.941860465116279,
"grad_norm": 0.32561731338500977,
"learning_rate": 2.8000000000000003e-06,
"loss": 0.052,
"step": 56
},
{
"epoch": 1.9767441860465116,
"grad_norm": 0.34972742199897766,
"learning_rate": 2.85e-06,
"loss": 0.0467,
"step": 57
},
{
"epoch": 2.0,
"grad_norm": 0.43575090169906616,
"learning_rate": 2.9e-06,
"loss": 0.0441,
"step": 58
},
{
"epoch": 2.0348837209302326,
"grad_norm": 0.3564951419830322,
"learning_rate": 2.95e-06,
"loss": 0.0471,
"step": 59
},
{
"epoch": 2.0697674418604652,
"grad_norm": 0.3126629889011383,
"learning_rate": 3e-06,
"loss": 0.0494,
"step": 60
},
{
"epoch": 2.104651162790698,
"grad_norm": 0.4542100429534912,
"learning_rate": 3.05e-06,
"loss": 0.0456,
"step": 61
},
{
"epoch": 2.13953488372093,
"grad_norm": 0.4597831666469574,
"learning_rate": 3.1000000000000004e-06,
"loss": 0.049,
"step": 62
},
{
"epoch": 2.1744186046511627,
"grad_norm": 0.3821645677089691,
"learning_rate": 3.1500000000000003e-06,
"loss": 0.0461,
"step": 63
},
{
"epoch": 2.2093023255813953,
"grad_norm": 0.3955609202384949,
"learning_rate": 3.2000000000000003e-06,
"loss": 0.0472,
"step": 64
},
{
"epoch": 2.244186046511628,
"grad_norm": 0.33170193433761597,
"learning_rate": 3.2500000000000002e-06,
"loss": 0.0433,
"step": 65
},
{
"epoch": 2.2790697674418605,
"grad_norm": 0.389060378074646,
"learning_rate": 3.3000000000000006e-06,
"loss": 0.0438,
"step": 66
},
{
"epoch": 2.313953488372093,
"grad_norm": 0.31669771671295166,
"learning_rate": 3.3500000000000005e-06,
"loss": 0.0403,
"step": 67
},
{
"epoch": 2.3488372093023258,
"grad_norm": 0.36098915338516235,
"learning_rate": 3.4000000000000005e-06,
"loss": 0.0392,
"step": 68
},
{
"epoch": 2.383720930232558,
"grad_norm": 0.4905363619327545,
"learning_rate": 3.45e-06,
"loss": 0.0391,
"step": 69
},
{
"epoch": 2.4186046511627906,
"grad_norm": 0.32984688878059387,
"learning_rate": 3.5e-06,
"loss": 0.0432,
"step": 70
},
{
"epoch": 2.453488372093023,
"grad_norm": 0.3890271782875061,
"learning_rate": 3.5500000000000003e-06,
"loss": 0.0411,
"step": 71
},
{
"epoch": 2.488372093023256,
"grad_norm": 0.4199780821800232,
"learning_rate": 3.6000000000000003e-06,
"loss": 0.0404,
"step": 72
},
{
"epoch": 2.5232558139534884,
"grad_norm": 0.5035303831100464,
"learning_rate": 3.65e-06,
"loss": 0.0417,
"step": 73
},
{
"epoch": 2.558139534883721,
"grad_norm": 0.45307648181915283,
"learning_rate": 3.7e-06,
"loss": 0.0397,
"step": 74
},
{
"epoch": 2.5930232558139537,
"grad_norm": 0.36111533641815186,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.0372,
"step": 75
},
{
"epoch": 2.6279069767441863,
"grad_norm": 0.4821806252002716,
"learning_rate": 3.8000000000000005e-06,
"loss": 0.0375,
"step": 76
},
{
"epoch": 2.6627906976744184,
"grad_norm": 0.4104216992855072,
"learning_rate": 3.85e-06,
"loss": 0.0363,
"step": 77
},
{
"epoch": 2.697674418604651,
"grad_norm": 0.4832363724708557,
"learning_rate": 3.900000000000001e-06,
"loss": 0.0406,
"step": 78
},
{
"epoch": 2.7325581395348837,
"grad_norm": 0.2809476852416992,
"learning_rate": 3.95e-06,
"loss": 0.037,
"step": 79
},
{
"epoch": 2.7674418604651163,
"grad_norm": 0.22463542222976685,
"learning_rate": 4.000000000000001e-06,
"loss": 0.0359,
"step": 80
},
{
"epoch": 2.802325581395349,
"grad_norm": 0.5055042505264282,
"learning_rate": 4.05e-06,
"loss": 0.0403,
"step": 81
},
{
"epoch": 2.8372093023255816,
"grad_norm": 0.26646772027015686,
"learning_rate": 4.1e-06,
"loss": 0.0346,
"step": 82
},
{
"epoch": 2.8720930232558137,
"grad_norm": 0.3047487437725067,
"learning_rate": 4.15e-06,
"loss": 0.0381,
"step": 83
},
{
"epoch": 2.9069767441860463,
"grad_norm": 0.24949510395526886,
"learning_rate": 4.2000000000000004e-06,
"loss": 0.0312,
"step": 84
},
{
"epoch": 2.941860465116279,
"grad_norm": 0.328715980052948,
"learning_rate": 4.25e-06,
"loss": 0.0305,
"step": 85
},
{
"epoch": 2.9767441860465116,
"grad_norm": 0.3866642117500305,
"learning_rate": 4.3e-06,
"loss": 0.0304,
"step": 86
},
{
"epoch": 3.0,
"grad_norm": 0.4193498194217682,
"learning_rate": 4.350000000000001e-06,
"loss": 0.033,
"step": 87
},
{
"epoch": 3.0348837209302326,
"grad_norm": 0.417618066072464,
"learning_rate": 4.4e-06,
"loss": 0.0276,
"step": 88
},
{
"epoch": 3.0697674418604652,
"grad_norm": 0.29087409377098083,
"learning_rate": 4.450000000000001e-06,
"loss": 0.0303,
"step": 89
},
{
"epoch": 3.104651162790698,
"grad_norm": 0.2996387779712677,
"learning_rate": 4.5e-06,
"loss": 0.0267,
"step": 90
},
{
"epoch": 3.13953488372093,
"grad_norm": 0.27138552069664,
"learning_rate": 4.5500000000000005e-06,
"loss": 0.0236,
"step": 91
},
{
"epoch": 3.1744186046511627,
"grad_norm": 0.30709055066108704,
"learning_rate": 4.600000000000001e-06,
"loss": 0.0176,
"step": 92
},
{
"epoch": 3.2093023255813953,
"grad_norm": 0.2744629979133606,
"learning_rate": 4.65e-06,
"loss": 0.023,
"step": 93
},
{
"epoch": 3.244186046511628,
"grad_norm": 0.30749836564064026,
"learning_rate": 4.7e-06,
"loss": 0.0228,
"step": 94
},
{
"epoch": 3.2790697674418605,
"grad_norm": 0.27703243494033813,
"learning_rate": 4.75e-06,
"loss": 0.0232,
"step": 95
},
{
"epoch": 3.313953488372093,
"grad_norm": 0.41885659098625183,
"learning_rate": 4.800000000000001e-06,
"loss": 0.0205,
"step": 96
},
{
"epoch": 3.3488372093023258,
"grad_norm": 0.32025253772735596,
"learning_rate": 4.85e-06,
"loss": 0.0199,
"step": 97
},
{
"epoch": 3.383720930232558,
"grad_norm": 0.2905832529067993,
"learning_rate": 4.9000000000000005e-06,
"loss": 0.0175,
"step": 98
},
{
"epoch": 3.4186046511627906,
"grad_norm": 0.34518706798553467,
"learning_rate": 4.95e-06,
"loss": 0.019,
"step": 99
},
{
"epoch": 3.453488372093023,
"grad_norm": 0.24052713811397552,
"learning_rate": 5e-06,
"loss": 0.0176,
"step": 100
},
{
"epoch": 3.488372093023256,
"grad_norm": 0.28509387373924255,
"learning_rate": 4.997332437005932e-06,
"loss": 0.0176,
"step": 101
},
{
"epoch": 3.5232558139534884,
"grad_norm": 0.30950990319252014,
"learning_rate": 4.989335440737587e-06,
"loss": 0.0165,
"step": 102
},
{
"epoch": 3.558139534883721,
"grad_norm": 0.2843819558620453,
"learning_rate": 4.976026077188013e-06,
"loss": 0.0142,
"step": 103
},
{
"epoch": 3.5930232558139537,
"grad_norm": 0.2819436192512512,
"learning_rate": 4.957432749209755e-06,
"loss": 0.0136,
"step": 104
},
{
"epoch": 3.6279069767441863,
"grad_norm": 0.2749376893043518,
"learning_rate": 4.933595135901733e-06,
"loss": 0.0132,
"step": 105
},
{
"epoch": 3.6627906976744184,
"grad_norm": 0.25022268295288086,
"learning_rate": 4.904564107932048e-06,
"loss": 0.0102,
"step": 106
},
{
"epoch": 3.697674418604651,
"grad_norm": 0.25575509667396545,
"learning_rate": 4.870401618977415e-06,
"loss": 0.0116,
"step": 107
},
{
"epoch": 3.7325581395348837,
"grad_norm": 0.2706020176410675,
"learning_rate": 4.83118057351089e-06,
"loss": 0.0111,
"step": 108
},
{
"epoch": 3.7674418604651163,
"grad_norm": 0.33668869733810425,
"learning_rate": 4.786984671220053e-06,
"loss": 0.0158,
"step": 109
},
{
"epoch": 3.802325581395349,
"grad_norm": 0.2773989737033844,
"learning_rate": 4.737908228387656e-06,
"loss": 0.01,
"step": 110
},
{
"epoch": 3.8372093023255816,
"grad_norm": 0.23735634982585907,
"learning_rate": 4.684055976615924e-06,
"loss": 0.0072,
"step": 111
},
{
"epoch": 3.8720930232558137,
"grad_norm": 0.30508551001548767,
"learning_rate": 4.625542839324036e-06,
"loss": 0.0084,
"step": 112
},
{
"epoch": 3.9069767441860463,
"grad_norm": 0.2335406094789505,
"learning_rate": 4.562493686495756e-06,
"loss": 0.0079,
"step": 113
},
{
"epoch": 3.941860465116279,
"grad_norm": 0.21952728927135468,
"learning_rate": 4.4950430682005995e-06,
"loss": 0.0073,
"step": 114
},
{
"epoch": 3.9767441860465116,
"grad_norm": 0.23103268444538116,
"learning_rate": 4.423334927457198e-06,
"loss": 0.0086,
"step": 115
},
{
"epoch": 4.0,
"grad_norm": 0.23103268444538116,
"learning_rate": 4.3475222930516484e-06,
"loss": 0.0091,
"step": 116
},
{
"epoch": 4.034883720930233,
"grad_norm": 0.39881983399391174,
"learning_rate": 4.267766952966369e-06,
"loss": 0.0044,
"step": 117
},
{
"epoch": 4.069767441860465,
"grad_norm": 0.15434899926185608,
"learning_rate": 4.184239109116393e-06,
"loss": 0.0034,
"step": 118
},
{
"epoch": 4.104651162790698,
"grad_norm": 0.1445930004119873,
"learning_rate": 4.097117014129903e-06,
"loss": 0.0039,
"step": 119
},
{
"epoch": 4.1395348837209305,
"grad_norm": 0.23641756176948547,
"learning_rate": 4.006586590948141e-06,
"loss": 0.0035,
"step": 120
},
{
"epoch": 4.174418604651163,
"grad_norm": 0.14500728249549866,
"learning_rate": 3.91284103605648e-06,
"loss": 0.0069,
"step": 121
},
{
"epoch": 4.209302325581396,
"grad_norm": 0.1661936342716217,
"learning_rate": 3.81608040719339e-06,
"loss": 0.0028,
"step": 122
},
{
"epoch": 4.2441860465116275,
"grad_norm": 0.15642984211444855,
"learning_rate": 3.7165111964171407e-06,
"loss": 0.0035,
"step": 123
},
{
"epoch": 4.27906976744186,
"grad_norm": 0.16263604164123535,
"learning_rate": 3.6143458894413463e-06,
"loss": 0.0047,
"step": 124
},
{
"epoch": 4.313953488372093,
"grad_norm": 0.13780659437179565,
"learning_rate": 3.5098025121797375e-06,
"loss": 0.0027,
"step": 125
},
{
"epoch": 4.348837209302325,
"grad_norm": 0.37652286887168884,
"learning_rate": 3.403104165467883e-06,
"loss": 0.0041,
"step": 126
},
{
"epoch": 4.383720930232558,
"grad_norm": 0.19331718981266022,
"learning_rate": 3.2944785489547544e-06,
"loss": 0.0032,
"step": 127
},
{
"epoch": 4.4186046511627906,
"grad_norm": 0.18035675585269928,
"learning_rate": 3.184157475180208e-06,
"loss": 0.003,
"step": 128
},
{
"epoch": 4.453488372093023,
"grad_norm": 0.2259574979543686,
"learning_rate": 3.0723763748753354e-06,
"loss": 0.004,
"step": 129
},
{
"epoch": 4.488372093023256,
"grad_norm": 0.31054359674453735,
"learning_rate": 2.9593737945414264e-06,
"loss": 0.003,
"step": 130
},
{
"epoch": 4.523255813953488,
"grad_norm": 0.11562985181808472,
"learning_rate": 2.845390887379706e-06,
"loss": 0.0016,
"step": 131
},
{
"epoch": 4.558139534883721,
"grad_norm": 0.16370315849781036,
"learning_rate": 2.730670898658255e-06,
"loss": 0.0049,
"step": 132
},
{
"epoch": 4.593023255813954,
"grad_norm": 0.2251535803079605,
"learning_rate": 2.6154586466143495e-06,
"loss": 0.0034,
"step": 133
},
{
"epoch": 4.627906976744186,
"grad_norm": 0.15986023843288422,
"learning_rate": 2.5e-06,
"loss": 0.0035,
"step": 134
},
{
"epoch": 4.662790697674419,
"grad_norm": 0.16980388760566711,
"learning_rate": 2.3845413533856517e-06,
"loss": 0.0029,
"step": 135
},
{
"epoch": 4.6976744186046515,
"grad_norm": 0.19148623943328857,
"learning_rate": 2.269329101341745e-06,
"loss": 0.0031,
"step": 136
},
{
"epoch": 4.732558139534884,
"grad_norm": 0.12943695485591888,
"learning_rate": 2.1546091126202955e-06,
"loss": 0.0035,
"step": 137
},
{
"epoch": 4.767441860465116,
"grad_norm": 0.11984245479106903,
"learning_rate": 2.040626205458574e-06,
"loss": 0.0028,
"step": 138
},
{
"epoch": 4.8023255813953485,
"grad_norm": 0.1532817780971527,
"learning_rate": 1.9276236251246655e-06,
"loss": 0.0028,
"step": 139
},
{
"epoch": 4.837209302325581,
"grad_norm": 0.5720587372779846,
"learning_rate": 1.8158425248197931e-06,
"loss": 0.0022,
"step": 140
},
{
"epoch": 4.872093023255814,
"grad_norm": 0.16564592719078064,
"learning_rate": 1.7055214510452462e-06,
"loss": 0.0019,
"step": 141
},
{
"epoch": 4.906976744186046,
"grad_norm": 0.12440971285104752,
"learning_rate": 1.5968958345321178e-06,
"loss": 0.0017,
"step": 142
},
{
"epoch": 4.941860465116279,
"grad_norm": 0.08577684313058853,
"learning_rate": 1.490197487820263e-06,
"loss": 0.0017,
"step": 143
},
{
"epoch": 4.976744186046512,
"grad_norm": 0.10730113834142685,
"learning_rate": 1.3856541105586545e-06,
"loss": 0.0016,
"step": 144
},
{
"epoch": 5.0,
"grad_norm": 0.11632312089204788,
"learning_rate": 1.2834888035828597e-06,
"loss": 0.002,
"step": 145
},
{
"epoch": 5.034883720930233,
"grad_norm": 0.11552423238754272,
"learning_rate": 1.1839195928066101e-06,
"loss": 0.0021,
"step": 146
},
{
"epoch": 5.069767441860465,
"grad_norm": 0.10733441263437271,
"learning_rate": 1.0871589639435204e-06,
"loss": 0.0016,
"step": 147
},
{
"epoch": 5.104651162790698,
"grad_norm": 0.06744810193777084,
"learning_rate": 9.934134090518593e-07,
"loss": 0.0017,
"step": 148
},
{
"epoch": 5.1395348837209305,
"grad_norm": 0.0750541016459465,
"learning_rate": 9.028829858700974e-07,
"loss": 0.0015,
"step": 149
},
{
"epoch": 5.174418604651163,
"grad_norm": 0.04486292973160744,
"learning_rate": 8.157608908836071e-07,
"loss": 0.0011,
"step": 150
},
{
"epoch": 5.209302325581396,
"grad_norm": 0.060016702860593796,
"learning_rate": 7.322330470336314e-07,
"loss": 0.0012,
"step": 151
},
{
"epoch": 5.2441860465116275,
"grad_norm": 0.056270696222782135,
"learning_rate": 6.524777069483526e-07,
"loss": 0.0009,
"step": 152
},
{
"epoch": 5.27906976744186,
"grad_norm": 0.06919720023870468,
"learning_rate": 5.766650725428027e-07,
"loss": 0.0023,
"step": 153
},
{
"epoch": 5.313953488372093,
"grad_norm": 0.10715026408433914,
"learning_rate": 5.049569317994013e-07,
"loss": 0.0016,
"step": 154
},
{
"epoch": 5.348837209302325,
"grad_norm": 0.07221261411905289,
"learning_rate": 4.3750631350424456e-07,
"loss": 0.0019,
"step": 155
},
{
"epoch": 5.383720930232558,
"grad_norm": 0.16049650311470032,
"learning_rate": 3.7445716067596506e-07,
"loss": 0.0022,
"step": 156
},
{
"epoch": 5.4186046511627906,
"grad_norm": 0.06509647518396378,
"learning_rate": 3.1594402338407633e-07,
"loss": 0.0017,
"step": 157
},
{
"epoch": 5.453488372093023,
"grad_norm": 0.22881199419498444,
"learning_rate": 2.620917716123444e-07,
"loss": 0.002,
"step": 158
},
{
"epoch": 5.488372093023256,
"grad_norm": 0.08407599478960037,
"learning_rate": 2.1301532877994747e-07,
"loss": 0.0012,
"step": 159
},
{
"epoch": 5.523255813953488,
"grad_norm": 0.048866819590330124,
"learning_rate": 1.6881942648911077e-07,
"loss": 0.0012,
"step": 160
},
{
"epoch": 5.558139534883721,
"grad_norm": 0.09164915978908539,
"learning_rate": 1.2959838102258537e-07,
"loss": 0.0023,
"step": 161
},
{
"epoch": 5.593023255813954,
"grad_norm": 0.07083190977573395,
"learning_rate": 9.54358920679524e-08,
"loss": 0.0009,
"step": 162
},
{
"epoch": 5.627906976744186,
"grad_norm": 0.054843056946992874,
"learning_rate": 6.640486409826785e-08,
"loss": 0.0007,
"step": 163
},
{
"epoch": 5.662790697674419,
"grad_norm": 0.06817606836557388,
"learning_rate": 4.256725079024554e-08,
"loss": 0.0015,
"step": 164
},
{
"epoch": 5.6976744186046515,
"grad_norm": 0.049567993730306625,
"learning_rate": 2.3973922811987295e-08,
"loss": 0.0014,
"step": 165
},
{
"epoch": 5.732558139534884,
"grad_norm": 0.07678244262933731,
"learning_rate": 1.0664559262413831e-08,
"loss": 0.0029,
"step": 166
},
{
"epoch": 5.767441860465116,
"grad_norm": 0.13374446332454681,
"learning_rate": 2.6675629940689508e-09,
"loss": 0.0015,
"step": 167
},
{
"epoch": 5.8023255813953485,
"grad_norm": 0.0840042233467102,
"learning_rate": 0.0,
"loss": 0.0019,
"step": 168
}
],
"logging_steps": 1,
"max_steps": 168,
"num_input_tokens_seen": 0,
"num_train_epochs": 6,
"save_steps": 28,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.4267371372895273e+19,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}