733342b0af
Model: longtermrisk/Qwen3-4B-Base-ftjob-6fd14d9c448d-ftjob-adf3bd7963be Source: Original Platform
1061 lines
24 KiB
JSON
1061 lines
24 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 3.0,
|
|
"eval_steps": 500,
|
|
"global_step": 144,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.020942408376963352,
|
|
"grad_norm": 0.421875,
|
|
"learning_rate": 0.0,
|
|
"loss": 1.9647,
|
|
"step": 1
|
|
},
|
|
{
|
|
"epoch": 0.041884816753926704,
|
|
"grad_norm": 0.4296875,
|
|
"learning_rate": 4.000000000000001e-06,
|
|
"loss": 1.9298,
|
|
"step": 2
|
|
},
|
|
{
|
|
"epoch": 0.06282722513089005,
|
|
"grad_norm": 0.412109375,
|
|
"learning_rate": 8.000000000000001e-06,
|
|
"loss": 1.9703,
|
|
"step": 3
|
|
},
|
|
{
|
|
"epoch": 0.08376963350785341,
|
|
"grad_norm": 0.4453125,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 1.9272,
|
|
"step": 4
|
|
},
|
|
{
|
|
"epoch": 0.10471204188481675,
|
|
"grad_norm": 0.412109375,
|
|
"learning_rate": 1.6000000000000003e-05,
|
|
"loss": 1.9523,
|
|
"step": 5
|
|
},
|
|
{
|
|
"epoch": 0.1256544502617801,
|
|
"grad_norm": 0.4140625,
|
|
"learning_rate": 2e-05,
|
|
"loss": 1.9899,
|
|
"step": 6
|
|
},
|
|
{
|
|
"epoch": 0.14659685863874344,
|
|
"grad_norm": 0.41015625,
|
|
"learning_rate": 1.985611510791367e-05,
|
|
"loss": 1.9546,
|
|
"step": 7
|
|
},
|
|
{
|
|
"epoch": 0.16753926701570682,
|
|
"grad_norm": 0.40625,
|
|
"learning_rate": 1.971223021582734e-05,
|
|
"loss": 1.9757,
|
|
"step": 8
|
|
},
|
|
{
|
|
"epoch": 0.18848167539267016,
|
|
"grad_norm": 0.451171875,
|
|
"learning_rate": 1.956834532374101e-05,
|
|
"loss": 2.0231,
|
|
"step": 9
|
|
},
|
|
{
|
|
"epoch": 0.2094240837696335,
|
|
"grad_norm": 0.4375,
|
|
"learning_rate": 1.9424460431654675e-05,
|
|
"loss": 2.0845,
|
|
"step": 10
|
|
},
|
|
{
|
|
"epoch": 0.23036649214659685,
|
|
"grad_norm": 0.419921875,
|
|
"learning_rate": 1.9280575539568347e-05,
|
|
"loss": 1.9888,
|
|
"step": 11
|
|
},
|
|
{
|
|
"epoch": 0.2513089005235602,
|
|
"grad_norm": 0.4140625,
|
|
"learning_rate": 1.9136690647482016e-05,
|
|
"loss": 1.9024,
|
|
"step": 12
|
|
},
|
|
{
|
|
"epoch": 0.27225130890052357,
|
|
"grad_norm": 0.44140625,
|
|
"learning_rate": 1.8992805755395684e-05,
|
|
"loss": 1.9497,
|
|
"step": 13
|
|
},
|
|
{
|
|
"epoch": 0.2931937172774869,
|
|
"grad_norm": 0.408203125,
|
|
"learning_rate": 1.8848920863309356e-05,
|
|
"loss": 2.0388,
|
|
"step": 14
|
|
},
|
|
{
|
|
"epoch": 0.31413612565445026,
|
|
"grad_norm": 0.431640625,
|
|
"learning_rate": 1.870503597122302e-05,
|
|
"loss": 1.9418,
|
|
"step": 15
|
|
},
|
|
{
|
|
"epoch": 0.33507853403141363,
|
|
"grad_norm": 0.4296875,
|
|
"learning_rate": 1.8561151079136693e-05,
|
|
"loss": 1.926,
|
|
"step": 16
|
|
},
|
|
{
|
|
"epoch": 0.35602094240837695,
|
|
"grad_norm": 0.421875,
|
|
"learning_rate": 1.841726618705036e-05,
|
|
"loss": 2.0346,
|
|
"step": 17
|
|
},
|
|
{
|
|
"epoch": 0.3769633507853403,
|
|
"grad_norm": 0.443359375,
|
|
"learning_rate": 1.827338129496403e-05,
|
|
"loss": 2.0832,
|
|
"step": 18
|
|
},
|
|
{
|
|
"epoch": 0.39790575916230364,
|
|
"grad_norm": 0.4296875,
|
|
"learning_rate": 1.8129496402877698e-05,
|
|
"loss": 1.9825,
|
|
"step": 19
|
|
},
|
|
{
|
|
"epoch": 0.418848167539267,
|
|
"grad_norm": 0.439453125,
|
|
"learning_rate": 1.7985611510791367e-05,
|
|
"loss": 2.0244,
|
|
"step": 20
|
|
},
|
|
{
|
|
"epoch": 0.4397905759162304,
|
|
"grad_norm": 0.431640625,
|
|
"learning_rate": 1.784172661870504e-05,
|
|
"loss": 1.9656,
|
|
"step": 21
|
|
},
|
|
{
|
|
"epoch": 0.4607329842931937,
|
|
"grad_norm": 0.451171875,
|
|
"learning_rate": 1.7697841726618707e-05,
|
|
"loss": 1.9011,
|
|
"step": 22
|
|
},
|
|
{
|
|
"epoch": 0.4816753926701571,
|
|
"grad_norm": 0.439453125,
|
|
"learning_rate": 1.7553956834532375e-05,
|
|
"loss": 1.9925,
|
|
"step": 23
|
|
},
|
|
{
|
|
"epoch": 0.5026178010471204,
|
|
"grad_norm": 0.4375,
|
|
"learning_rate": 1.7410071942446044e-05,
|
|
"loss": 2.0238,
|
|
"step": 24
|
|
},
|
|
{
|
|
"epoch": 0.5235602094240838,
|
|
"grad_norm": 0.42578125,
|
|
"learning_rate": 1.7266187050359712e-05,
|
|
"loss": 2.0006,
|
|
"step": 25
|
|
},
|
|
{
|
|
"epoch": 0.5445026178010471,
|
|
"grad_norm": 0.431640625,
|
|
"learning_rate": 1.7122302158273384e-05,
|
|
"loss": 2.0042,
|
|
"step": 26
|
|
},
|
|
{
|
|
"epoch": 0.5654450261780105,
|
|
"grad_norm": 0.431640625,
|
|
"learning_rate": 1.6978417266187053e-05,
|
|
"loss": 1.9056,
|
|
"step": 27
|
|
},
|
|
{
|
|
"epoch": 0.5863874345549738,
|
|
"grad_norm": 0.44921875,
|
|
"learning_rate": 1.683453237410072e-05,
|
|
"loss": 2.0109,
|
|
"step": 28
|
|
},
|
|
{
|
|
"epoch": 0.6073298429319371,
|
|
"grad_norm": 0.451171875,
|
|
"learning_rate": 1.669064748201439e-05,
|
|
"loss": 1.9771,
|
|
"step": 29
|
|
},
|
|
{
|
|
"epoch": 0.6282722513089005,
|
|
"grad_norm": 0.447265625,
|
|
"learning_rate": 1.6546762589928058e-05,
|
|
"loss": 1.9996,
|
|
"step": 30
|
|
},
|
|
{
|
|
"epoch": 0.6492146596858639,
|
|
"grad_norm": 0.443359375,
|
|
"learning_rate": 1.640287769784173e-05,
|
|
"loss": 2.0592,
|
|
"step": 31
|
|
},
|
|
{
|
|
"epoch": 0.6701570680628273,
|
|
"grad_norm": 0.443359375,
|
|
"learning_rate": 1.6258992805755398e-05,
|
|
"loss": 1.9724,
|
|
"step": 32
|
|
},
|
|
{
|
|
"epoch": 0.6910994764397905,
|
|
"grad_norm": 0.45703125,
|
|
"learning_rate": 1.6115107913669067e-05,
|
|
"loss": 2.0015,
|
|
"step": 33
|
|
},
|
|
{
|
|
"epoch": 0.7120418848167539,
|
|
"grad_norm": 0.4453125,
|
|
"learning_rate": 1.5971223021582735e-05,
|
|
"loss": 2.0208,
|
|
"step": 34
|
|
},
|
|
{
|
|
"epoch": 0.7329842931937173,
|
|
"grad_norm": 0.458984375,
|
|
"learning_rate": 1.5827338129496403e-05,
|
|
"loss": 1.9894,
|
|
"step": 35
|
|
},
|
|
{
|
|
"epoch": 0.7539267015706806,
|
|
"grad_norm": 0.43359375,
|
|
"learning_rate": 1.5683453237410072e-05,
|
|
"loss": 1.9313,
|
|
"step": 36
|
|
},
|
|
{
|
|
"epoch": 0.774869109947644,
|
|
"grad_norm": 0.439453125,
|
|
"learning_rate": 1.5539568345323744e-05,
|
|
"loss": 1.8912,
|
|
"step": 37
|
|
},
|
|
{
|
|
"epoch": 0.7958115183246073,
|
|
"grad_norm": 0.44140625,
|
|
"learning_rate": 1.5395683453237412e-05,
|
|
"loss": 1.9814,
|
|
"step": 38
|
|
},
|
|
{
|
|
"epoch": 0.8167539267015707,
|
|
"grad_norm": 0.443359375,
|
|
"learning_rate": 1.525179856115108e-05,
|
|
"loss": 1.9366,
|
|
"step": 39
|
|
},
|
|
{
|
|
"epoch": 0.837696335078534,
|
|
"grad_norm": 0.447265625,
|
|
"learning_rate": 1.5107913669064749e-05,
|
|
"loss": 1.9994,
|
|
"step": 40
|
|
},
|
|
{
|
|
"epoch": 0.8586387434554974,
|
|
"grad_norm": 0.4453125,
|
|
"learning_rate": 1.496402877697842e-05,
|
|
"loss": 1.9949,
|
|
"step": 41
|
|
},
|
|
{
|
|
"epoch": 0.8795811518324608,
|
|
"grad_norm": 0.5,
|
|
"learning_rate": 1.4820143884892086e-05,
|
|
"loss": 2.0622,
|
|
"step": 42
|
|
},
|
|
{
|
|
"epoch": 0.900523560209424,
|
|
"grad_norm": 0.453125,
|
|
"learning_rate": 1.4676258992805756e-05,
|
|
"loss": 1.9898,
|
|
"step": 43
|
|
},
|
|
{
|
|
"epoch": 0.9214659685863874,
|
|
"grad_norm": 0.455078125,
|
|
"learning_rate": 1.4532374100719426e-05,
|
|
"loss": 2.0145,
|
|
"step": 44
|
|
},
|
|
{
|
|
"epoch": 0.9424083769633508,
|
|
"grad_norm": 0.482421875,
|
|
"learning_rate": 1.4388489208633095e-05,
|
|
"loss": 1.9582,
|
|
"step": 45
|
|
},
|
|
{
|
|
"epoch": 0.9633507853403142,
|
|
"grad_norm": 0.470703125,
|
|
"learning_rate": 1.4244604316546765e-05,
|
|
"loss": 2.0705,
|
|
"step": 46
|
|
},
|
|
{
|
|
"epoch": 0.9842931937172775,
|
|
"grad_norm": 0.462890625,
|
|
"learning_rate": 1.4100719424460432e-05,
|
|
"loss": 1.9609,
|
|
"step": 47
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"grad_norm": 0.53125,
|
|
"learning_rate": 1.3956834532374102e-05,
|
|
"loss": 1.9681,
|
|
"step": 48
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"eval_loss": 2.0093743801116943,
|
|
"eval_model_preparation_time": 0.0211,
|
|
"eval_runtime": 2.3083,
|
|
"eval_samples_per_second": 36.823,
|
|
"eval_steps_per_second": 18.628,
|
|
"step": 48
|
|
},
|
|
{
|
|
"epoch": 1.0209424083769634,
|
|
"grad_norm": 0.439453125,
|
|
"learning_rate": 1.3812949640287772e-05,
|
|
"loss": 1.9578,
|
|
"step": 49
|
|
},
|
|
{
|
|
"epoch": 1.0418848167539267,
|
|
"grad_norm": 0.439453125,
|
|
"learning_rate": 1.3669064748201439e-05,
|
|
"loss": 1.9415,
|
|
"step": 50
|
|
},
|
|
{
|
|
"epoch": 1.0628272251308901,
|
|
"grad_norm": 0.453125,
|
|
"learning_rate": 1.3525179856115109e-05,
|
|
"loss": 1.8854,
|
|
"step": 51
|
|
},
|
|
{
|
|
"epoch": 1.0837696335078535,
|
|
"grad_norm": 0.48046875,
|
|
"learning_rate": 1.3381294964028777e-05,
|
|
"loss": 2.002,
|
|
"step": 52
|
|
},
|
|
{
|
|
"epoch": 1.1047120418848166,
|
|
"grad_norm": 0.455078125,
|
|
"learning_rate": 1.3237410071942447e-05,
|
|
"loss": 1.9128,
|
|
"step": 53
|
|
},
|
|
{
|
|
"epoch": 1.12565445026178,
|
|
"grad_norm": 0.46875,
|
|
"learning_rate": 1.3093525179856117e-05,
|
|
"loss": 1.9928,
|
|
"step": 54
|
|
},
|
|
{
|
|
"epoch": 1.1465968586387434,
|
|
"grad_norm": 0.439453125,
|
|
"learning_rate": 1.2949640287769784e-05,
|
|
"loss": 1.9237,
|
|
"step": 55
|
|
},
|
|
{
|
|
"epoch": 1.1675392670157068,
|
|
"grad_norm": 0.47265625,
|
|
"learning_rate": 1.2805755395683454e-05,
|
|
"loss": 1.8692,
|
|
"step": 56
|
|
},
|
|
{
|
|
"epoch": 1.1884816753926701,
|
|
"grad_norm": 0.5078125,
|
|
"learning_rate": 1.2661870503597123e-05,
|
|
"loss": 2.0276,
|
|
"step": 57
|
|
},
|
|
{
|
|
"epoch": 1.2094240837696335,
|
|
"grad_norm": 0.455078125,
|
|
"learning_rate": 1.2517985611510793e-05,
|
|
"loss": 2.007,
|
|
"step": 58
|
|
},
|
|
{
|
|
"epoch": 1.2303664921465969,
|
|
"grad_norm": 0.482421875,
|
|
"learning_rate": 1.2374100719424463e-05,
|
|
"loss": 1.8595,
|
|
"step": 59
|
|
},
|
|
{
|
|
"epoch": 1.2513089005235603,
|
|
"grad_norm": 0.453125,
|
|
"learning_rate": 1.223021582733813e-05,
|
|
"loss": 1.9615,
|
|
"step": 60
|
|
},
|
|
{
|
|
"epoch": 1.2722513089005236,
|
|
"grad_norm": 0.48828125,
|
|
"learning_rate": 1.20863309352518e-05,
|
|
"loss": 1.9661,
|
|
"step": 61
|
|
},
|
|
{
|
|
"epoch": 1.2931937172774868,
|
|
"grad_norm": 0.47265625,
|
|
"learning_rate": 1.1942446043165468e-05,
|
|
"loss": 1.9818,
|
|
"step": 62
|
|
},
|
|
{
|
|
"epoch": 1.3141361256544504,
|
|
"grad_norm": 0.46875,
|
|
"learning_rate": 1.1798561151079137e-05,
|
|
"loss": 1.8661,
|
|
"step": 63
|
|
},
|
|
{
|
|
"epoch": 1.3350785340314135,
|
|
"grad_norm": 0.5078125,
|
|
"learning_rate": 1.1654676258992807e-05,
|
|
"loss": 1.9231,
|
|
"step": 64
|
|
},
|
|
{
|
|
"epoch": 1.356020942408377,
|
|
"grad_norm": 0.49609375,
|
|
"learning_rate": 1.1510791366906475e-05,
|
|
"loss": 1.9471,
|
|
"step": 65
|
|
},
|
|
{
|
|
"epoch": 1.3769633507853403,
|
|
"grad_norm": 0.484375,
|
|
"learning_rate": 1.1366906474820146e-05,
|
|
"loss": 1.9238,
|
|
"step": 66
|
|
},
|
|
{
|
|
"epoch": 1.3979057591623036,
|
|
"grad_norm": 0.484375,
|
|
"learning_rate": 1.1223021582733812e-05,
|
|
"loss": 1.9119,
|
|
"step": 67
|
|
},
|
|
{
|
|
"epoch": 1.418848167539267,
|
|
"grad_norm": 0.48046875,
|
|
"learning_rate": 1.1079136690647482e-05,
|
|
"loss": 1.9054,
|
|
"step": 68
|
|
},
|
|
{
|
|
"epoch": 1.4397905759162304,
|
|
"grad_norm": 0.466796875,
|
|
"learning_rate": 1.0935251798561153e-05,
|
|
"loss": 1.9729,
|
|
"step": 69
|
|
},
|
|
{
|
|
"epoch": 1.4607329842931938,
|
|
"grad_norm": 0.478515625,
|
|
"learning_rate": 1.0791366906474821e-05,
|
|
"loss": 1.9446,
|
|
"step": 70
|
|
},
|
|
{
|
|
"epoch": 1.4816753926701571,
|
|
"grad_norm": 0.474609375,
|
|
"learning_rate": 1.0647482014388491e-05,
|
|
"loss": 1.9389,
|
|
"step": 71
|
|
},
|
|
{
|
|
"epoch": 1.5026178010471205,
|
|
"grad_norm": 0.49609375,
|
|
"learning_rate": 1.0503597122302158e-05,
|
|
"loss": 1.9624,
|
|
"step": 72
|
|
},
|
|
{
|
|
"epoch": 1.5235602094240837,
|
|
"grad_norm": 0.48828125,
|
|
"learning_rate": 1.0359712230215828e-05,
|
|
"loss": 1.8873,
|
|
"step": 73
|
|
},
|
|
{
|
|
"epoch": 1.5445026178010473,
|
|
"grad_norm": 0.50390625,
|
|
"learning_rate": 1.0215827338129498e-05,
|
|
"loss": 1.9917,
|
|
"step": 74
|
|
},
|
|
{
|
|
"epoch": 1.5654450261780104,
|
|
"grad_norm": 0.47265625,
|
|
"learning_rate": 1.0071942446043167e-05,
|
|
"loss": 1.862,
|
|
"step": 75
|
|
},
|
|
{
|
|
"epoch": 1.5863874345549738,
|
|
"grad_norm": 0.494140625,
|
|
"learning_rate": 9.928057553956835e-06,
|
|
"loss": 2.0042,
|
|
"step": 76
|
|
},
|
|
{
|
|
"epoch": 1.6073298429319371,
|
|
"grad_norm": 0.48828125,
|
|
"learning_rate": 9.784172661870505e-06,
|
|
"loss": 1.9189,
|
|
"step": 77
|
|
},
|
|
{
|
|
"epoch": 1.6282722513089005,
|
|
"grad_norm": 0.55859375,
|
|
"learning_rate": 9.640287769784174e-06,
|
|
"loss": 1.9742,
|
|
"step": 78
|
|
},
|
|
{
|
|
"epoch": 1.649214659685864,
|
|
"grad_norm": 0.515625,
|
|
"learning_rate": 9.496402877697842e-06,
|
|
"loss": 1.8749,
|
|
"step": 79
|
|
},
|
|
{
|
|
"epoch": 1.6701570680628273,
|
|
"grad_norm": 0.494140625,
|
|
"learning_rate": 9.35251798561151e-06,
|
|
"loss": 1.9488,
|
|
"step": 80
|
|
},
|
|
{
|
|
"epoch": 1.6910994764397906,
|
|
"grad_norm": 0.47265625,
|
|
"learning_rate": 9.20863309352518e-06,
|
|
"loss": 1.881,
|
|
"step": 81
|
|
},
|
|
{
|
|
"epoch": 1.7120418848167538,
|
|
"grad_norm": 0.490234375,
|
|
"learning_rate": 9.064748201438849e-06,
|
|
"loss": 1.9373,
|
|
"step": 82
|
|
},
|
|
{
|
|
"epoch": 1.7329842931937174,
|
|
"grad_norm": 0.49609375,
|
|
"learning_rate": 8.92086330935252e-06,
|
|
"loss": 1.8781,
|
|
"step": 83
|
|
},
|
|
{
|
|
"epoch": 1.7539267015706805,
|
|
"grad_norm": 0.5078125,
|
|
"learning_rate": 8.776978417266188e-06,
|
|
"loss": 1.9445,
|
|
"step": 84
|
|
},
|
|
{
|
|
"epoch": 1.7748691099476441,
|
|
"grad_norm": 0.484375,
|
|
"learning_rate": 8.633093525179856e-06,
|
|
"loss": 1.9793,
|
|
"step": 85
|
|
},
|
|
{
|
|
"epoch": 1.7958115183246073,
|
|
"grad_norm": 0.5078125,
|
|
"learning_rate": 8.489208633093526e-06,
|
|
"loss": 1.9069,
|
|
"step": 86
|
|
},
|
|
{
|
|
"epoch": 1.8167539267015707,
|
|
"grad_norm": 0.5078125,
|
|
"learning_rate": 8.345323741007195e-06,
|
|
"loss": 1.9321,
|
|
"step": 87
|
|
},
|
|
{
|
|
"epoch": 1.837696335078534,
|
|
"grad_norm": 0.474609375,
|
|
"learning_rate": 8.201438848920865e-06,
|
|
"loss": 1.8512,
|
|
"step": 88
|
|
},
|
|
{
|
|
"epoch": 1.8586387434554974,
|
|
"grad_norm": 0.51171875,
|
|
"learning_rate": 8.057553956834533e-06,
|
|
"loss": 1.9338,
|
|
"step": 89
|
|
},
|
|
{
|
|
"epoch": 1.8795811518324608,
|
|
"grad_norm": 0.498046875,
|
|
"learning_rate": 7.913669064748202e-06,
|
|
"loss": 1.9625,
|
|
"step": 90
|
|
},
|
|
{
|
|
"epoch": 1.900523560209424,
|
|
"grad_norm": 0.51171875,
|
|
"learning_rate": 7.769784172661872e-06,
|
|
"loss": 1.9461,
|
|
"step": 91
|
|
},
|
|
{
|
|
"epoch": 1.9214659685863875,
|
|
"grad_norm": 0.50390625,
|
|
"learning_rate": 7.62589928057554e-06,
|
|
"loss": 1.9645,
|
|
"step": 92
|
|
},
|
|
{
|
|
"epoch": 1.9424083769633507,
|
|
"grad_norm": 0.490234375,
|
|
"learning_rate": 7.48201438848921e-06,
|
|
"loss": 1.9475,
|
|
"step": 93
|
|
},
|
|
{
|
|
"epoch": 1.9633507853403143,
|
|
"grad_norm": 0.49609375,
|
|
"learning_rate": 7.338129496402878e-06,
|
|
"loss": 1.8948,
|
|
"step": 94
|
|
},
|
|
{
|
|
"epoch": 1.9842931937172774,
|
|
"grad_norm": 0.5,
|
|
"learning_rate": 7.194244604316547e-06,
|
|
"loss": 1.9867,
|
|
"step": 95
|
|
},
|
|
{
|
|
"epoch": 2.0,
|
|
"grad_norm": 0.6015625,
|
|
"learning_rate": 7.050359712230216e-06,
|
|
"loss": 1.9697,
|
|
"step": 96
|
|
},
|
|
{
|
|
"epoch": 2.0,
|
|
"eval_loss": 1.9895108938217163,
|
|
"eval_model_preparation_time": 0.0211,
|
|
"eval_runtime": 2.5043,
|
|
"eval_samples_per_second": 33.942,
|
|
"eval_steps_per_second": 17.171,
|
|
"step": 96
|
|
},
|
|
{
|
|
"epoch": 2.020942408376963,
|
|
"grad_norm": 0.515625,
|
|
"learning_rate": 6.906474820143886e-06,
|
|
"loss": 1.9634,
|
|
"step": 97
|
|
},
|
|
{
|
|
"epoch": 2.0418848167539267,
|
|
"grad_norm": 0.48046875,
|
|
"learning_rate": 6.762589928057554e-06,
|
|
"loss": 1.9503,
|
|
"step": 98
|
|
},
|
|
{
|
|
"epoch": 2.06282722513089,
|
|
"grad_norm": 0.494140625,
|
|
"learning_rate": 6.618705035971224e-06,
|
|
"loss": 1.9514,
|
|
"step": 99
|
|
},
|
|
{
|
|
"epoch": 2.0837696335078535,
|
|
"grad_norm": 0.515625,
|
|
"learning_rate": 6.474820143884892e-06,
|
|
"loss": 1.8892,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 2.1047120418848166,
|
|
"grad_norm": 0.48828125,
|
|
"learning_rate": 6.330935251798561e-06,
|
|
"loss": 1.94,
|
|
"step": 101
|
|
},
|
|
{
|
|
"epoch": 2.1256544502617802,
|
|
"grad_norm": 0.490234375,
|
|
"learning_rate": 6.1870503597122315e-06,
|
|
"loss": 1.8321,
|
|
"step": 102
|
|
},
|
|
{
|
|
"epoch": 2.1465968586387434,
|
|
"grad_norm": 0.48828125,
|
|
"learning_rate": 6.0431654676259e-06,
|
|
"loss": 1.9178,
|
|
"step": 103
|
|
},
|
|
{
|
|
"epoch": 2.167539267015707,
|
|
"grad_norm": 0.498046875,
|
|
"learning_rate": 5.899280575539568e-06,
|
|
"loss": 1.9368,
|
|
"step": 104
|
|
},
|
|
{
|
|
"epoch": 2.18848167539267,
|
|
"grad_norm": 0.50390625,
|
|
"learning_rate": 5.755395683453238e-06,
|
|
"loss": 1.9388,
|
|
"step": 105
|
|
},
|
|
{
|
|
"epoch": 2.2094240837696333,
|
|
"grad_norm": 0.498046875,
|
|
"learning_rate": 5.611510791366906e-06,
|
|
"loss": 1.8343,
|
|
"step": 106
|
|
},
|
|
{
|
|
"epoch": 2.230366492146597,
|
|
"grad_norm": 0.5,
|
|
"learning_rate": 5.467625899280576e-06,
|
|
"loss": 1.9223,
|
|
"step": 107
|
|
},
|
|
{
|
|
"epoch": 2.25130890052356,
|
|
"grad_norm": 0.49609375,
|
|
"learning_rate": 5.3237410071942456e-06,
|
|
"loss": 1.8655,
|
|
"step": 108
|
|
},
|
|
{
|
|
"epoch": 2.2722513089005236,
|
|
"grad_norm": 0.490234375,
|
|
"learning_rate": 5.179856115107914e-06,
|
|
"loss": 1.8889,
|
|
"step": 109
|
|
},
|
|
{
|
|
"epoch": 2.2931937172774868,
|
|
"grad_norm": 0.5234375,
|
|
"learning_rate": 5.035971223021583e-06,
|
|
"loss": 1.8139,
|
|
"step": 110
|
|
},
|
|
{
|
|
"epoch": 2.3141361256544504,
|
|
"grad_norm": 0.57421875,
|
|
"learning_rate": 4.892086330935253e-06,
|
|
"loss": 1.9506,
|
|
"step": 111
|
|
},
|
|
{
|
|
"epoch": 2.3350785340314135,
|
|
"grad_norm": 0.50390625,
|
|
"learning_rate": 4.748201438848921e-06,
|
|
"loss": 1.8646,
|
|
"step": 112
|
|
},
|
|
{
|
|
"epoch": 2.356020942408377,
|
|
"grad_norm": 0.51171875,
|
|
"learning_rate": 4.60431654676259e-06,
|
|
"loss": 1.8975,
|
|
"step": 113
|
|
},
|
|
{
|
|
"epoch": 2.3769633507853403,
|
|
"grad_norm": 0.51171875,
|
|
"learning_rate": 4.46043165467626e-06,
|
|
"loss": 1.9361,
|
|
"step": 114
|
|
},
|
|
{
|
|
"epoch": 2.3979057591623034,
|
|
"grad_norm": 0.515625,
|
|
"learning_rate": 4.316546762589928e-06,
|
|
"loss": 1.9321,
|
|
"step": 115
|
|
},
|
|
{
|
|
"epoch": 2.418848167539267,
|
|
"grad_norm": 0.54296875,
|
|
"learning_rate": 4.172661870503597e-06,
|
|
"loss": 1.9547,
|
|
"step": 116
|
|
},
|
|
{
|
|
"epoch": 2.4397905759162306,
|
|
"grad_norm": 0.5078125,
|
|
"learning_rate": 4.028776978417267e-06,
|
|
"loss": 1.9316,
|
|
"step": 117
|
|
},
|
|
{
|
|
"epoch": 2.4607329842931938,
|
|
"grad_norm": 0.494140625,
|
|
"learning_rate": 3.884892086330936e-06,
|
|
"loss": 1.9229,
|
|
"step": 118
|
|
},
|
|
{
|
|
"epoch": 2.481675392670157,
|
|
"grad_norm": 0.5078125,
|
|
"learning_rate": 3.741007194244605e-06,
|
|
"loss": 1.8711,
|
|
"step": 119
|
|
},
|
|
{
|
|
"epoch": 2.5026178010471205,
|
|
"grad_norm": 0.55078125,
|
|
"learning_rate": 3.5971223021582737e-06,
|
|
"loss": 2.0226,
|
|
"step": 120
|
|
},
|
|
{
|
|
"epoch": 2.5235602094240837,
|
|
"grad_norm": 0.484375,
|
|
"learning_rate": 3.453237410071943e-06,
|
|
"loss": 1.8823,
|
|
"step": 121
|
|
},
|
|
{
|
|
"epoch": 2.5445026178010473,
|
|
"grad_norm": 0.515625,
|
|
"learning_rate": 3.309352517985612e-06,
|
|
"loss": 1.9203,
|
|
"step": 122
|
|
},
|
|
{
|
|
"epoch": 2.5654450261780104,
|
|
"grad_norm": 0.5078125,
|
|
"learning_rate": 3.1654676258992807e-06,
|
|
"loss": 1.873,
|
|
"step": 123
|
|
},
|
|
{
|
|
"epoch": 2.5863874345549736,
|
|
"grad_norm": 0.5234375,
|
|
"learning_rate": 3.02158273381295e-06,
|
|
"loss": 1.9312,
|
|
"step": 124
|
|
},
|
|
{
|
|
"epoch": 2.607329842931937,
|
|
"grad_norm": 0.5234375,
|
|
"learning_rate": 2.877697841726619e-06,
|
|
"loss": 2.0012,
|
|
"step": 125
|
|
},
|
|
{
|
|
"epoch": 2.6282722513089007,
|
|
"grad_norm": 0.515625,
|
|
"learning_rate": 2.733812949640288e-06,
|
|
"loss": 1.9038,
|
|
"step": 126
|
|
},
|
|
{
|
|
"epoch": 2.649214659685864,
|
|
"grad_norm": 0.515625,
|
|
"learning_rate": 2.589928057553957e-06,
|
|
"loss": 1.891,
|
|
"step": 127
|
|
},
|
|
{
|
|
"epoch": 2.670157068062827,
|
|
"grad_norm": 0.5078125,
|
|
"learning_rate": 2.4460431654676263e-06,
|
|
"loss": 1.9418,
|
|
"step": 128
|
|
},
|
|
{
|
|
"epoch": 2.6910994764397906,
|
|
"grad_norm": 0.498046875,
|
|
"learning_rate": 2.302158273381295e-06,
|
|
"loss": 1.8493,
|
|
"step": 129
|
|
},
|
|
{
|
|
"epoch": 2.712041884816754,
|
|
"grad_norm": 0.490234375,
|
|
"learning_rate": 2.158273381294964e-06,
|
|
"loss": 1.8732,
|
|
"step": 130
|
|
},
|
|
{
|
|
"epoch": 2.7329842931937174,
|
|
"grad_norm": 0.5234375,
|
|
"learning_rate": 2.0143884892086333e-06,
|
|
"loss": 1.8554,
|
|
"step": 131
|
|
},
|
|
{
|
|
"epoch": 2.7539267015706805,
|
|
"grad_norm": 0.53515625,
|
|
"learning_rate": 1.8705035971223024e-06,
|
|
"loss": 1.9015,
|
|
"step": 132
|
|
},
|
|
{
|
|
"epoch": 2.774869109947644,
|
|
"grad_norm": 0.50390625,
|
|
"learning_rate": 1.7266187050359715e-06,
|
|
"loss": 1.9164,
|
|
"step": 133
|
|
},
|
|
{
|
|
"epoch": 2.7958115183246073,
|
|
"grad_norm": 0.5078125,
|
|
"learning_rate": 1.5827338129496403e-06,
|
|
"loss": 1.856,
|
|
"step": 134
|
|
},
|
|
{
|
|
"epoch": 2.816753926701571,
|
|
"grad_norm": 0.50390625,
|
|
"learning_rate": 1.4388489208633094e-06,
|
|
"loss": 1.8792,
|
|
"step": 135
|
|
},
|
|
{
|
|
"epoch": 2.837696335078534,
|
|
"grad_norm": 0.490234375,
|
|
"learning_rate": 1.2949640287769785e-06,
|
|
"loss": 1.8862,
|
|
"step": 136
|
|
},
|
|
{
|
|
"epoch": 2.858638743455497,
|
|
"grad_norm": 0.51171875,
|
|
"learning_rate": 1.1510791366906476e-06,
|
|
"loss": 1.936,
|
|
"step": 137
|
|
},
|
|
{
|
|
"epoch": 2.8795811518324608,
|
|
"grad_norm": 0.515625,
|
|
"learning_rate": 1.0071942446043167e-06,
|
|
"loss": 1.8862,
|
|
"step": 138
|
|
},
|
|
{
|
|
"epoch": 2.900523560209424,
|
|
"grad_norm": 0.51171875,
|
|
"learning_rate": 8.633093525179857e-07,
|
|
"loss": 1.9041,
|
|
"step": 139
|
|
},
|
|
{
|
|
"epoch": 2.9214659685863875,
|
|
"grad_norm": 0.486328125,
|
|
"learning_rate": 7.194244604316547e-07,
|
|
"loss": 1.86,
|
|
"step": 140
|
|
},
|
|
{
|
|
"epoch": 2.9424083769633507,
|
|
"grad_norm": 0.5234375,
|
|
"learning_rate": 5.755395683453238e-07,
|
|
"loss": 1.9579,
|
|
"step": 141
|
|
},
|
|
{
|
|
"epoch": 2.9633507853403143,
|
|
"grad_norm": 0.5234375,
|
|
"learning_rate": 4.3165467625899287e-07,
|
|
"loss": 1.9625,
|
|
"step": 142
|
|
},
|
|
{
|
|
"epoch": 2.9842931937172774,
|
|
"grad_norm": 0.5,
|
|
"learning_rate": 2.877697841726619e-07,
|
|
"loss": 1.9296,
|
|
"step": 143
|
|
},
|
|
{
|
|
"epoch": 3.0,
|
|
"grad_norm": 0.5859375,
|
|
"learning_rate": 1.4388489208633095e-07,
|
|
"loss": 1.9113,
|
|
"step": 144
|
|
}
|
|
],
|
|
"logging_steps": 1,
|
|
"max_steps": 144,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 3,
|
|
"save_steps": 5000,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": true
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 4.364548133830656e+16,
|
|
"train_batch_size": 2,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|