codegen1-6B-ds-zero3 / trainer_state.json
0xk1h0's picture
Upload folder using huggingface_hub
0655cdb
raw
history blame
No virus
17.1 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"global_step": 142,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 0,
"loss": 3.8301,
"step": 1
},
{
"epoch": 0.01,
"learning_rate": 0,
"loss": 3.648,
"step": 2
},
{
"epoch": 0.02,
"learning_rate": 0,
"loss": 3.6693,
"step": 3
},
{
"epoch": 0.03,
"learning_rate": 0,
"loss": 3.6809,
"step": 4
},
{
"epoch": 0.04,
"learning_rate": 0,
"loss": 3.6776,
"step": 5
},
{
"epoch": 0.04,
"learning_rate": 0.0,
"loss": 3.6498,
"step": 6
},
{
"epoch": 0.05,
"learning_rate": 4.306765580733931e-06,
"loss": 2.0524,
"step": 7
},
{
"epoch": 0.06,
"learning_rate": 6.826061944859854e-06,
"loss": 1.9607,
"step": 8
},
{
"epoch": 0.06,
"learning_rate": 8.613531161467863e-06,
"loss": 0.8966,
"step": 9
},
{
"epoch": 0.07,
"learning_rate": 1e-05,
"loss": 0.292,
"step": 10
},
{
"epoch": 0.08,
"learning_rate": 1e-05,
"loss": 0.2447,
"step": 11
},
{
"epoch": 0.08,
"learning_rate": 9.97624703087886e-06,
"loss": 0.1913,
"step": 12
},
{
"epoch": 0.09,
"learning_rate": 9.95249406175772e-06,
"loss": 0.1367,
"step": 13
},
{
"epoch": 0.1,
"learning_rate": 9.92874109263658e-06,
"loss": 0.1032,
"step": 14
},
{
"epoch": 0.11,
"learning_rate": 9.90498812351544e-06,
"loss": 0.0797,
"step": 15
},
{
"epoch": 0.11,
"learning_rate": 9.8812351543943e-06,
"loss": 0.0641,
"step": 16
},
{
"epoch": 0.12,
"learning_rate": 9.857482185273159e-06,
"loss": 0.0501,
"step": 17
},
{
"epoch": 0.13,
"learning_rate": 9.83372921615202e-06,
"loss": 0.0494,
"step": 18
},
{
"epoch": 0.13,
"learning_rate": 9.80997624703088e-06,
"loss": 0.0497,
"step": 19
},
{
"epoch": 0.14,
"learning_rate": 9.78622327790974e-06,
"loss": 0.0453,
"step": 20
},
{
"epoch": 0.15,
"learning_rate": 9.7624703087886e-06,
"loss": 0.0384,
"step": 21
},
{
"epoch": 0.15,
"learning_rate": 9.73871733966746e-06,
"loss": 0.0332,
"step": 22
},
{
"epoch": 0.16,
"learning_rate": 9.714964370546319e-06,
"loss": 0.0328,
"step": 23
},
{
"epoch": 0.17,
"learning_rate": 9.69121140142518e-06,
"loss": 0.0335,
"step": 24
},
{
"epoch": 0.18,
"learning_rate": 9.66745843230404e-06,
"loss": 0.0337,
"step": 25
},
{
"epoch": 0.18,
"learning_rate": 9.643705463182899e-06,
"loss": 0.0335,
"step": 26
},
{
"epoch": 0.19,
"learning_rate": 9.619952494061758e-06,
"loss": 0.0348,
"step": 27
},
{
"epoch": 0.2,
"learning_rate": 9.596199524940617e-06,
"loss": 0.0318,
"step": 28
},
{
"epoch": 0.2,
"learning_rate": 9.572446555819479e-06,
"loss": 0.0309,
"step": 29
},
{
"epoch": 0.21,
"learning_rate": 9.548693586698338e-06,
"loss": 0.0303,
"step": 30
},
{
"epoch": 0.22,
"learning_rate": 9.524940617577197e-06,
"loss": 0.032,
"step": 31
},
{
"epoch": 0.23,
"learning_rate": 9.501187648456057e-06,
"loss": 0.0303,
"step": 32
},
{
"epoch": 0.23,
"learning_rate": 9.477434679334918e-06,
"loss": 0.0294,
"step": 33
},
{
"epoch": 0.24,
"learning_rate": 9.453681710213777e-06,
"loss": 0.0288,
"step": 34
},
{
"epoch": 0.25,
"learning_rate": 9.429928741092638e-06,
"loss": 0.0288,
"step": 35
},
{
"epoch": 0.25,
"learning_rate": 9.406175771971498e-06,
"loss": 0.0301,
"step": 36
},
{
"epoch": 0.26,
"learning_rate": 9.382422802850357e-06,
"loss": 0.0314,
"step": 37
},
{
"epoch": 0.27,
"learning_rate": 9.358669833729217e-06,
"loss": 0.0288,
"step": 38
},
{
"epoch": 0.27,
"learning_rate": 9.334916864608076e-06,
"loss": 0.0287,
"step": 39
},
{
"epoch": 0.28,
"learning_rate": 9.311163895486937e-06,
"loss": 0.0299,
"step": 40
},
{
"epoch": 0.29,
"learning_rate": 9.287410926365797e-06,
"loss": 0.0288,
"step": 41
},
{
"epoch": 0.3,
"learning_rate": 9.263657957244656e-06,
"loss": 0.0302,
"step": 42
},
{
"epoch": 0.3,
"learning_rate": 9.239904988123515e-06,
"loss": 0.0277,
"step": 43
},
{
"epoch": 0.31,
"learning_rate": 9.216152019002376e-06,
"loss": 0.0279,
"step": 44
},
{
"epoch": 0.32,
"learning_rate": 9.192399049881236e-06,
"loss": 0.0285,
"step": 45
},
{
"epoch": 0.32,
"learning_rate": 9.168646080760095e-06,
"loss": 0.0299,
"step": 46
},
{
"epoch": 0.33,
"learning_rate": 9.144893111638956e-06,
"loss": 0.029,
"step": 47
},
{
"epoch": 0.34,
"learning_rate": 9.121140142517816e-06,
"loss": 0.0293,
"step": 48
},
{
"epoch": 0.35,
"learning_rate": 9.097387173396675e-06,
"loss": 0.0293,
"step": 49
},
{
"epoch": 0.35,
"learning_rate": 9.073634204275536e-06,
"loss": 0.029,
"step": 50
},
{
"epoch": 0.36,
"learning_rate": 9.049881235154396e-06,
"loss": 0.0295,
"step": 51
},
{
"epoch": 0.37,
"learning_rate": 9.026128266033255e-06,
"loss": 0.0284,
"step": 52
},
{
"epoch": 0.37,
"learning_rate": 9.002375296912114e-06,
"loss": 0.0298,
"step": 53
},
{
"epoch": 0.38,
"learning_rate": 8.978622327790974e-06,
"loss": 0.0286,
"step": 54
},
{
"epoch": 0.39,
"learning_rate": 8.954869358669835e-06,
"loss": 0.0292,
"step": 55
},
{
"epoch": 0.39,
"learning_rate": 8.931116389548694e-06,
"loss": 0.0283,
"step": 56
},
{
"epoch": 0.4,
"learning_rate": 8.907363420427554e-06,
"loss": 0.0286,
"step": 57
},
{
"epoch": 0.41,
"learning_rate": 8.883610451306413e-06,
"loss": 0.029,
"step": 58
},
{
"epoch": 0.42,
"learning_rate": 8.859857482185273e-06,
"loss": 0.0282,
"step": 59
},
{
"epoch": 0.42,
"learning_rate": 8.836104513064134e-06,
"loss": 0.0281,
"step": 60
},
{
"epoch": 0.43,
"learning_rate": 8.812351543942995e-06,
"loss": 0.0281,
"step": 61
},
{
"epoch": 0.44,
"learning_rate": 8.788598574821854e-06,
"loss": 0.028,
"step": 62
},
{
"epoch": 0.44,
"learning_rate": 8.764845605700714e-06,
"loss": 0.0286,
"step": 63
},
{
"epoch": 0.45,
"learning_rate": 8.741092636579573e-06,
"loss": 0.0293,
"step": 64
},
{
"epoch": 0.46,
"learning_rate": 8.717339667458432e-06,
"loss": 0.0285,
"step": 65
},
{
"epoch": 0.46,
"learning_rate": 8.693586698337293e-06,
"loss": 0.0281,
"step": 66
},
{
"epoch": 0.47,
"learning_rate": 8.669833729216153e-06,
"loss": 0.0284,
"step": 67
},
{
"epoch": 0.48,
"learning_rate": 8.646080760095012e-06,
"loss": 0.0288,
"step": 68
},
{
"epoch": 0.49,
"learning_rate": 8.622327790973872e-06,
"loss": 0.0288,
"step": 69
},
{
"epoch": 0.49,
"learning_rate": 8.598574821852733e-06,
"loss": 0.0283,
"step": 70
},
{
"epoch": 0.5,
"learning_rate": 8.574821852731592e-06,
"loss": 0.0278,
"step": 71
},
{
"epoch": 0.51,
"learning_rate": 8.551068883610452e-06,
"loss": 0.0281,
"step": 72
},
{
"epoch": 0.51,
"learning_rate": 8.527315914489311e-06,
"loss": 0.0276,
"step": 73
},
{
"epoch": 0.52,
"learning_rate": 8.50356294536817e-06,
"loss": 0.0282,
"step": 74
},
{
"epoch": 0.53,
"learning_rate": 8.479809976247032e-06,
"loss": 0.0281,
"step": 75
},
{
"epoch": 0.54,
"learning_rate": 8.456057007125893e-06,
"loss": 0.0282,
"step": 76
},
{
"epoch": 0.54,
"learning_rate": 8.432304038004752e-06,
"loss": 0.0282,
"step": 77
},
{
"epoch": 0.55,
"learning_rate": 8.408551068883611e-06,
"loss": 0.0281,
"step": 78
},
{
"epoch": 0.56,
"learning_rate": 8.38479809976247e-06,
"loss": 0.0282,
"step": 79
},
{
"epoch": 0.56,
"learning_rate": 8.36104513064133e-06,
"loss": 0.0293,
"step": 80
},
{
"epoch": 0.57,
"learning_rate": 8.337292161520191e-06,
"loss": 0.0284,
"step": 81
},
{
"epoch": 0.58,
"learning_rate": 8.31353919239905e-06,
"loss": 0.028,
"step": 82
},
{
"epoch": 0.58,
"learning_rate": 8.28978622327791e-06,
"loss": 0.0281,
"step": 83
},
{
"epoch": 0.59,
"learning_rate": 8.26603325415677e-06,
"loss": 0.0285,
"step": 84
},
{
"epoch": 0.6,
"learning_rate": 8.24228028503563e-06,
"loss": 0.0271,
"step": 85
},
{
"epoch": 0.61,
"learning_rate": 8.21852731591449e-06,
"loss": 0.0284,
"step": 86
},
{
"epoch": 0.61,
"learning_rate": 8.19477434679335e-06,
"loss": 0.0294,
"step": 87
},
{
"epoch": 0.62,
"learning_rate": 8.171021377672209e-06,
"loss": 0.0286,
"step": 88
},
{
"epoch": 0.63,
"learning_rate": 8.14726840855107e-06,
"loss": 0.0276,
"step": 89
},
{
"epoch": 0.63,
"learning_rate": 8.12351543942993e-06,
"loss": 0.028,
"step": 90
},
{
"epoch": 0.64,
"learning_rate": 8.099762470308789e-06,
"loss": 0.0274,
"step": 91
},
{
"epoch": 0.65,
"learning_rate": 8.07600950118765e-06,
"loss": 0.028,
"step": 92
},
{
"epoch": 0.65,
"learning_rate": 8.05225653206651e-06,
"loss": 0.0287,
"step": 93
},
{
"epoch": 0.66,
"learning_rate": 8.028503562945369e-06,
"loss": 0.0297,
"step": 94
},
{
"epoch": 0.67,
"learning_rate": 8.004750593824228e-06,
"loss": 0.028,
"step": 95
},
{
"epoch": 0.68,
"learning_rate": 7.98099762470309e-06,
"loss": 0.0288,
"step": 96
},
{
"epoch": 0.68,
"learning_rate": 7.957244655581949e-06,
"loss": 0.0278,
"step": 97
},
{
"epoch": 0.69,
"learning_rate": 7.933491686460808e-06,
"loss": 0.0281,
"step": 98
},
{
"epoch": 0.7,
"learning_rate": 7.909738717339667e-06,
"loss": 0.0286,
"step": 99
},
{
"epoch": 0.7,
"learning_rate": 7.885985748218527e-06,
"loss": 0.0282,
"step": 100
},
{
"epoch": 0.71,
"learning_rate": 7.862232779097388e-06,
"loss": 0.0279,
"step": 101
},
{
"epoch": 0.72,
"learning_rate": 7.838479809976247e-06,
"loss": 0.0283,
"step": 102
},
{
"epoch": 0.73,
"learning_rate": 7.814726840855108e-06,
"loss": 0.0293,
"step": 103
},
{
"epoch": 0.73,
"learning_rate": 7.790973871733968e-06,
"loss": 0.0287,
"step": 104
},
{
"epoch": 0.74,
"learning_rate": 7.767220902612827e-06,
"loss": 0.0272,
"step": 105
},
{
"epoch": 0.75,
"learning_rate": 7.743467933491687e-06,
"loss": 0.0288,
"step": 106
},
{
"epoch": 0.75,
"learning_rate": 7.719714964370548e-06,
"loss": 0.0292,
"step": 107
},
{
"epoch": 0.76,
"learning_rate": 7.695961995249407e-06,
"loss": 0.0287,
"step": 108
},
{
"epoch": 0.77,
"learning_rate": 7.672209026128267e-06,
"loss": 0.0286,
"step": 109
},
{
"epoch": 0.77,
"learning_rate": 7.648456057007126e-06,
"loss": 0.0293,
"step": 110
},
{
"epoch": 0.78,
"learning_rate": 7.624703087885986e-06,
"loss": 0.0278,
"step": 111
},
{
"epoch": 0.79,
"learning_rate": 7.600950118764846e-06,
"loss": 0.028,
"step": 112
},
{
"epoch": 0.8,
"learning_rate": 7.577197149643706e-06,
"loss": 0.0284,
"step": 113
},
{
"epoch": 0.8,
"learning_rate": 7.553444180522565e-06,
"loss": 0.0272,
"step": 114
},
{
"epoch": 0.81,
"learning_rate": 7.5296912114014255e-06,
"loss": 0.028,
"step": 115
},
{
"epoch": 0.82,
"learning_rate": 7.505938242280285e-06,
"loss": 0.0292,
"step": 116
},
{
"epoch": 0.82,
"learning_rate": 7.482185273159146e-06,
"loss": 0.0273,
"step": 117
},
{
"epoch": 0.83,
"learning_rate": 7.458432304038005e-06,
"loss": 0.0287,
"step": 118
},
{
"epoch": 0.84,
"learning_rate": 7.434679334916866e-06,
"loss": 0.0297,
"step": 119
},
{
"epoch": 0.85,
"learning_rate": 7.410926365795725e-06,
"loss": 0.0299,
"step": 120
},
{
"epoch": 0.85,
"learning_rate": 7.387173396674585e-06,
"loss": 0.0276,
"step": 121
},
{
"epoch": 0.86,
"learning_rate": 7.363420427553445e-06,
"loss": 0.0288,
"step": 122
},
{
"epoch": 0.87,
"learning_rate": 7.339667458432305e-06,
"loss": 0.0276,
"step": 123
},
{
"epoch": 0.87,
"learning_rate": 7.315914489311164e-06,
"loss": 0.0278,
"step": 124
},
{
"epoch": 0.88,
"learning_rate": 7.292161520190024e-06,
"loss": 0.0283,
"step": 125
},
{
"epoch": 0.89,
"learning_rate": 7.268408551068884e-06,
"loss": 0.0282,
"step": 126
},
{
"epoch": 0.89,
"learning_rate": 7.2446555819477435e-06,
"loss": 0.0288,
"step": 127
},
{
"epoch": 0.9,
"learning_rate": 7.220902612826604e-06,
"loss": 0.0289,
"step": 128
},
{
"epoch": 0.91,
"learning_rate": 7.197149643705463e-06,
"loss": 0.0283,
"step": 129
},
{
"epoch": 0.92,
"learning_rate": 7.173396674584323e-06,
"loss": 0.0284,
"step": 130
},
{
"epoch": 0.92,
"learning_rate": 7.149643705463184e-06,
"loss": 0.0291,
"step": 131
},
{
"epoch": 0.93,
"learning_rate": 7.125890736342044e-06,
"loss": 0.0286,
"step": 132
},
{
"epoch": 0.94,
"learning_rate": 7.102137767220903e-06,
"loss": 0.0276,
"step": 133
},
{
"epoch": 0.94,
"learning_rate": 7.0783847980997635e-06,
"loss": 0.0283,
"step": 134
},
{
"epoch": 0.95,
"learning_rate": 7.054631828978623e-06,
"loss": 0.0296,
"step": 135
},
{
"epoch": 0.96,
"learning_rate": 7.030878859857483e-06,
"loss": 0.028,
"step": 136
},
{
"epoch": 0.96,
"learning_rate": 7.007125890736343e-06,
"loss": 0.0289,
"step": 137
},
{
"epoch": 0.97,
"learning_rate": 6.983372921615203e-06,
"loss": 0.0289,
"step": 138
},
{
"epoch": 0.98,
"learning_rate": 6.959619952494062e-06,
"loss": 0.0286,
"step": 139
},
{
"epoch": 0.99,
"learning_rate": 6.935866983372922e-06,
"loss": 0.0276,
"step": 140
},
{
"epoch": 0.99,
"learning_rate": 6.912114014251782e-06,
"loss": 0.0277,
"step": 141
},
{
"epoch": 1.0,
"learning_rate": 6.888361045130641e-06,
"loss": 0.0279,
"step": 142
}
],
"max_steps": 426,
"num_train_epochs": 3,
"total_flos": 86302131093504.0,
"trial_name": null,
"trial_params": null
}