Upload trainer_state.json
Browse files- trainer_state.json +160 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,160 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.09445525705814362,
|
| 3 |
+
"best_model_checkpoint": "/kaggle/working/checkpoint-48000",
|
| 4 |
+
"epoch": 1.9991670137442732,
|
| 5 |
+
"global_step": 48000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.25,
|
| 12 |
+
"learning_rate": 1.750104123281966e-05,
|
| 13 |
+
"loss": 0.0319,
|
| 14 |
+
"step": 6000
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.25,
|
| 18 |
+
"eval_accuracy": 0.9821939782823297,
|
| 19 |
+
"eval_f1": 0.9897475612268823,
|
| 20 |
+
"eval_loss": 0.16341181099414825,
|
| 21 |
+
"eval_precision": 0.9992109831150386,
|
| 22 |
+
"eval_recall": 0.9804617117117117,
|
| 23 |
+
"eval_runtime": 789.0185,
|
| 24 |
+
"eval_samples_per_second": 102.71,
|
| 25 |
+
"eval_steps_per_second": 12.839,
|
| 26 |
+
"step": 6000
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.5,
|
| 30 |
+
"learning_rate": 1.5002082465639318e-05,
|
| 31 |
+
"loss": 0.0147,
|
| 32 |
+
"step": 12000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.5,
|
| 36 |
+
"eval_accuracy": 0.9766658440276407,
|
| 37 |
+
"eval_f1": 0.9865158764680296,
|
| 38 |
+
"eval_loss": 0.10613188147544861,
|
| 39 |
+
"eval_precision": 0.999638723102935,
|
| 40 |
+
"eval_recall": 0.9737331081081081,
|
| 41 |
+
"eval_runtime": 793.5025,
|
| 42 |
+
"eval_samples_per_second": 102.129,
|
| 43 |
+
"eval_steps_per_second": 12.766,
|
| 44 |
+
"step": 12000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.75,
|
| 48 |
+
"learning_rate": 1.2503123698458976e-05,
|
| 49 |
+
"loss": 0.0107,
|
| 50 |
+
"step": 18000
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.75,
|
| 54 |
+
"eval_accuracy": 0.9957428430404739,
|
| 55 |
+
"eval_f1": 0.9975677676336846,
|
| 56 |
+
"eval_loss": 0.043012455105781555,
|
| 57 |
+
"eval_precision": 0.9992232186992444,
|
| 58 |
+
"eval_recall": 0.9959177927927928,
|
| 59 |
+
"eval_runtime": 792.2966,
|
| 60 |
+
"eval_samples_per_second": 102.285,
|
| 61 |
+
"eval_steps_per_second": 12.786,
|
| 62 |
+
"step": 18000
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 1.0,
|
| 66 |
+
"learning_rate": 1.0004164931278636e-05,
|
| 67 |
+
"loss": 0.0074,
|
| 68 |
+
"step": 24000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 1.0,
|
| 72 |
+
"eval_accuracy": 0.9847235932872656,
|
| 73 |
+
"eval_f1": 0.9912112564069798,
|
| 74 |
+
"eval_loss": 0.09504717588424683,
|
| 75 |
+
"eval_precision": 0.9998567786657501,
|
| 76 |
+
"eval_recall": 0.9827139639639639,
|
| 77 |
+
"eval_runtime": 790.115,
|
| 78 |
+
"eval_samples_per_second": 102.567,
|
| 79 |
+
"eval_steps_per_second": 12.821,
|
| 80 |
+
"step": 24000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 1.25,
|
| 84 |
+
"learning_rate": 7.505206164098294e-06,
|
| 85 |
+
"loss": 0.0041,
|
| 86 |
+
"step": 30000
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 1.25,
|
| 90 |
+
"eval_accuracy": 0.9903998025666337,
|
| 91 |
+
"eval_f1": 0.9944949973819397,
|
| 92 |
+
"eval_loss": 0.08727405965328217,
|
| 93 |
+
"eval_precision": 0.9998292689867114,
|
| 94 |
+
"eval_recall": 0.9892173423423424,
|
| 95 |
+
"eval_runtime": 788.7991,
|
| 96 |
+
"eval_samples_per_second": 102.738,
|
| 97 |
+
"eval_steps_per_second": 12.842,
|
| 98 |
+
"step": 30000
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.5,
|
| 102 |
+
"learning_rate": 5.006247396917952e-06,
|
| 103 |
+
"loss": 0.0031,
|
| 104 |
+
"step": 36000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 1.5,
|
| 108 |
+
"eval_accuracy": 0.9819718657453109,
|
| 109 |
+
"eval_f1": 0.9896102233695304,
|
| 110 |
+
"eval_loss": 0.1629093438386917,
|
| 111 |
+
"eval_precision": 1.0,
|
| 112 |
+
"eval_recall": 0.9794341216216216,
|
| 113 |
+
"eval_runtime": 788.9221,
|
| 114 |
+
"eval_samples_per_second": 102.722,
|
| 115 |
+
"eval_steps_per_second": 12.84,
|
| 116 |
+
"step": 36000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.75,
|
| 120 |
+
"learning_rate": 2.5072886297376097e-06,
|
| 121 |
+
"loss": 0.0022,
|
| 122 |
+
"step": 42000
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 1.75,
|
| 126 |
+
"eval_accuracy": 0.9804540967423495,
|
| 127 |
+
"eval_f1": 0.9887258181611126,
|
| 128 |
+
"eval_loss": 0.1487206518650055,
|
| 129 |
+
"eval_precision": 0.9999856028103314,
|
| 130 |
+
"eval_recall": 0.9777167792792792,
|
| 131 |
+
"eval_runtime": 791.8683,
|
| 132 |
+
"eval_samples_per_second": 102.34,
|
| 133 |
+
"eval_steps_per_second": 12.793,
|
| 134 |
+
"step": 42000
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 2.0,
|
| 138 |
+
"learning_rate": 8.329862557267807e-09,
|
| 139 |
+
"loss": 0.001,
|
| 140 |
+
"step": 48000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 2.0,
|
| 144 |
+
"eval_accuracy": 0.9904368213228035,
|
| 145 |
+
"eval_f1": 0.9945157204220418,
|
| 146 |
+
"eval_loss": 0.09445525705814362,
|
| 147 |
+
"eval_precision": 0.9999430791342336,
|
| 148 |
+
"eval_recall": 0.9891469594594594,
|
| 149 |
+
"eval_runtime": 792.2508,
|
| 150 |
+
"eval_samples_per_second": 102.291,
|
| 151 |
+
"eval_steps_per_second": 12.786,
|
| 152 |
+
"step": 48000
|
| 153 |
+
}
|
| 154 |
+
],
|
| 155 |
+
"max_steps": 48020,
|
| 156 |
+
"num_train_epochs": 2,
|
| 157 |
+
"total_flos": 7.687934370520642e+16,
|
| 158 |
+
"trial_name": null,
|
| 159 |
+
"trial_params": null
|
| 160 |
+
}
|