wav2vec2-large-ft-fake-detection / trainer_state.json
alexandreacff's picture
End of training
194dcc8 verified
{
"best_metric": 0.7102803738317757,
"best_model_checkpoint": "wav2vec2-large-ft-fake-detection/checkpoint-301",
"epoch": 9.850746268656717,
"eval_steps": 500,
"global_step": 330,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.29850746268656714,
"grad_norm": 0.7566494941711426,
"learning_rate": 9.090909090909091e-06,
"loss": 0.6829,
"step": 10
},
{
"epoch": 0.5970149253731343,
"grad_norm": 0.6273395419120789,
"learning_rate": 1.8181818181818182e-05,
"loss": 0.6512,
"step": 20
},
{
"epoch": 0.8955223880597015,
"grad_norm": 1.1124577522277832,
"learning_rate": 2.7272727272727273e-05,
"loss": 0.6274,
"step": 30
},
{
"epoch": 0.9850746268656716,
"eval_accuracy": 0.6205607476635514,
"eval_loss": 0.62535560131073,
"eval_runtime": 39.1441,
"eval_samples_per_second": 27.335,
"eval_steps_per_second": 1.712,
"step": 33
},
{
"epoch": 1.1940298507462686,
"grad_norm": 1.8781670331954956,
"learning_rate": 2.9292929292929294e-05,
"loss": 0.577,
"step": 40
},
{
"epoch": 1.4925373134328357,
"grad_norm": 6.504702568054199,
"learning_rate": 2.8282828282828285e-05,
"loss": 0.5106,
"step": 50
},
{
"epoch": 1.7910447761194028,
"grad_norm": 2.83394455909729,
"learning_rate": 2.7272727272727273e-05,
"loss": 0.4961,
"step": 60
},
{
"epoch": 2.0,
"eval_accuracy": 0.6158878504672897,
"eval_loss": 0.947680652141571,
"eval_runtime": 43.4278,
"eval_samples_per_second": 24.639,
"eval_steps_per_second": 1.543,
"step": 67
},
{
"epoch": 2.08955223880597,
"grad_norm": 2.453080177307129,
"learning_rate": 2.6262626262626265e-05,
"loss": 0.4237,
"step": 70
},
{
"epoch": 2.388059701492537,
"grad_norm": 2.8955702781677246,
"learning_rate": 2.5252525252525256e-05,
"loss": 0.3841,
"step": 80
},
{
"epoch": 2.6865671641791042,
"grad_norm": 2.1122732162475586,
"learning_rate": 2.4242424242424244e-05,
"loss": 0.3749,
"step": 90
},
{
"epoch": 2.9850746268656714,
"grad_norm": 3.88417649269104,
"learning_rate": 2.3232323232323235e-05,
"loss": 0.3391,
"step": 100
},
{
"epoch": 2.9850746268656714,
"eval_accuracy": 0.6411214953271028,
"eval_loss": 0.9273380041122437,
"eval_runtime": 39.2854,
"eval_samples_per_second": 27.237,
"eval_steps_per_second": 1.705,
"step": 100
},
{
"epoch": 3.283582089552239,
"grad_norm": 2.406618595123291,
"learning_rate": 2.222222222222222e-05,
"loss": 0.3291,
"step": 110
},
{
"epoch": 3.582089552238806,
"grad_norm": 2.385477066040039,
"learning_rate": 2.121212121212121e-05,
"loss": 0.3334,
"step": 120
},
{
"epoch": 3.8805970149253732,
"grad_norm": 5.02505350112915,
"learning_rate": 2.0202020202020203e-05,
"loss": 0.2857,
"step": 130
},
{
"epoch": 4.0,
"eval_accuracy": 0.6616822429906543,
"eval_loss": 0.6611053347587585,
"eval_runtime": 39.7207,
"eval_samples_per_second": 26.938,
"eval_steps_per_second": 1.687,
"step": 134
},
{
"epoch": 4.17910447761194,
"grad_norm": 2.3081231117248535,
"learning_rate": 1.919191919191919e-05,
"loss": 0.3724,
"step": 140
},
{
"epoch": 4.477611940298507,
"grad_norm": 3.455447196960449,
"learning_rate": 1.8181818181818182e-05,
"loss": 0.2929,
"step": 150
},
{
"epoch": 4.776119402985074,
"grad_norm": 3.234449625015259,
"learning_rate": 1.717171717171717e-05,
"loss": 0.3186,
"step": 160
},
{
"epoch": 4.985074626865671,
"eval_accuracy": 0.6214953271028038,
"eval_loss": 0.7654162049293518,
"eval_runtime": 40.6872,
"eval_samples_per_second": 26.298,
"eval_steps_per_second": 1.647,
"step": 167
},
{
"epoch": 5.074626865671641,
"grad_norm": 1.7067972421646118,
"learning_rate": 1.616161616161616e-05,
"loss": 0.308,
"step": 170
},
{
"epoch": 5.373134328358209,
"grad_norm": 2.3431663513183594,
"learning_rate": 1.5151515151515153e-05,
"loss": 0.2597,
"step": 180
},
{
"epoch": 5.6716417910447765,
"grad_norm": 5.564834117889404,
"learning_rate": 1.4141414141414143e-05,
"loss": 0.2341,
"step": 190
},
{
"epoch": 5.970149253731344,
"grad_norm": 4.282162189483643,
"learning_rate": 1.3131313131313132e-05,
"loss": 0.2483,
"step": 200
},
{
"epoch": 6.0,
"eval_accuracy": 0.6224299065420561,
"eval_loss": 0.9394506216049194,
"eval_runtime": 37.7276,
"eval_samples_per_second": 28.361,
"eval_steps_per_second": 1.776,
"step": 201
},
{
"epoch": 6.268656716417911,
"grad_norm": 3.2581288814544678,
"learning_rate": 1.2121212121212122e-05,
"loss": 0.2363,
"step": 210
},
{
"epoch": 6.567164179104478,
"grad_norm": 3.974890947341919,
"learning_rate": 1.111111111111111e-05,
"loss": 0.2808,
"step": 220
},
{
"epoch": 6.865671641791045,
"grad_norm": 2.1842823028564453,
"learning_rate": 1.0101010101010101e-05,
"loss": 0.239,
"step": 230
},
{
"epoch": 6.985074626865671,
"eval_accuracy": 0.6542056074766355,
"eval_loss": 0.8366522192955017,
"eval_runtime": 42.0615,
"eval_samples_per_second": 25.439,
"eval_steps_per_second": 1.593,
"step": 234
},
{
"epoch": 7.164179104477612,
"grad_norm": 2.455822706222534,
"learning_rate": 9.090909090909091e-06,
"loss": 0.2156,
"step": 240
},
{
"epoch": 7.462686567164179,
"grad_norm": 4.788252830505371,
"learning_rate": 8.08080808080808e-06,
"loss": 0.2297,
"step": 250
},
{
"epoch": 7.7611940298507465,
"grad_norm": 3.2945239543914795,
"learning_rate": 7.070707070707071e-06,
"loss": 0.2049,
"step": 260
},
{
"epoch": 8.0,
"eval_accuracy": 0.685981308411215,
"eval_loss": 0.7708710432052612,
"eval_runtime": 43.0532,
"eval_samples_per_second": 24.853,
"eval_steps_per_second": 1.556,
"step": 268
},
{
"epoch": 8.059701492537313,
"grad_norm": 3.367969512939453,
"learning_rate": 6.060606060606061e-06,
"loss": 0.2485,
"step": 270
},
{
"epoch": 8.35820895522388,
"grad_norm": 2.869590997695923,
"learning_rate": 5.050505050505051e-06,
"loss": 0.2318,
"step": 280
},
{
"epoch": 8.656716417910447,
"grad_norm": 3.1148149967193604,
"learning_rate": 4.04040404040404e-06,
"loss": 0.2058,
"step": 290
},
{
"epoch": 8.955223880597014,
"grad_norm": 4.1323065757751465,
"learning_rate": 3.0303030303030305e-06,
"loss": 0.224,
"step": 300
},
{
"epoch": 8.985074626865671,
"eval_accuracy": 0.7102803738317757,
"eval_loss": 0.669384777545929,
"eval_runtime": 42.9166,
"eval_samples_per_second": 24.932,
"eval_steps_per_second": 1.561,
"step": 301
},
{
"epoch": 9.253731343283581,
"grad_norm": 3.609107494354248,
"learning_rate": 2.02020202020202e-06,
"loss": 0.2291,
"step": 310
},
{
"epoch": 9.552238805970148,
"grad_norm": 2.175992250442505,
"learning_rate": 1.01010101010101e-06,
"loss": 0.2151,
"step": 320
},
{
"epoch": 9.850746268656717,
"grad_norm": 3.807249069213867,
"learning_rate": 0.0,
"loss": 0.2279,
"step": 330
},
{
"epoch": 9.850746268656717,
"eval_accuracy": 0.6822429906542056,
"eval_loss": 0.786749541759491,
"eval_runtime": 39.843,
"eval_samples_per_second": 26.855,
"eval_steps_per_second": 1.682,
"step": 330
},
{
"epoch": 9.850746268656717,
"step": 330,
"total_flos": 1.277142970631616e+18,
"train_loss": 0.33447390686381945,
"train_runtime": 2006.3999,
"train_samples_per_second": 21.317,
"train_steps_per_second": 0.164
}
],
"logging_steps": 10,
"max_steps": 330,
"num_input_tokens_seen": 0,
"num_train_epochs": 10,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.277142970631616e+18,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}