wav2vec2-base-ft-fake-detection / trainer_state.json
alexandreacff's picture
End of training
22c5ea2 verified
{
"best_metric": 0.6523364485981309,
"best_model_checkpoint": "wav2vec2-base-ft-fake-detection/checkpoint-33",
"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": 0.6689966320991516,
"learning_rate": 9.090909090909091e-06,
"loss": 0.696,
"step": 10
},
{
"epoch": 0.5970149253731343,
"grad_norm": 0.972612738609314,
"learning_rate": 1.8181818181818182e-05,
"loss": 0.6649,
"step": 20
},
{
"epoch": 0.8955223880597015,
"grad_norm": 0.47211921215057373,
"learning_rate": 2.7272727272727273e-05,
"loss": 0.6253,
"step": 30
},
{
"epoch": 0.9850746268656716,
"eval_accuracy": 0.6523364485981309,
"eval_loss": 0.6260867714881897,
"eval_runtime": 42.0969,
"eval_samples_per_second": 25.418,
"eval_steps_per_second": 1.592,
"step": 33
},
{
"epoch": 1.1940298507462686,
"grad_norm": 1.006468653678894,
"learning_rate": 2.9292929292929294e-05,
"loss": 0.5877,
"step": 40
},
{
"epoch": 1.4925373134328357,
"grad_norm": 7.379218101501465,
"learning_rate": 2.8282828282828285e-05,
"loss": 0.4855,
"step": 50
},
{
"epoch": 1.7910447761194028,
"grad_norm": 3.237988233566284,
"learning_rate": 2.7272727272727273e-05,
"loss": 0.4394,
"step": 60
},
{
"epoch": 2.0,
"eval_accuracy": 0.5644859813084112,
"eval_loss": 0.7140001058578491,
"eval_runtime": 39.3742,
"eval_samples_per_second": 27.175,
"eval_steps_per_second": 1.702,
"step": 67
},
{
"epoch": 2.08955223880597,
"grad_norm": 2.6573326587677,
"learning_rate": 2.6262626262626265e-05,
"loss": 0.402,
"step": 70
},
{
"epoch": 2.388059701492537,
"grad_norm": 3.1143898963928223,
"learning_rate": 2.5252525252525256e-05,
"loss": 0.3622,
"step": 80
},
{
"epoch": 2.6865671641791042,
"grad_norm": 3.858957529067993,
"learning_rate": 2.4242424242424244e-05,
"loss": 0.3682,
"step": 90
},
{
"epoch": 2.9850746268656714,
"grad_norm": 2.6959176063537598,
"learning_rate": 2.3232323232323235e-05,
"loss": 0.3685,
"step": 100
},
{
"epoch": 2.9850746268656714,
"eval_accuracy": 0.5850467289719626,
"eval_loss": 0.7180566191673279,
"eval_runtime": 34.4048,
"eval_samples_per_second": 31.1,
"eval_steps_per_second": 1.947,
"step": 100
},
{
"epoch": 3.283582089552239,
"grad_norm": 7.184162616729736,
"learning_rate": 2.222222222222222e-05,
"loss": 0.3469,
"step": 110
},
{
"epoch": 3.582089552238806,
"grad_norm": 3.8342511653900146,
"learning_rate": 2.121212121212121e-05,
"loss": 0.3181,
"step": 120
},
{
"epoch": 3.8805970149253732,
"grad_norm": 3.2687244415283203,
"learning_rate": 2.0202020202020203e-05,
"loss": 0.317,
"step": 130
},
{
"epoch": 4.0,
"eval_accuracy": 0.6149532710280374,
"eval_loss": 0.7291006445884705,
"eval_runtime": 38.9708,
"eval_samples_per_second": 27.456,
"eval_steps_per_second": 1.719,
"step": 134
},
{
"epoch": 4.17910447761194,
"grad_norm": 2.616014003753662,
"learning_rate": 1.919191919191919e-05,
"loss": 0.3229,
"step": 140
},
{
"epoch": 4.477611940298507,
"grad_norm": 6.417969226837158,
"learning_rate": 1.8181818181818182e-05,
"loss": 0.2673,
"step": 150
},
{
"epoch": 4.776119402985074,
"grad_norm": 6.794923782348633,
"learning_rate": 1.717171717171717e-05,
"loss": 0.3027,
"step": 160
},
{
"epoch": 4.985074626865671,
"eval_accuracy": 0.6158878504672897,
"eval_loss": 0.7457267045974731,
"eval_runtime": 37.5396,
"eval_samples_per_second": 28.503,
"eval_steps_per_second": 1.785,
"step": 167
},
{
"epoch": 5.074626865671641,
"grad_norm": 2.6645431518554688,
"learning_rate": 1.616161616161616e-05,
"loss": 0.2804,
"step": 170
},
{
"epoch": 5.373134328358209,
"grad_norm": 3.8583788871765137,
"learning_rate": 1.5151515151515153e-05,
"loss": 0.2613,
"step": 180
},
{
"epoch": 5.6716417910447765,
"grad_norm": 2.621940851211548,
"learning_rate": 1.4141414141414143e-05,
"loss": 0.2578,
"step": 190
},
{
"epoch": 5.970149253731344,
"grad_norm": 4.610651969909668,
"learning_rate": 1.3131313131313132e-05,
"loss": 0.2672,
"step": 200
},
{
"epoch": 6.0,
"eval_accuracy": 0.6242990654205608,
"eval_loss": 0.7805104851722717,
"eval_runtime": 35.9857,
"eval_samples_per_second": 29.734,
"eval_steps_per_second": 1.862,
"step": 201
},
{
"epoch": 6.268656716417911,
"grad_norm": 3.4527761936187744,
"learning_rate": 1.2121212121212122e-05,
"loss": 0.2166,
"step": 210
},
{
"epoch": 6.567164179104478,
"grad_norm": 3.452103614807129,
"learning_rate": 1.111111111111111e-05,
"loss": 0.2513,
"step": 220
},
{
"epoch": 6.865671641791045,
"grad_norm": 3.1861703395843506,
"learning_rate": 1.0101010101010101e-05,
"loss": 0.2711,
"step": 230
},
{
"epoch": 6.985074626865671,
"eval_accuracy": 0.6214953271028038,
"eval_loss": 0.8112902641296387,
"eval_runtime": 36.6515,
"eval_samples_per_second": 29.194,
"eval_steps_per_second": 1.828,
"step": 234
},
{
"epoch": 7.164179104477612,
"grad_norm": 4.2410993576049805,
"learning_rate": 9.090909090909091e-06,
"loss": 0.23,
"step": 240
},
{
"epoch": 7.462686567164179,
"grad_norm": 2.9806125164031982,
"learning_rate": 8.08080808080808e-06,
"loss": 0.2123,
"step": 250
},
{
"epoch": 7.7611940298507465,
"grad_norm": 2.821526527404785,
"learning_rate": 7.070707070707071e-06,
"loss": 0.2086,
"step": 260
},
{
"epoch": 8.0,
"eval_accuracy": 0.5962616822429907,
"eval_loss": 0.9130397439002991,
"eval_runtime": 34.0895,
"eval_samples_per_second": 31.388,
"eval_steps_per_second": 1.965,
"step": 268
},
{
"epoch": 8.059701492537313,
"grad_norm": 4.247228622436523,
"learning_rate": 6.060606060606061e-06,
"loss": 0.2475,
"step": 270
},
{
"epoch": 8.35820895522388,
"grad_norm": 2.185816764831543,
"learning_rate": 5.050505050505051e-06,
"loss": 0.208,
"step": 280
},
{
"epoch": 8.656716417910447,
"grad_norm": 4.324727535247803,
"learning_rate": 4.04040404040404e-06,
"loss": 0.1971,
"step": 290
},
{
"epoch": 8.955223880597014,
"grad_norm": 4.010946750640869,
"learning_rate": 3.0303030303030305e-06,
"loss": 0.2077,
"step": 300
},
{
"epoch": 8.985074626865671,
"eval_accuracy": 0.616822429906542,
"eval_loss": 0.9042268991470337,
"eval_runtime": 33.3964,
"eval_samples_per_second": 32.039,
"eval_steps_per_second": 2.006,
"step": 301
},
{
"epoch": 9.253731343283581,
"grad_norm": 5.064040184020996,
"learning_rate": 2.02020202020202e-06,
"loss": 0.2026,
"step": 310
},
{
"epoch": 9.552238805970148,
"grad_norm": 2.497706890106201,
"learning_rate": 1.01010101010101e-06,
"loss": 0.202,
"step": 320
},
{
"epoch": 9.850746268656717,
"grad_norm": 4.643800735473633,
"learning_rate": 0.0,
"loss": 0.223,
"step": 330
},
{
"epoch": 9.850746268656717,
"eval_accuracy": 0.6177570093457944,
"eval_loss": 0.8923965096473694,
"eval_runtime": 38.1063,
"eval_samples_per_second": 28.079,
"eval_steps_per_second": 1.758,
"step": 330
},
{
"epoch": 9.850746268656717,
"step": 330,
"total_flos": 3.8258265808224e+17,
"train_loss": 0.3278515060742696,
"train_runtime": 1794.8751,
"train_samples_per_second": 23.829,
"train_steps_per_second": 0.184
}
],
"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": 3.8258265808224e+17,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}