wav2vec2-base-EMOPIA / trainer_state.json
kurosekurose's picture
End of training
0605546
raw
history blame
7.91 kB
{
"best_metric": 0.7888200283050537,
"best_model_checkpoint": "kurosekurose/wav2vec2-base-EMOPIA/checkpoint-812",
"epoch": 19.994285714285713,
"global_step": 1160,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.99,
"learning_rate": 1.1600000000000001e-06,
"loss": 1.393,
"step": 58
},
{
"epoch": 0.99,
"eval_accuracy": 0.5285714268684387,
"eval_loss": 1.3617215156555176,
"eval_runtime": 48.0313,
"eval_samples_per_second": 1.457,
"eval_steps_per_second": 0.375,
"step": 58
},
{
"epoch": 1.99,
"learning_rate": 2.3200000000000002e-06,
"loss": 1.3257,
"step": 116
},
{
"epoch": 1.99,
"eval_accuracy": 0.4000000059604645,
"eval_loss": 1.2633056640625,
"eval_runtime": 48.3007,
"eval_samples_per_second": 1.449,
"eval_steps_per_second": 0.373,
"step": 116
},
{
"epoch": 2.99,
"learning_rate": 3.48e-06,
"loss": 1.175,
"step": 174
},
{
"epoch": 2.99,
"eval_accuracy": 0.4000000059604645,
"eval_loss": 1.11065673828125,
"eval_runtime": 48.0736,
"eval_samples_per_second": 1.456,
"eval_steps_per_second": 0.374,
"step": 174
},
{
"epoch": 3.99,
"learning_rate": 4.6400000000000005e-06,
"loss": 1.0229,
"step": 232
},
{
"epoch": 3.99,
"eval_accuracy": 0.4285714328289032,
"eval_loss": 1.0352715253829956,
"eval_runtime": 48.1873,
"eval_samples_per_second": 1.453,
"eval_steps_per_second": 0.374,
"step": 232
},
{
"epoch": 4.99,
"learning_rate": 5.8e-06,
"loss": 0.9276,
"step": 290
},
{
"epoch": 4.99,
"eval_accuracy": 0.37142857909202576,
"eval_loss": 1.0740830898284912,
"eval_runtime": 48.3346,
"eval_samples_per_second": 1.448,
"eval_steps_per_second": 0.372,
"step": 290
},
{
"epoch": 5.99,
"learning_rate": 6.96e-06,
"loss": 0.8505,
"step": 348
},
{
"epoch": 5.99,
"eval_accuracy": 0.6285714507102966,
"eval_loss": 0.9529690146446228,
"eval_runtime": 48.6491,
"eval_samples_per_second": 1.439,
"eval_steps_per_second": 0.37,
"step": 348
},
{
"epoch": 6.99,
"learning_rate": 8.120000000000002e-06,
"loss": 0.7713,
"step": 406
},
{
"epoch": 6.99,
"eval_accuracy": 0.6142857074737549,
"eval_loss": 0.9922705888748169,
"eval_runtime": 48.6353,
"eval_samples_per_second": 1.439,
"eval_steps_per_second": 0.37,
"step": 406
},
{
"epoch": 7.99,
"learning_rate": 9.280000000000001e-06,
"loss": 0.7208,
"step": 464
},
{
"epoch": 7.99,
"eval_accuracy": 0.6142857074737549,
"eval_loss": 0.8938446044921875,
"eval_runtime": 48.2669,
"eval_samples_per_second": 1.45,
"eval_steps_per_second": 0.373,
"step": 464
},
{
"epoch": 8.99,
"learning_rate": 9.666666666666667e-06,
"loss": 0.6573,
"step": 522
},
{
"epoch": 8.99,
"eval_accuracy": 0.6142857074737549,
"eval_loss": 1.0785200595855713,
"eval_runtime": 47.8068,
"eval_samples_per_second": 1.464,
"eval_steps_per_second": 0.377,
"step": 522
},
{
"epoch": 9.99,
"learning_rate": 8.787878787878788e-06,
"loss": 0.6134,
"step": 580
},
{
"epoch": 9.99,
"eval_accuracy": 0.6428571343421936,
"eval_loss": 0.9662051796913147,
"eval_runtime": 48.1596,
"eval_samples_per_second": 1.454,
"eval_steps_per_second": 0.374,
"step": 580
},
{
"epoch": 10.99,
"learning_rate": 7.909090909090909e-06,
"loss": 0.5514,
"step": 638
},
{
"epoch": 10.99,
"eval_accuracy": 0.6571428775787354,
"eval_loss": 0.9189529418945312,
"eval_runtime": 48.0311,
"eval_samples_per_second": 1.457,
"eval_steps_per_second": 0.375,
"step": 638
},
{
"epoch": 11.99,
"learning_rate": 7.030303030303031e-06,
"loss": 0.4643,
"step": 696
},
{
"epoch": 11.99,
"eval_accuracy": 0.6857143044471741,
"eval_loss": 0.8748092651367188,
"eval_runtime": 48.3032,
"eval_samples_per_second": 1.449,
"eval_steps_per_second": 0.373,
"step": 696
},
{
"epoch": 12.99,
"learning_rate": 6.166666666666667e-06,
"loss": 0.3946,
"step": 754
},
{
"epoch": 12.99,
"eval_accuracy": 0.6571428775787354,
"eval_loss": 0.9243893027305603,
"eval_runtime": 48.1936,
"eval_samples_per_second": 1.452,
"eval_steps_per_second": 0.373,
"step": 754
},
{
"epoch": 13.99,
"learning_rate": 5.287878787878788e-06,
"loss": 0.3875,
"step": 812
},
{
"epoch": 13.99,
"eval_accuracy": 0.7285714149475098,
"eval_loss": 0.7888200283050537,
"eval_runtime": 47.9295,
"eval_samples_per_second": 1.46,
"eval_steps_per_second": 0.376,
"step": 812
},
{
"epoch": 14.99,
"learning_rate": 4.409090909090909e-06,
"loss": 0.3044,
"step": 870
},
{
"epoch": 14.99,
"eval_accuracy": 0.6714285612106323,
"eval_loss": 0.8981361389160156,
"eval_runtime": 47.9952,
"eval_samples_per_second": 1.458,
"eval_steps_per_second": 0.375,
"step": 870
},
{
"epoch": 15.99,
"learning_rate": 3.5303030303030304e-06,
"loss": 0.2794,
"step": 928
},
{
"epoch": 15.99,
"eval_accuracy": 0.6714285612106323,
"eval_loss": 0.8927078247070312,
"eval_runtime": 47.849,
"eval_samples_per_second": 1.463,
"eval_steps_per_second": 0.376,
"step": 928
},
{
"epoch": 16.99,
"learning_rate": 2.6515151515151514e-06,
"loss": 0.2307,
"step": 986
},
{
"epoch": 16.99,
"eval_accuracy": 0.6714285612106323,
"eval_loss": 0.9921481013298035,
"eval_runtime": 48.0049,
"eval_samples_per_second": 1.458,
"eval_steps_per_second": 0.375,
"step": 986
},
{
"epoch": 17.99,
"learning_rate": 1.7727272727272729e-06,
"loss": 0.2153,
"step": 1044
},
{
"epoch": 17.99,
"eval_accuracy": 0.6714285612106323,
"eval_loss": 0.979206919670105,
"eval_runtime": 47.8808,
"eval_samples_per_second": 1.462,
"eval_steps_per_second": 0.376,
"step": 1044
},
{
"epoch": 18.99,
"learning_rate": 8.93939393939394e-07,
"loss": 0.1805,
"step": 1102
},
{
"epoch": 18.99,
"eval_accuracy": 0.6714285612106323,
"eval_loss": 1.1085455417633057,
"eval_runtime": 48.1686,
"eval_samples_per_second": 1.453,
"eval_steps_per_second": 0.374,
"step": 1102
},
{
"epoch": 19.99,
"learning_rate": 3.0303030303030305e-08,
"loss": 0.1625,
"step": 1160
},
{
"epoch": 19.99,
"eval_accuracy": 0.6571428775787354,
"eval_loss": 1.0330772399902344,
"eval_runtime": 48.6,
"eval_samples_per_second": 1.44,
"eval_steps_per_second": 0.37,
"step": 1160
},
{
"epoch": 19.99,
"step": 1160,
"total_flos": 6.655036715585741e+18,
"train_loss": 0.6313976797564277,
"train_runtime": 27101.5799,
"train_samples_per_second": 0.517,
"train_steps_per_second": 0.043
}
],
"max_steps": 1160,
"num_train_epochs": 20,
"total_flos": 6.655036715585741e+18,
"trial_name": null,
"trial_params": null
}