sanchit-gandhi's picture
Add training and evaluation results
4909500
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.444987775061125,
"global_step": 1000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.06,
"learning_rate": 4.6000000000000004e-07,
"loss": 0.8917,
"step": 25
},
{
"epoch": 0.12,
"learning_rate": 9.400000000000001e-07,
"loss": 0.8196,
"step": 50
},
{
"epoch": 0.18,
"learning_rate": 1.44e-06,
"loss": 0.6476,
"step": 75
},
{
"epoch": 0.24,
"learning_rate": 1.94e-06,
"loss": 0.4701,
"step": 100
},
{
"epoch": 0.31,
"learning_rate": 2.4400000000000004e-06,
"loss": 0.5104,
"step": 125
},
{
"epoch": 0.37,
"learning_rate": 2.9400000000000002e-06,
"loss": 0.3547,
"step": 150
},
{
"epoch": 0.43,
"learning_rate": 3.44e-06,
"loss": 0.4427,
"step": 175
},
{
"epoch": 0.49,
"learning_rate": 3.94e-06,
"loss": 0.3209,
"step": 200
},
{
"epoch": 0.55,
"learning_rate": 4.440000000000001e-06,
"loss": 0.3846,
"step": 225
},
{
"epoch": 0.61,
"learning_rate": 4.94e-06,
"loss": 0.252,
"step": 250
},
{
"epoch": 0.67,
"learning_rate": 5.4400000000000004e-06,
"loss": 0.3344,
"step": 275
},
{
"epoch": 0.73,
"learning_rate": 5.94e-06,
"loss": 0.2421,
"step": 300
},
{
"epoch": 0.79,
"learning_rate": 6.440000000000001e-06,
"loss": 0.3204,
"step": 325
},
{
"epoch": 0.86,
"learning_rate": 6.9400000000000005e-06,
"loss": 0.2263,
"step": 350
},
{
"epoch": 0.92,
"learning_rate": 7.440000000000001e-06,
"loss": 0.2931,
"step": 375
},
{
"epoch": 0.98,
"learning_rate": 7.94e-06,
"loss": 0.1939,
"step": 400
},
{
"epoch": 1.04,
"learning_rate": 8.44e-06,
"loss": 0.2428,
"step": 425
},
{
"epoch": 1.1,
"learning_rate": 8.94e-06,
"loss": 0.1915,
"step": 450
},
{
"epoch": 1.16,
"learning_rate": 9.440000000000001e-06,
"loss": 0.1955,
"step": 475
},
{
"epoch": 1.22,
"learning_rate": 9.940000000000001e-06,
"loss": 0.1694,
"step": 500
},
{
"epoch": 1.28,
"learning_rate": 9.951111111111111e-06,
"loss": 0.1946,
"step": 525
},
{
"epoch": 1.34,
"learning_rate": 9.895555555555557e-06,
"loss": 0.1883,
"step": 550
},
{
"epoch": 1.41,
"learning_rate": 9.84e-06,
"loss": 0.1889,
"step": 575
},
{
"epoch": 1.47,
"learning_rate": 9.784444444444445e-06,
"loss": 0.1598,
"step": 600
},
{
"epoch": 1.53,
"learning_rate": 9.72888888888889e-06,
"loss": 0.1764,
"step": 625
},
{
"epoch": 1.59,
"learning_rate": 9.673333333333334e-06,
"loss": 0.1652,
"step": 650
},
{
"epoch": 1.65,
"learning_rate": 9.617777777777778e-06,
"loss": 0.1718,
"step": 675
},
{
"epoch": 1.71,
"learning_rate": 9.562222222222223e-06,
"loss": 0.157,
"step": 700
},
{
"epoch": 1.77,
"learning_rate": 9.506666666666667e-06,
"loss": 0.1685,
"step": 725
},
{
"epoch": 1.83,
"learning_rate": 9.451111111111112e-06,
"loss": 0.1597,
"step": 750
},
{
"epoch": 1.89,
"learning_rate": 9.395555555555556e-06,
"loss": 0.1641,
"step": 775
},
{
"epoch": 1.96,
"learning_rate": 9.340000000000002e-06,
"loss": 0.1446,
"step": 800
},
{
"epoch": 2.02,
"learning_rate": 9.284444444444444e-06,
"loss": 0.1322,
"step": 825
},
{
"epoch": 2.08,
"learning_rate": 9.22888888888889e-06,
"loss": 0.0919,
"step": 850
},
{
"epoch": 2.14,
"learning_rate": 9.173333333333334e-06,
"loss": 0.0799,
"step": 875
},
{
"epoch": 2.2,
"learning_rate": 9.117777777777778e-06,
"loss": 0.0925,
"step": 900
},
{
"epoch": 2.26,
"learning_rate": 9.062222222222224e-06,
"loss": 0.0725,
"step": 925
},
{
"epoch": 2.32,
"learning_rate": 9.006666666666666e-06,
"loss": 0.1034,
"step": 950
},
{
"epoch": 2.38,
"learning_rate": 8.951111111111112e-06,
"loss": 0.082,
"step": 975
},
{
"epoch": 2.44,
"learning_rate": 8.895555555555556e-06,
"loss": 0.1011,
"step": 1000
},
{
"epoch": 2.44,
"eval_loss": 0.30747494101524353,
"eval_runtime": 799.5663,
"eval_samples_per_second": 3.619,
"eval_steps_per_second": 0.226,
"eval_wer": 0.3463133835604842,
"step": 1000
}
],
"max_steps": 5000,
"num_train_epochs": 13,
"total_flos": 4.61505772191744e+18,
"trial_name": null,
"trial_params": null
}