mode_test_001 / trainer_state.json
arthoho66's picture
Upload 19 files
14d9c4f
{
"best_metric": 114.85714285714286,
"best_model_checkpoint": "./medium_TH/checkpoint-1000",
"epoch": 55.55555555555556,
"global_step": 1000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 1.39,
"learning_rate": 4.6000000000000004e-07,
"loss": 0.8382,
"step": 25
},
{
"epoch": 2.78,
"learning_rate": 9.400000000000001e-07,
"loss": 0.7178,
"step": 50
},
{
"epoch": 4.17,
"learning_rate": 1.44e-06,
"loss": 0.5937,
"step": 75
},
{
"epoch": 5.56,
"learning_rate": 1.94e-06,
"loss": 0.5143,
"step": 100
},
{
"epoch": 6.94,
"learning_rate": 2.4400000000000004e-06,
"loss": 0.3861,
"step": 125
},
{
"epoch": 8.33,
"learning_rate": 2.9400000000000002e-06,
"loss": 0.3526,
"step": 150
},
{
"epoch": 9.72,
"learning_rate": 3.44e-06,
"loss": 0.287,
"step": 175
},
{
"epoch": 11.11,
"learning_rate": 3.94e-06,
"loss": 0.1953,
"step": 200
},
{
"epoch": 12.5,
"learning_rate": 4.440000000000001e-06,
"loss": 0.1874,
"step": 225
},
{
"epoch": 13.89,
"learning_rate": 4.94e-06,
"loss": 0.172,
"step": 250
},
{
"epoch": 15.28,
"learning_rate": 5.4400000000000004e-06,
"loss": 0.0642,
"step": 275
},
{
"epoch": 16.67,
"learning_rate": 5.94e-06,
"loss": 0.09,
"step": 300
},
{
"epoch": 18.06,
"learning_rate": 6.440000000000001e-06,
"loss": 0.0578,
"step": 325
},
{
"epoch": 19.44,
"learning_rate": 6.9400000000000005e-06,
"loss": 0.0384,
"step": 350
},
{
"epoch": 20.83,
"learning_rate": 7.440000000000001e-06,
"loss": 0.0672,
"step": 375
},
{
"epoch": 22.22,
"learning_rate": 7.94e-06,
"loss": 0.0689,
"step": 400
},
{
"epoch": 23.61,
"learning_rate": 8.44e-06,
"loss": 0.0902,
"step": 425
},
{
"epoch": 25.0,
"learning_rate": 8.94e-06,
"loss": 0.0497,
"step": 450
},
{
"epoch": 26.39,
"learning_rate": 9.440000000000001e-06,
"loss": 0.0232,
"step": 475
},
{
"epoch": 27.78,
"learning_rate": 9.940000000000001e-06,
"loss": 0.0465,
"step": 500
},
{
"epoch": 29.17,
"learning_rate": 9.937142857142858e-06,
"loss": 0.0282,
"step": 525
},
{
"epoch": 30.56,
"learning_rate": 9.865714285714285e-06,
"loss": 0.0546,
"step": 550
},
{
"epoch": 31.94,
"learning_rate": 9.794285714285714e-06,
"loss": 0.0786,
"step": 575
},
{
"epoch": 33.33,
"learning_rate": 9.722857142857143e-06,
"loss": 0.0628,
"step": 600
},
{
"epoch": 34.72,
"learning_rate": 9.651428571428572e-06,
"loss": 0.0325,
"step": 625
},
{
"epoch": 36.11,
"learning_rate": 9.58e-06,
"loss": 0.0254,
"step": 650
},
{
"epoch": 37.5,
"learning_rate": 9.508571428571429e-06,
"loss": 0.0358,
"step": 675
},
{
"epoch": 38.89,
"learning_rate": 9.437142857142858e-06,
"loss": 0.0331,
"step": 700
},
{
"epoch": 40.28,
"learning_rate": 9.365714285714287e-06,
"loss": 0.0281,
"step": 725
},
{
"epoch": 41.67,
"learning_rate": 9.294285714285714e-06,
"loss": 0.0314,
"step": 750
},
{
"epoch": 43.06,
"learning_rate": 9.222857142857143e-06,
"loss": 0.0131,
"step": 775
},
{
"epoch": 44.44,
"learning_rate": 9.151428571428572e-06,
"loss": 0.0245,
"step": 800
},
{
"epoch": 45.83,
"learning_rate": 9.080000000000001e-06,
"loss": 0.0071,
"step": 825
},
{
"epoch": 47.22,
"learning_rate": 9.00857142857143e-06,
"loss": 0.0249,
"step": 850
},
{
"epoch": 48.61,
"learning_rate": 8.937142857142857e-06,
"loss": 0.0186,
"step": 875
},
{
"epoch": 50.0,
"learning_rate": 8.865714285714287e-06,
"loss": 0.0066,
"step": 900
},
{
"epoch": 51.39,
"learning_rate": 8.794285714285716e-06,
"loss": 0.0114,
"step": 925
},
{
"epoch": 52.78,
"learning_rate": 8.722857142857145e-06,
"loss": 0.0178,
"step": 950
},
{
"epoch": 54.17,
"learning_rate": 8.651428571428572e-06,
"loss": 0.0188,
"step": 975
},
{
"epoch": 55.56,
"learning_rate": 8.580000000000001e-06,
"loss": 0.0272,
"step": 1000
},
{
"epoch": 55.56,
"eval_loss": 0.37656262516975403,
"eval_runtime": 10.4546,
"eval_samples_per_second": 2.2,
"eval_steps_per_second": 0.287,
"eval_wer": 114.85714285714286,
"step": 1000
}
],
"max_steps": 4000,
"num_train_epochs": 223,
"total_flos": 5.842963095552e+18,
"trial_name": null,
"trial_params": null
}