JRHuy's picture
Training in progress, step 2000
b7e39b2
raw
history blame
9.54 kB
{
"best_metric": 16.120763315294788,
"best_model_checkpoint": "./whisper-small-vietnamese/checkpoint-1000",
"epoch": 5.730659025787966,
"global_step": 2000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.07,
"learning_rate": 4.6000000000000004e-07,
"loss": 6.1527,
"step": 25
},
{
"epoch": 0.14,
"learning_rate": 9.600000000000001e-07,
"loss": 3.4261,
"step": 50
},
{
"epoch": 0.21,
"learning_rate": 1.46e-06,
"loss": 1.7644,
"step": 75
},
{
"epoch": 0.29,
"learning_rate": 1.9600000000000003e-06,
"loss": 1.0756,
"step": 100
},
{
"epoch": 0.36,
"learning_rate": 2.46e-06,
"loss": 0.7441,
"step": 125
},
{
"epoch": 0.43,
"learning_rate": 2.96e-06,
"loss": 0.6409,
"step": 150
},
{
"epoch": 0.5,
"learning_rate": 3.46e-06,
"loss": 0.5309,
"step": 175
},
{
"epoch": 0.57,
"learning_rate": 3.96e-06,
"loss": 0.5382,
"step": 200
},
{
"epoch": 0.64,
"learning_rate": 4.4600000000000005e-06,
"loss": 0.4923,
"step": 225
},
{
"epoch": 0.72,
"learning_rate": 4.960000000000001e-06,
"loss": 0.4704,
"step": 250
},
{
"epoch": 0.79,
"learning_rate": 5.460000000000001e-06,
"loss": 0.4312,
"step": 275
},
{
"epoch": 0.86,
"learning_rate": 5.9600000000000005e-06,
"loss": 0.4088,
"step": 300
},
{
"epoch": 0.93,
"learning_rate": 6.460000000000001e-06,
"loss": 0.3795,
"step": 325
},
{
"epoch": 1.0,
"learning_rate": 6.96e-06,
"loss": 0.3983,
"step": 350
},
{
"epoch": 1.07,
"learning_rate": 7.4600000000000006e-06,
"loss": 0.3188,
"step": 375
},
{
"epoch": 1.15,
"learning_rate": 7.960000000000002e-06,
"loss": 0.2895,
"step": 400
},
{
"epoch": 1.22,
"learning_rate": 8.46e-06,
"loss": 0.3059,
"step": 425
},
{
"epoch": 1.29,
"learning_rate": 8.96e-06,
"loss": 0.3012,
"step": 450
},
{
"epoch": 1.36,
"learning_rate": 9.460000000000001e-06,
"loss": 0.2889,
"step": 475
},
{
"epoch": 1.43,
"learning_rate": 9.960000000000001e-06,
"loss": 0.2848,
"step": 500
},
{
"epoch": 1.5,
"learning_rate": 1e-05,
"loss": 0.287,
"step": 525
},
{
"epoch": 1.58,
"learning_rate": 1e-05,
"loss": 0.2979,
"step": 550
},
{
"epoch": 1.65,
"learning_rate": 1e-05,
"loss": 0.2735,
"step": 575
},
{
"epoch": 1.72,
"learning_rate": 1e-05,
"loss": 0.2613,
"step": 600
},
{
"epoch": 1.79,
"learning_rate": 1e-05,
"loss": 0.3091,
"step": 625
},
{
"epoch": 1.86,
"learning_rate": 1e-05,
"loss": 0.2647,
"step": 650
},
{
"epoch": 1.93,
"learning_rate": 1e-05,
"loss": 0.248,
"step": 675
},
{
"epoch": 2.01,
"learning_rate": 1e-05,
"loss": 0.2379,
"step": 700
},
{
"epoch": 2.08,
"learning_rate": 1e-05,
"loss": 0.1577,
"step": 725
},
{
"epoch": 2.15,
"learning_rate": 1e-05,
"loss": 0.1555,
"step": 750
},
{
"epoch": 2.22,
"learning_rate": 1e-05,
"loss": 0.1546,
"step": 775
},
{
"epoch": 2.29,
"learning_rate": 1e-05,
"loss": 0.1377,
"step": 800
},
{
"epoch": 2.36,
"learning_rate": 1e-05,
"loss": 0.1655,
"step": 825
},
{
"epoch": 2.44,
"learning_rate": 1e-05,
"loss": 0.1463,
"step": 850
},
{
"epoch": 2.51,
"learning_rate": 1e-05,
"loss": 0.1485,
"step": 875
},
{
"epoch": 2.58,
"learning_rate": 1e-05,
"loss": 0.1525,
"step": 900
},
{
"epoch": 2.65,
"learning_rate": 1e-05,
"loss": 0.1471,
"step": 925
},
{
"epoch": 2.72,
"learning_rate": 1e-05,
"loss": 0.1496,
"step": 950
},
{
"epoch": 2.79,
"learning_rate": 1e-05,
"loss": 0.1487,
"step": 975
},
{
"epoch": 2.87,
"learning_rate": 1e-05,
"loss": 0.1455,
"step": 1000
},
{
"epoch": 2.87,
"eval_cer": 8.51668353666116,
"eval_loss": 0.3513067960739136,
"eval_runtime": 693.7124,
"eval_samples_per_second": 1.591,
"eval_steps_per_second": 0.099,
"eval_wer": 16.120763315294788,
"step": 1000
},
{
"epoch": 2.94,
"learning_rate": 1e-05,
"loss": 0.1508,
"step": 1025
},
{
"epoch": 3.01,
"learning_rate": 1e-05,
"loss": 0.1379,
"step": 1050
},
{
"epoch": 3.08,
"learning_rate": 1e-05,
"loss": 0.0688,
"step": 1075
},
{
"epoch": 3.15,
"learning_rate": 1e-05,
"loss": 0.0738,
"step": 1100
},
{
"epoch": 3.22,
"learning_rate": 1e-05,
"loss": 0.0731,
"step": 1125
},
{
"epoch": 3.3,
"learning_rate": 1e-05,
"loss": 0.0767,
"step": 1150
},
{
"epoch": 3.37,
"learning_rate": 1e-05,
"loss": 0.0742,
"step": 1175
},
{
"epoch": 3.44,
"learning_rate": 1e-05,
"loss": 0.0734,
"step": 1200
},
{
"epoch": 3.51,
"learning_rate": 1e-05,
"loss": 0.0774,
"step": 1225
},
{
"epoch": 3.58,
"learning_rate": 1e-05,
"loss": 0.0738,
"step": 1250
},
{
"epoch": 3.65,
"learning_rate": 1e-05,
"loss": 0.0742,
"step": 1275
},
{
"epoch": 3.72,
"learning_rate": 1e-05,
"loss": 0.0737,
"step": 1300
},
{
"epoch": 3.8,
"learning_rate": 1e-05,
"loss": 0.0781,
"step": 1325
},
{
"epoch": 3.87,
"learning_rate": 1e-05,
"loss": 0.0741,
"step": 1350
},
{
"epoch": 3.94,
"learning_rate": 1e-05,
"loss": 0.0732,
"step": 1375
},
{
"epoch": 4.01,
"learning_rate": 1e-05,
"loss": 0.0726,
"step": 1400
},
{
"epoch": 4.08,
"learning_rate": 1e-05,
"loss": 0.0364,
"step": 1425
},
{
"epoch": 4.15,
"learning_rate": 1e-05,
"loss": 0.0324,
"step": 1450
},
{
"epoch": 4.23,
"learning_rate": 1e-05,
"loss": 0.0331,
"step": 1475
},
{
"epoch": 4.3,
"learning_rate": 1e-05,
"loss": 0.0339,
"step": 1500
},
{
"epoch": 4.37,
"learning_rate": 1e-05,
"loss": 0.043,
"step": 1525
},
{
"epoch": 4.44,
"learning_rate": 1e-05,
"loss": 0.0402,
"step": 1550
},
{
"epoch": 4.51,
"learning_rate": 1e-05,
"loss": 0.0359,
"step": 1575
},
{
"epoch": 4.58,
"learning_rate": 1e-05,
"loss": 0.0351,
"step": 1600
},
{
"epoch": 4.66,
"learning_rate": 1e-05,
"loss": 0.0381,
"step": 1625
},
{
"epoch": 4.73,
"learning_rate": 1e-05,
"loss": 0.0391,
"step": 1650
},
{
"epoch": 4.8,
"learning_rate": 1e-05,
"loss": 0.0357,
"step": 1675
},
{
"epoch": 4.87,
"learning_rate": 1e-05,
"loss": 0.0372,
"step": 1700
},
{
"epoch": 4.94,
"learning_rate": 1e-05,
"loss": 0.039,
"step": 1725
},
{
"epoch": 5.01,
"learning_rate": 1e-05,
"loss": 0.0318,
"step": 1750
},
{
"epoch": 5.09,
"learning_rate": 1e-05,
"loss": 0.0191,
"step": 1775
},
{
"epoch": 5.16,
"learning_rate": 1e-05,
"loss": 0.0216,
"step": 1800
},
{
"epoch": 5.23,
"learning_rate": 1e-05,
"loss": 0.019,
"step": 1825
},
{
"epoch": 5.3,
"learning_rate": 1e-05,
"loss": 0.0203,
"step": 1850
},
{
"epoch": 5.37,
"learning_rate": 1e-05,
"loss": 0.0198,
"step": 1875
},
{
"epoch": 5.44,
"learning_rate": 1e-05,
"loss": 0.0191,
"step": 1900
},
{
"epoch": 5.52,
"learning_rate": 1e-05,
"loss": 0.018,
"step": 1925
},
{
"epoch": 5.59,
"learning_rate": 1e-05,
"loss": 0.0198,
"step": 1950
},
{
"epoch": 5.66,
"learning_rate": 1e-05,
"loss": 0.0236,
"step": 1975
},
{
"epoch": 5.73,
"learning_rate": 1e-05,
"loss": 0.0198,
"step": 2000
},
{
"epoch": 5.73,
"eval_cer": 9.205097697985662,
"eval_loss": 0.4001982510089874,
"eval_runtime": 712.4165,
"eval_samples_per_second": 1.55,
"eval_steps_per_second": 0.097,
"eval_wer": 16.708526449341033,
"step": 2000
}
],
"max_steps": 4000,
"num_train_epochs": 12,
"total_flos": 9.2174176862208e+18,
"trial_name": null,
"trial_params": null
}