whisper-large-v2-hi / trainer_state.json
anuragshas's picture
End of training
75c7b10
raw
history blame
5.62 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.18,
"global_step": 1000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.03,
"learning_rate": 6.712113404111031e-06,
"loss": 0.3544,
"step": 25
},
{
"epoch": 0.05,
"learning_rate": 8.31378915840787e-06,
"loss": 0.1956,
"step": 50
},
{
"epoch": 0.07,
"learning_rate": 9.256291743595376e-06,
"loss": 0.1798,
"step": 75
},
{
"epoch": 0.1,
"learning_rate": 9.911356165197841e-06,
"loss": 0.1901,
"step": 100
},
{
"epoch": 0.12,
"learning_rate": 9.777777777777779e-06,
"loss": 0.1771,
"step": 125
},
{
"epoch": 0.15,
"learning_rate": 9.5e-06,
"loss": 0.151,
"step": 150
},
{
"epoch": 0.17,
"learning_rate": 9.222222222222224e-06,
"loss": 0.1219,
"step": 175
},
{
"epoch": 0.2,
"learning_rate": 8.944444444444446e-06,
"loss": 0.1307,
"step": 200
},
{
"epoch": 1.02,
"learning_rate": 8.666666666666668e-06,
"loss": 0.1204,
"step": 225
},
{
"epoch": 1.04,
"learning_rate": 8.38888888888889e-06,
"loss": 0.097,
"step": 250
},
{
"epoch": 1.07,
"learning_rate": 8.111111111111112e-06,
"loss": 0.0844,
"step": 275
},
{
"epoch": 1.09,
"learning_rate": 7.833333333333333e-06,
"loss": 0.0889,
"step": 300
},
{
"epoch": 1.12,
"learning_rate": 7.555555555555556e-06,
"loss": 0.0807,
"step": 325
},
{
"epoch": 1.15,
"learning_rate": 7.277777777777778e-06,
"loss": 0.072,
"step": 350
},
{
"epoch": 1.17,
"learning_rate": 7e-06,
"loss": 0.0621,
"step": 375
},
{
"epoch": 1.2,
"learning_rate": 6.7222222222222235e-06,
"loss": 0.0598,
"step": 400
},
{
"epoch": 2.02,
"learning_rate": 6.444444444444445e-06,
"loss": 0.0585,
"step": 425
},
{
"epoch": 2.04,
"learning_rate": 6.166666666666667e-06,
"loss": 0.0501,
"step": 450
},
{
"epoch": 2.06,
"learning_rate": 5.9e-06,
"loss": 0.0446,
"step": 475
},
{
"epoch": 2.09,
"learning_rate": 5.622222222222222e-06,
"loss": 0.0395,
"step": 500
},
{
"epoch": 2.12,
"learning_rate": 5.344444444444446e-06,
"loss": 0.0417,
"step": 525
},
{
"epoch": 2.14,
"learning_rate": 5.0666666666666676e-06,
"loss": 0.0359,
"step": 550
},
{
"epoch": 2.17,
"learning_rate": 4.7888888888888894e-06,
"loss": 0.0319,
"step": 575
},
{
"epoch": 2.19,
"learning_rate": 4.511111111111111e-06,
"loss": 0.0304,
"step": 600
},
{
"epoch": 3.01,
"learning_rate": 4.233333333333334e-06,
"loss": 0.0287,
"step": 625
},
{
"epoch": 3.04,
"learning_rate": 3.955555555555556e-06,
"loss": 0.0242,
"step": 650
},
{
"epoch": 3.06,
"learning_rate": 3.6777777777777778e-06,
"loss": 0.0198,
"step": 675
},
{
"epoch": 3.08,
"learning_rate": 3.4000000000000005e-06,
"loss": 0.0182,
"step": 700
},
{
"epoch": 3.11,
"learning_rate": 3.1222222222222228e-06,
"loss": 0.018,
"step": 725
},
{
"epoch": 3.13,
"learning_rate": 2.8444444444444446e-06,
"loss": 0.0154,
"step": 750
},
{
"epoch": 3.16,
"learning_rate": 2.566666666666667e-06,
"loss": 0.0129,
"step": 775
},
{
"epoch": 3.19,
"learning_rate": 2.2888888888888892e-06,
"loss": 0.0134,
"step": 800
},
{
"epoch": 4.0,
"learning_rate": 2.011111111111111e-06,
"loss": 0.0128,
"step": 825
},
{
"epoch": 4.03,
"learning_rate": 1.7333333333333336e-06,
"loss": 0.0074,
"step": 850
},
{
"epoch": 4.05,
"learning_rate": 1.4555555555555557e-06,
"loss": 0.0063,
"step": 875
},
{
"epoch": 4.08,
"learning_rate": 1.1777777777777778e-06,
"loss": 0.0044,
"step": 900
},
{
"epoch": 4.11,
"learning_rate": 9.000000000000001e-07,
"loss": 0.0049,
"step": 925
},
{
"epoch": 4.13,
"learning_rate": 6.222222222222223e-07,
"loss": 0.0056,
"step": 950
},
{
"epoch": 4.16,
"learning_rate": 3.444444444444445e-07,
"loss": 0.0068,
"step": 975
},
{
"epoch": 4.18,
"learning_rate": 6.666666666666668e-08,
"loss": 0.0067,
"step": 1000
},
{
"epoch": 4.18,
"eval_loss": 0.2325439453125,
"eval_runtime": 1399.9283,
"eval_samples_per_second": 2.067,
"eval_steps_per_second": 0.129,
"eval_wer": 10.360772823001557,
"step": 1000
},
{
"epoch": 4.18,
"step": 1000,
"total_flos": 6.777161964753152e+19,
"train_loss": 0.06760481643676758,
"train_runtime": 9362.2814,
"train_samples_per_second": 3.418,
"train_steps_per_second": 0.107
}
],
"max_steps": 1000,
"num_train_epochs": 9223372036854775807,
"total_flos": 6.777161964753152e+19,
"trial_name": null,
"trial_params": null
}