whisper-medium-eu / checkpoint-2000 /trainer_state.json
xezpeleta's picture
Training in progress, step 4000
e50d5d9
raw
history blame
10.4 kB
{
"best_metric": 22.052330997610273,
"best_model_checkpoint": "./checkpoint-2000",
"epoch": 0.2857142857142857,
"global_step": 2000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 4.6000000000000004e-07,
"loss": 1.4182,
"step": 25
},
{
"epoch": 0.01,
"learning_rate": 9.400000000000001e-07,
"loss": 1.292,
"step": 50
},
{
"epoch": 0.01,
"learning_rate": 1.44e-06,
"loss": 1.0018,
"step": 75
},
{
"epoch": 0.01,
"learning_rate": 1.94e-06,
"loss": 0.7765,
"step": 100
},
{
"epoch": 0.02,
"learning_rate": 2.4400000000000004e-06,
"loss": 0.7103,
"step": 125
},
{
"epoch": 0.02,
"learning_rate": 2.9400000000000002e-06,
"loss": 0.6597,
"step": 150
},
{
"epoch": 0.03,
"learning_rate": 3.44e-06,
"loss": 0.6657,
"step": 175
},
{
"epoch": 0.03,
"learning_rate": 3.94e-06,
"loss": 0.5853,
"step": 200
},
{
"epoch": 0.03,
"learning_rate": 4.440000000000001e-06,
"loss": 0.5273,
"step": 225
},
{
"epoch": 0.04,
"learning_rate": 4.94e-06,
"loss": 0.5979,
"step": 250
},
{
"epoch": 0.04,
"learning_rate": 5.4400000000000004e-06,
"loss": 0.5861,
"step": 275
},
{
"epoch": 0.04,
"learning_rate": 5.94e-06,
"loss": 0.5085,
"step": 300
},
{
"epoch": 0.05,
"learning_rate": 6.440000000000001e-06,
"loss": 0.4827,
"step": 325
},
{
"epoch": 0.05,
"learning_rate": 6.9400000000000005e-06,
"loss": 0.4909,
"step": 350
},
{
"epoch": 0.05,
"learning_rate": 7.440000000000001e-06,
"loss": 0.4651,
"step": 375
},
{
"epoch": 0.06,
"learning_rate": 7.94e-06,
"loss": 0.494,
"step": 400
},
{
"epoch": 0.06,
"learning_rate": 8.44e-06,
"loss": 0.4188,
"step": 425
},
{
"epoch": 0.06,
"learning_rate": 8.94e-06,
"loss": 0.3849,
"step": 450
},
{
"epoch": 0.07,
"learning_rate": 9.440000000000001e-06,
"loss": 0.4577,
"step": 475
},
{
"epoch": 0.07,
"learning_rate": 9.940000000000001e-06,
"loss": 0.4415,
"step": 500
},
{
"epoch": 0.07,
"learning_rate": 9.966153846153847e-06,
"loss": 0.4614,
"step": 525
},
{
"epoch": 0.08,
"learning_rate": 9.927692307692309e-06,
"loss": 0.4283,
"step": 550
},
{
"epoch": 0.08,
"learning_rate": 9.88923076923077e-06,
"loss": 0.4486,
"step": 575
},
{
"epoch": 0.09,
"learning_rate": 9.850769230769231e-06,
"loss": 0.4434,
"step": 600
},
{
"epoch": 0.09,
"learning_rate": 9.812307692307694e-06,
"loss": 0.4245,
"step": 625
},
{
"epoch": 0.09,
"learning_rate": 9.773846153846154e-06,
"loss": 0.4254,
"step": 650
},
{
"epoch": 0.1,
"learning_rate": 9.735384615384616e-06,
"loss": 0.4143,
"step": 675
},
{
"epoch": 0.1,
"learning_rate": 9.696923076923078e-06,
"loss": 0.4115,
"step": 700
},
{
"epoch": 0.1,
"learning_rate": 9.658461538461539e-06,
"loss": 0.3656,
"step": 725
},
{
"epoch": 0.11,
"learning_rate": 9.620000000000001e-06,
"loss": 0.4004,
"step": 750
},
{
"epoch": 0.11,
"learning_rate": 9.581538461538462e-06,
"loss": 0.388,
"step": 775
},
{
"epoch": 0.11,
"learning_rate": 9.543076923076924e-06,
"loss": 0.3665,
"step": 800
},
{
"epoch": 0.12,
"learning_rate": 9.504615384615386e-06,
"loss": 0.3777,
"step": 825
},
{
"epoch": 0.12,
"learning_rate": 9.466153846153846e-06,
"loss": 0.4338,
"step": 850
},
{
"epoch": 0.12,
"learning_rate": 9.427692307692309e-06,
"loss": 0.4017,
"step": 875
},
{
"epoch": 0.13,
"learning_rate": 9.38923076923077e-06,
"loss": 0.3757,
"step": 900
},
{
"epoch": 0.13,
"learning_rate": 9.350769230769231e-06,
"loss": 0.4084,
"step": 925
},
{
"epoch": 0.14,
"learning_rate": 9.312307692307693e-06,
"loss": 0.3821,
"step": 950
},
{
"epoch": 0.14,
"learning_rate": 9.273846153846154e-06,
"loss": 0.421,
"step": 975
},
{
"epoch": 0.14,
"learning_rate": 9.235384615384616e-06,
"loss": 0.4203,
"step": 1000
},
{
"epoch": 0.14,
"eval_loss": 0.4127572178840637,
"eval_runtime": 1802.5775,
"eval_samples_per_second": 3.656,
"eval_steps_per_second": 0.457,
"eval_wer": 28.265624367127064,
"step": 1000
},
{
"epoch": 0.15,
"learning_rate": 9.196923076923078e-06,
"loss": 0.4094,
"step": 1025
},
{
"epoch": 0.15,
"learning_rate": 9.158461538461539e-06,
"loss": 0.3674,
"step": 1050
},
{
"epoch": 0.15,
"learning_rate": 9.12e-06,
"loss": 0.3613,
"step": 1075
},
{
"epoch": 0.16,
"learning_rate": 9.081538461538463e-06,
"loss": 0.3904,
"step": 1100
},
{
"epoch": 0.16,
"learning_rate": 9.043076923076923e-06,
"loss": 0.3865,
"step": 1125
},
{
"epoch": 0.16,
"learning_rate": 9.004615384615386e-06,
"loss": 0.3677,
"step": 1150
},
{
"epoch": 0.17,
"learning_rate": 8.966153846153846e-06,
"loss": 0.3295,
"step": 1175
},
{
"epoch": 0.17,
"learning_rate": 8.927692307692308e-06,
"loss": 0.3207,
"step": 1200
},
{
"epoch": 0.17,
"learning_rate": 8.88923076923077e-06,
"loss": 0.2817,
"step": 1225
},
{
"epoch": 0.18,
"learning_rate": 8.850769230769231e-06,
"loss": 0.3218,
"step": 1250
},
{
"epoch": 0.18,
"learning_rate": 8.812307692307693e-06,
"loss": 0.3502,
"step": 1275
},
{
"epoch": 0.19,
"learning_rate": 8.773846153846155e-06,
"loss": 0.3432,
"step": 1300
},
{
"epoch": 0.19,
"learning_rate": 8.735384615384616e-06,
"loss": 0.3244,
"step": 1325
},
{
"epoch": 0.19,
"learning_rate": 8.696923076923078e-06,
"loss": 0.2607,
"step": 1350
},
{
"epoch": 0.2,
"learning_rate": 8.658461538461538e-06,
"loss": 0.3469,
"step": 1375
},
{
"epoch": 0.2,
"learning_rate": 8.62e-06,
"loss": 0.2756,
"step": 1400
},
{
"epoch": 0.2,
"learning_rate": 8.581538461538463e-06,
"loss": 0.294,
"step": 1425
},
{
"epoch": 0.21,
"learning_rate": 8.543076923076923e-06,
"loss": 0.3915,
"step": 1450
},
{
"epoch": 0.21,
"learning_rate": 8.504615384615385e-06,
"loss": 0.3237,
"step": 1475
},
{
"epoch": 0.21,
"learning_rate": 8.466153846153847e-06,
"loss": 0.2723,
"step": 1500
},
{
"epoch": 0.22,
"learning_rate": 8.427692307692308e-06,
"loss": 0.2917,
"step": 1525
},
{
"epoch": 0.22,
"learning_rate": 8.38923076923077e-06,
"loss": 0.2921,
"step": 1550
},
{
"epoch": 0.23,
"learning_rate": 8.35076923076923e-06,
"loss": 0.3724,
"step": 1575
},
{
"epoch": 0.23,
"learning_rate": 8.312307692307693e-06,
"loss": 0.3364,
"step": 1600
},
{
"epoch": 0.23,
"learning_rate": 8.273846153846155e-06,
"loss": 0.3661,
"step": 1625
},
{
"epoch": 0.24,
"learning_rate": 8.235384615384615e-06,
"loss": 0.3307,
"step": 1650
},
{
"epoch": 0.24,
"learning_rate": 8.196923076923078e-06,
"loss": 0.3201,
"step": 1675
},
{
"epoch": 0.24,
"learning_rate": 8.15846153846154e-06,
"loss": 0.2973,
"step": 1700
},
{
"epoch": 0.25,
"learning_rate": 8.120000000000002e-06,
"loss": 0.3013,
"step": 1725
},
{
"epoch": 0.25,
"learning_rate": 8.081538461538462e-06,
"loss": 0.3349,
"step": 1750
},
{
"epoch": 0.25,
"learning_rate": 8.043076923076923e-06,
"loss": 0.3244,
"step": 1775
},
{
"epoch": 0.26,
"learning_rate": 8.004615384615385e-06,
"loss": 0.3331,
"step": 1800
},
{
"epoch": 0.26,
"learning_rate": 7.966153846153847e-06,
"loss": 0.2723,
"step": 1825
},
{
"epoch": 0.26,
"learning_rate": 7.927692307692308e-06,
"loss": 0.316,
"step": 1850
},
{
"epoch": 0.27,
"learning_rate": 7.88923076923077e-06,
"loss": 0.2648,
"step": 1875
},
{
"epoch": 0.27,
"learning_rate": 7.850769230769232e-06,
"loss": 0.2965,
"step": 1900
},
{
"epoch": 0.28,
"learning_rate": 7.812307692307694e-06,
"loss": 0.256,
"step": 1925
},
{
"epoch": 0.28,
"learning_rate": 7.773846153846155e-06,
"loss": 0.2156,
"step": 1950
},
{
"epoch": 0.28,
"learning_rate": 7.735384615384615e-06,
"loss": 0.2418,
"step": 1975
},
{
"epoch": 0.29,
"learning_rate": 7.696923076923077e-06,
"loss": 0.2693,
"step": 2000
},
{
"epoch": 0.29,
"eval_loss": 0.3239877223968506,
"eval_runtime": 1817.7748,
"eval_samples_per_second": 3.626,
"eval_steps_per_second": 0.453,
"eval_wer": 22.052330997610273,
"step": 2000
}
],
"max_steps": 7000,
"num_train_epochs": 9223372036854775807,
"total_flos": 8.16483926016e+18,
"trial_name": null,
"trial_params": null
}