| { | |
| "best_metric": 24.86069836552749, | |
| "best_model_checkpoint": "whisper-el-small-augmented/checkpoint-5000", | |
| "epoch": 23.474178403755868, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.6e-06, | |
| "loss": 2.0417, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 9.600000000000001e-06, | |
| "loss": 0.9587, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.4599999999999999e-05, | |
| "loss": 0.646, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 1.9600000000000002e-05, | |
| "loss": 0.5162, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 2.46e-05, | |
| "loss": 0.3936, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "learning_rate": 2.96e-05, | |
| "loss": 0.3532, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 3.46e-05, | |
| "loss": 0.3432, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "learning_rate": 3.960000000000001e-05, | |
| "loss": 0.3449, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "learning_rate": 4.46e-05, | |
| "loss": 0.2857, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "learning_rate": 4.96e-05, | |
| "loss": 0.2223, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "eval_loss": 0.4402843415737152, | |
| "eval_runtime": 350.2815, | |
| "eval_samples_per_second": 4.842, | |
| "eval_steps_per_second": 0.605, | |
| "eval_wer": 37.992199108469535, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "learning_rate": 4.948888888888889e-05, | |
| "loss": 0.2421, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "learning_rate": 4.8933333333333335e-05, | |
| "loss": 0.255, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "learning_rate": 4.837777777777778e-05, | |
| "loss": 0.2059, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "learning_rate": 4.782222222222222e-05, | |
| "loss": 0.1474, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "learning_rate": 4.726666666666667e-05, | |
| "loss": 0.1376, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "learning_rate": 4.671111111111111e-05, | |
| "loss": 0.1433, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.99, | |
| "learning_rate": 4.615555555555556e-05, | |
| "loss": 0.1527, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 4.23, | |
| "learning_rate": 4.5600000000000004e-05, | |
| "loss": 0.0922, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.46, | |
| "learning_rate": 4.504444444444445e-05, | |
| "loss": 0.0894, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 4.69, | |
| "learning_rate": 4.448888888888889e-05, | |
| "loss": 0.0908, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.69, | |
| "eval_loss": 0.40411776304244995, | |
| "eval_runtime": 397.0474, | |
| "eval_samples_per_second": 4.272, | |
| "eval_steps_per_second": 0.534, | |
| "eval_wer": 35.65193164933135, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.93, | |
| "learning_rate": 4.3933333333333335e-05, | |
| "loss": 0.092, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 5.16, | |
| "learning_rate": 4.337777777777778e-05, | |
| "loss": 0.07, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "learning_rate": 4.282222222222222e-05, | |
| "loss": 0.0631, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 5.63, | |
| "learning_rate": 4.226666666666667e-05, | |
| "loss": 0.0661, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 5.87, | |
| "learning_rate": 4.171111111111111e-05, | |
| "loss": 0.0603, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "learning_rate": 4.115555555555556e-05, | |
| "loss": 0.058, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 6.34, | |
| "learning_rate": 4.0600000000000004e-05, | |
| "loss": 0.0459, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 6.57, | |
| "learning_rate": 4.004444444444445e-05, | |
| "loss": 0.0422, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 6.81, | |
| "learning_rate": 3.948888888888889e-05, | |
| "loss": 0.05, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 7.04, | |
| "learning_rate": 3.8933333333333336e-05, | |
| "loss": 0.0465, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 7.04, | |
| "eval_loss": 0.4188905656337738, | |
| "eval_runtime": 368.3644, | |
| "eval_samples_per_second": 4.604, | |
| "eval_steps_per_second": 0.576, | |
| "eval_wer": 34.30534918276374, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 7.28, | |
| "learning_rate": 3.837777777777778e-05, | |
| "loss": 0.0395, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 7.51, | |
| "learning_rate": 3.782222222222222e-05, | |
| "loss": 0.0285, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 7.75, | |
| "learning_rate": 3.726666666666667e-05, | |
| "loss": 0.0354, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 7.98, | |
| "learning_rate": 3.671111111111111e-05, | |
| "loss": 0.0399, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 8.22, | |
| "learning_rate": 3.615555555555556e-05, | |
| "loss": 0.0303, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 8.45, | |
| "learning_rate": 3.56e-05, | |
| "loss": 0.0232, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 8.69, | |
| "learning_rate": 3.504444444444445e-05, | |
| "loss": 0.0293, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 8.92, | |
| "learning_rate": 3.448888888888889e-05, | |
| "loss": 0.0288, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 9.15, | |
| "learning_rate": 3.3933333333333336e-05, | |
| "loss": 0.0197, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 9.39, | |
| "learning_rate": 3.337777777777778e-05, | |
| "loss": 0.0168, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 9.39, | |
| "eval_loss": 0.3971741199493408, | |
| "eval_runtime": 351.6489, | |
| "eval_samples_per_second": 4.823, | |
| "eval_steps_per_second": 0.603, | |
| "eval_wer": 29.91270430906389, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 9.62, | |
| "learning_rate": 3.2822222222222223e-05, | |
| "loss": 0.022, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 9.86, | |
| "learning_rate": 3.226666666666667e-05, | |
| "loss": 0.021, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 10.09, | |
| "learning_rate": 3.171111111111111e-05, | |
| "loss": 0.0199, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 10.33, | |
| "learning_rate": 3.1155555555555555e-05, | |
| "loss": 0.0132, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 10.56, | |
| "learning_rate": 3.06e-05, | |
| "loss": 0.02, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "learning_rate": 3.004444444444445e-05, | |
| "loss": 0.0164, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 11.03, | |
| "learning_rate": 2.948888888888889e-05, | |
| "loss": 0.0147, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 11.27, | |
| "learning_rate": 2.8933333333333333e-05, | |
| "loss": 0.0121, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 11.5, | |
| "learning_rate": 2.837777777777778e-05, | |
| "loss": 0.0129, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 11.74, | |
| "learning_rate": 2.782222222222222e-05, | |
| "loss": 0.0118, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 11.74, | |
| "eval_loss": 0.40434157848358154, | |
| "eval_runtime": 348.0717, | |
| "eval_samples_per_second": 4.873, | |
| "eval_steps_per_second": 0.609, | |
| "eval_wer": 28.99331352154532, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 11.97, | |
| "learning_rate": 2.7266666666666668e-05, | |
| "loss": 0.0138, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 12.21, | |
| "learning_rate": 2.6711111111111115e-05, | |
| "loss": 0.0075, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 12.44, | |
| "learning_rate": 2.6155555555555555e-05, | |
| "loss": 0.0081, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 12.68, | |
| "learning_rate": 2.5600000000000002e-05, | |
| "loss": 0.0087, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 12.91, | |
| "learning_rate": 2.504444444444445e-05, | |
| "loss": 0.0084, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 13.15, | |
| "learning_rate": 2.448888888888889e-05, | |
| "loss": 0.0059, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 13.38, | |
| "learning_rate": 2.3933333333333337e-05, | |
| "loss": 0.006, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 13.62, | |
| "learning_rate": 2.337777777777778e-05, | |
| "loss": 0.006, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 13.85, | |
| "learning_rate": 2.282222222222222e-05, | |
| "loss": 0.0073, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 14.08, | |
| "learning_rate": 2.2266666666666668e-05, | |
| "loss": 0.0053, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 14.08, | |
| "eval_loss": 0.3968341648578644, | |
| "eval_runtime": 360.6055, | |
| "eval_samples_per_second": 4.703, | |
| "eval_steps_per_second": 0.588, | |
| "eval_wer": 28.593982169390785, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 14.32, | |
| "learning_rate": 2.1711111111111112e-05, | |
| "loss": 0.0048, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 14.55, | |
| "learning_rate": 2.1155555555555556e-05, | |
| "loss": 0.0042, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 14.79, | |
| "learning_rate": 2.06e-05, | |
| "loss": 0.0038, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 15.02, | |
| "learning_rate": 2.0044444444444446e-05, | |
| "loss": 0.005, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 15.26, | |
| "learning_rate": 1.948888888888889e-05, | |
| "loss": 0.0029, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 15.49, | |
| "learning_rate": 1.8933333333333334e-05, | |
| "loss": 0.0067, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 15.73, | |
| "learning_rate": 1.837777777777778e-05, | |
| "loss": 0.0036, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 15.96, | |
| "learning_rate": 1.7822222222222225e-05, | |
| "loss": 0.0029, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 16.2, | |
| "learning_rate": 1.726666666666667e-05, | |
| "loss": 0.0025, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 16.43, | |
| "learning_rate": 1.6711111111111112e-05, | |
| "loss": 0.0032, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 16.43, | |
| "eval_loss": 0.3663780689239502, | |
| "eval_runtime": 347.1158, | |
| "eval_samples_per_second": 4.886, | |
| "eval_steps_per_second": 0.611, | |
| "eval_wer": 25.677934621099553, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 16.67, | |
| "learning_rate": 1.6155555555555556e-05, | |
| "loss": 0.0022, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 16.9, | |
| "learning_rate": 1.56e-05, | |
| "loss": 0.0017, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 17.14, | |
| "learning_rate": 1.5044444444444445e-05, | |
| "loss": 0.0025, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 17.37, | |
| "learning_rate": 1.448888888888889e-05, | |
| "loss": 0.0013, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 17.61, | |
| "learning_rate": 1.3933333333333334e-05, | |
| "loss": 0.0019, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 17.84, | |
| "learning_rate": 1.3377777777777778e-05, | |
| "loss": 0.0017, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 18.08, | |
| "learning_rate": 1.2822222222222222e-05, | |
| "loss": 0.0014, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 18.31, | |
| "learning_rate": 1.2266666666666667e-05, | |
| "loss": 0.0017, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 18.54, | |
| "learning_rate": 1.1711111111111111e-05, | |
| "loss": 0.002, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 18.78, | |
| "learning_rate": 1.1155555555555556e-05, | |
| "loss": 0.0009, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 18.78, | |
| "eval_loss": 0.366485595703125, | |
| "eval_runtime": 364.7578, | |
| "eval_samples_per_second": 4.65, | |
| "eval_steps_per_second": 0.581, | |
| "eval_wer": 26.2444279346211, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 19.01, | |
| "learning_rate": 1.06e-05, | |
| "loss": 0.0019, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 19.25, | |
| "learning_rate": 1.0044444444444446e-05, | |
| "loss": 0.0013, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 19.48, | |
| "learning_rate": 9.48888888888889e-06, | |
| "loss": 0.0005, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 19.72, | |
| "learning_rate": 8.933333333333333e-06, | |
| "loss": 0.0011, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 19.95, | |
| "learning_rate": 8.377777777777779e-06, | |
| "loss": 0.0007, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 20.19, | |
| "learning_rate": 7.822222222222222e-06, | |
| "loss": 0.0013, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 20.42, | |
| "learning_rate": 7.266666666666668e-06, | |
| "loss": 0.0004, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 20.66, | |
| "learning_rate": 6.711111111111111e-06, | |
| "loss": 0.0007, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 20.89, | |
| "learning_rate": 6.155555555555556e-06, | |
| "loss": 0.0006, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 21.13, | |
| "learning_rate": 5.600000000000001e-06, | |
| "loss": 0.0003, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 21.13, | |
| "eval_loss": 0.3619721531867981, | |
| "eval_runtime": 350.8892, | |
| "eval_samples_per_second": 4.833, | |
| "eval_steps_per_second": 0.604, | |
| "eval_wer": 25.28789004457652, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 21.36, | |
| "learning_rate": 5.044444444444444e-06, | |
| "loss": 0.0005, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 21.6, | |
| "learning_rate": 4.488888888888889e-06, | |
| "loss": 0.0005, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 21.83, | |
| "learning_rate": 3.9333333333333335e-06, | |
| "loss": 0.0025, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 22.07, | |
| "learning_rate": 3.3777777777777777e-06, | |
| "loss": 0.0002, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 22.3, | |
| "learning_rate": 2.8222222222222223e-06, | |
| "loss": 0.0002, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 22.54, | |
| "learning_rate": 2.266666666666667e-06, | |
| "loss": 0.0002, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 22.77, | |
| "learning_rate": 1.7111111111111112e-06, | |
| "loss": 0.0003, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "learning_rate": 1.1555555555555556e-06, | |
| "loss": 0.0003, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 23.24, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 0.0002, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 23.47, | |
| "learning_rate": 4.444444444444445e-08, | |
| "loss": 0.0004, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 23.47, | |
| "eval_loss": 0.35695433616638184, | |
| "eval_runtime": 359.944, | |
| "eval_samples_per_second": 4.712, | |
| "eval_steps_per_second": 0.589, | |
| "eval_wer": 24.86069836552749, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 23.47, | |
| "step": 5000, | |
| "total_flos": 2.304700724035584e+19, | |
| "train_loss": 0.08859642370911315, | |
| "train_runtime": 9402.6841, | |
| "train_samples_per_second": 8.508, | |
| "train_steps_per_second": 0.532 | |
| } | |
| ], | |
| "max_steps": 5000, | |
| "num_train_epochs": 24, | |
| "total_flos": 2.304700724035584e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |