|
{ |
|
"best_metric": 0.46134513589253145, |
|
"best_model_checkpoint": "./whisper-small-ar_tsize_1.0/checkpoint-5000", |
|
"epoch": 1.2883277505797475, |
|
"eval_steps": 1000, |
|
"global_step": 5000, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.01, |
|
"grad_norm": 113.85204315185547, |
|
"learning_rate": 5.000000000000001e-07, |
|
"loss": 3.119, |
|
"step": 25 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"grad_norm": 21.994482040405273, |
|
"learning_rate": 1.0000000000000002e-06, |
|
"loss": 2.7327, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"grad_norm": 17.54548454284668, |
|
"learning_rate": 1.5e-06, |
|
"loss": 1.8873, |
|
"step": 75 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"grad_norm": 19.763961791992188, |
|
"learning_rate": 2.0000000000000003e-06, |
|
"loss": 1.2914, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"grad_norm": 22.205575942993164, |
|
"learning_rate": 2.5e-06, |
|
"loss": 1.0909, |
|
"step": 125 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"grad_norm": 13.473732948303223, |
|
"learning_rate": 3e-06, |
|
"loss": 1.0554, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"grad_norm": 14.15457820892334, |
|
"learning_rate": 3.5e-06, |
|
"loss": 0.9368, |
|
"step": 175 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"grad_norm": 21.088159561157227, |
|
"learning_rate": 4.000000000000001e-06, |
|
"loss": 0.8853, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"grad_norm": 16.64329719543457, |
|
"learning_rate": 4.5e-06, |
|
"loss": 0.8381, |
|
"step": 225 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"grad_norm": 17.755544662475586, |
|
"learning_rate": 5e-06, |
|
"loss": 0.8109, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"grad_norm": 15.66264820098877, |
|
"learning_rate": 5.500000000000001e-06, |
|
"loss": 0.7065, |
|
"step": 275 |
|
}, |
|
{ |
|
"epoch": 0.08, |
|
"grad_norm": 16.0575008392334, |
|
"learning_rate": 6e-06, |
|
"loss": 0.5877, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.08, |
|
"grad_norm": 9.758112907409668, |
|
"learning_rate": 6.5000000000000004e-06, |
|
"loss": 0.465, |
|
"step": 325 |
|
}, |
|
{ |
|
"epoch": 0.09, |
|
"grad_norm": 11.255630493164062, |
|
"learning_rate": 7e-06, |
|
"loss": 0.4381, |
|
"step": 350 |
|
}, |
|
{ |
|
"epoch": 0.1, |
|
"grad_norm": 7.868831157684326, |
|
"learning_rate": 7.500000000000001e-06, |
|
"loss": 0.4518, |
|
"step": 375 |
|
}, |
|
{ |
|
"epoch": 0.1, |
|
"grad_norm": 14.623653411865234, |
|
"learning_rate": 8.000000000000001e-06, |
|
"loss": 0.4146, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.11, |
|
"grad_norm": 12.925063133239746, |
|
"learning_rate": 8.5e-06, |
|
"loss": 0.4376, |
|
"step": 425 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"grad_norm": 11.95840835571289, |
|
"learning_rate": 9e-06, |
|
"loss": 0.4082, |
|
"step": 450 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"grad_norm": 11.043086051940918, |
|
"learning_rate": 9.5e-06, |
|
"loss": 0.4204, |
|
"step": 475 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"grad_norm": 13.569351196289062, |
|
"learning_rate": 1e-05, |
|
"loss": 0.4574, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.14, |
|
"grad_norm": 12.956459999084473, |
|
"learning_rate": 9.944444444444445e-06, |
|
"loss": 0.3668, |
|
"step": 525 |
|
}, |
|
{ |
|
"epoch": 0.14, |
|
"grad_norm": 12.411469459533691, |
|
"learning_rate": 9.88888888888889e-06, |
|
"loss": 0.392, |
|
"step": 550 |
|
}, |
|
{ |
|
"epoch": 0.15, |
|
"grad_norm": 11.523972511291504, |
|
"learning_rate": 9.833333333333333e-06, |
|
"loss": 0.3887, |
|
"step": 575 |
|
}, |
|
{ |
|
"epoch": 0.15, |
|
"grad_norm": 9.563281059265137, |
|
"learning_rate": 9.777777777777779e-06, |
|
"loss": 0.4055, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"grad_norm": 13.318591117858887, |
|
"learning_rate": 9.722222222222223e-06, |
|
"loss": 0.3857, |
|
"step": 625 |
|
}, |
|
{ |
|
"epoch": 0.17, |
|
"grad_norm": 7.878371238708496, |
|
"learning_rate": 9.666666666666667e-06, |
|
"loss": 0.3565, |
|
"step": 650 |
|
}, |
|
{ |
|
"epoch": 0.17, |
|
"grad_norm": 10.200860023498535, |
|
"learning_rate": 9.611111111111112e-06, |
|
"loss": 0.3702, |
|
"step": 675 |
|
}, |
|
{ |
|
"epoch": 0.18, |
|
"grad_norm": 10.580467224121094, |
|
"learning_rate": 9.555555555555556e-06, |
|
"loss": 0.3776, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 0.19, |
|
"grad_norm": 12.521218299865723, |
|
"learning_rate": 9.5e-06, |
|
"loss": 0.3601, |
|
"step": 725 |
|
}, |
|
{ |
|
"epoch": 0.19, |
|
"grad_norm": 13.661670684814453, |
|
"learning_rate": 9.444444444444445e-06, |
|
"loss": 0.3829, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 0.2, |
|
"grad_norm": 12.150840759277344, |
|
"learning_rate": 9.38888888888889e-06, |
|
"loss": 0.3913, |
|
"step": 775 |
|
}, |
|
{ |
|
"epoch": 0.21, |
|
"grad_norm": 11.47882080078125, |
|
"learning_rate": 9.333333333333334e-06, |
|
"loss": 0.4022, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 0.21, |
|
"grad_norm": 10.39133071899414, |
|
"learning_rate": 9.277777777777778e-06, |
|
"loss": 0.3931, |
|
"step": 825 |
|
}, |
|
{ |
|
"epoch": 0.22, |
|
"grad_norm": 9.383455276489258, |
|
"learning_rate": 9.222222222222224e-06, |
|
"loss": 0.3703, |
|
"step": 850 |
|
}, |
|
{ |
|
"epoch": 0.23, |
|
"grad_norm": 11.980205535888672, |
|
"learning_rate": 9.166666666666666e-06, |
|
"loss": 0.3724, |
|
"step": 875 |
|
}, |
|
{ |
|
"epoch": 0.23, |
|
"grad_norm": 16.279930114746094, |
|
"learning_rate": 9.111111111111112e-06, |
|
"loss": 0.3365, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 0.24, |
|
"grad_norm": 8.593803405761719, |
|
"learning_rate": 9.055555555555556e-06, |
|
"loss": 0.3397, |
|
"step": 925 |
|
}, |
|
{ |
|
"epoch": 0.24, |
|
"grad_norm": 12.553435325622559, |
|
"learning_rate": 9e-06, |
|
"loss": 0.3542, |
|
"step": 950 |
|
}, |
|
{ |
|
"epoch": 0.25, |
|
"grad_norm": 9.077679634094238, |
|
"learning_rate": 8.944444444444446e-06, |
|
"loss": 0.3606, |
|
"step": 975 |
|
}, |
|
{ |
|
"epoch": 0.26, |
|
"grad_norm": 8.542105674743652, |
|
"learning_rate": 8.888888888888888e-06, |
|
"loss": 0.3577, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 0.26, |
|
"eval_loss": 0.34235334396362305, |
|
"eval_runtime": 2902.7007, |
|
"eval_samples_per_second": 2.674, |
|
"eval_steps_per_second": 0.335, |
|
"eval_wer": 0.575619411947867, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 0.26, |
|
"grad_norm": 18.058244705200195, |
|
"learning_rate": 8.833333333333334e-06, |
|
"loss": 0.3599, |
|
"step": 1025 |
|
}, |
|
{ |
|
"epoch": 0.27, |
|
"grad_norm": 11.067248344421387, |
|
"learning_rate": 8.777777777777778e-06, |
|
"loss": 0.3432, |
|
"step": 1050 |
|
}, |
|
{ |
|
"epoch": 0.28, |
|
"grad_norm": 12.799885749816895, |
|
"learning_rate": 8.722222222222224e-06, |
|
"loss": 0.3552, |
|
"step": 1075 |
|
}, |
|
{ |
|
"epoch": 0.28, |
|
"grad_norm": 10.083304405212402, |
|
"learning_rate": 8.666666666666668e-06, |
|
"loss": 0.357, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 0.29, |
|
"grad_norm": 11.970526695251465, |
|
"learning_rate": 8.611111111111112e-06, |
|
"loss": 0.323, |
|
"step": 1125 |
|
}, |
|
{ |
|
"epoch": 0.3, |
|
"grad_norm": 10.243406295776367, |
|
"learning_rate": 8.555555555555556e-06, |
|
"loss": 0.3695, |
|
"step": 1150 |
|
}, |
|
{ |
|
"epoch": 0.3, |
|
"grad_norm": 7.486917972564697, |
|
"learning_rate": 8.5e-06, |
|
"loss": 0.3229, |
|
"step": 1175 |
|
}, |
|
{ |
|
"epoch": 0.31, |
|
"grad_norm": 11.434945106506348, |
|
"learning_rate": 8.444444444444446e-06, |
|
"loss": 0.3559, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 0.32, |
|
"grad_norm": 12.598071098327637, |
|
"learning_rate": 8.38888888888889e-06, |
|
"loss": 0.312, |
|
"step": 1225 |
|
}, |
|
{ |
|
"epoch": 0.32, |
|
"grad_norm": 14.671950340270996, |
|
"learning_rate": 8.333333333333334e-06, |
|
"loss": 0.3205, |
|
"step": 1250 |
|
}, |
|
{ |
|
"epoch": 0.33, |
|
"grad_norm": 12.891206741333008, |
|
"learning_rate": 8.277777777777778e-06, |
|
"loss": 0.3112, |
|
"step": 1275 |
|
}, |
|
{ |
|
"epoch": 0.33, |
|
"grad_norm": 7.274331092834473, |
|
"learning_rate": 8.222222222222222e-06, |
|
"loss": 0.3434, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 0.34, |
|
"grad_norm": 7.594877243041992, |
|
"learning_rate": 8.166666666666668e-06, |
|
"loss": 0.3087, |
|
"step": 1325 |
|
}, |
|
{ |
|
"epoch": 0.35, |
|
"grad_norm": 9.791447639465332, |
|
"learning_rate": 8.111111111111112e-06, |
|
"loss": 0.3337, |
|
"step": 1350 |
|
}, |
|
{ |
|
"epoch": 0.35, |
|
"grad_norm": 11.591399192810059, |
|
"learning_rate": 8.055555555555557e-06, |
|
"loss": 0.3143, |
|
"step": 1375 |
|
}, |
|
{ |
|
"epoch": 0.36, |
|
"grad_norm": 10.390214920043945, |
|
"learning_rate": 8.000000000000001e-06, |
|
"loss": 0.3216, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 0.37, |
|
"grad_norm": 16.003393173217773, |
|
"learning_rate": 7.944444444444445e-06, |
|
"loss": 0.3405, |
|
"step": 1425 |
|
}, |
|
{ |
|
"epoch": 0.37, |
|
"grad_norm": 6.049975395202637, |
|
"learning_rate": 7.88888888888889e-06, |
|
"loss": 0.3171, |
|
"step": 1450 |
|
}, |
|
{ |
|
"epoch": 0.38, |
|
"grad_norm": 11.825002670288086, |
|
"learning_rate": 7.833333333333333e-06, |
|
"loss": 0.3348, |
|
"step": 1475 |
|
}, |
|
{ |
|
"epoch": 0.39, |
|
"grad_norm": 9.303126335144043, |
|
"learning_rate": 7.77777777777778e-06, |
|
"loss": 0.3446, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 0.39, |
|
"grad_norm": 11.592827796936035, |
|
"learning_rate": 7.722222222222223e-06, |
|
"loss": 0.3402, |
|
"step": 1525 |
|
}, |
|
{ |
|
"epoch": 0.4, |
|
"grad_norm": 11.242114067077637, |
|
"learning_rate": 7.666666666666667e-06, |
|
"loss": 0.2915, |
|
"step": 1550 |
|
}, |
|
{ |
|
"epoch": 0.41, |
|
"grad_norm": 5.611409664154053, |
|
"learning_rate": 7.611111111111111e-06, |
|
"loss": 0.2787, |
|
"step": 1575 |
|
}, |
|
{ |
|
"epoch": 0.41, |
|
"grad_norm": 11.119479179382324, |
|
"learning_rate": 7.555555555555556e-06, |
|
"loss": 0.32, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 0.42, |
|
"grad_norm": 8.848244667053223, |
|
"learning_rate": 7.500000000000001e-06, |
|
"loss": 0.307, |
|
"step": 1625 |
|
}, |
|
{ |
|
"epoch": 0.43, |
|
"grad_norm": 15.980573654174805, |
|
"learning_rate": 7.444444444444445e-06, |
|
"loss": 0.3652, |
|
"step": 1650 |
|
}, |
|
{ |
|
"epoch": 0.43, |
|
"grad_norm": 13.006916999816895, |
|
"learning_rate": 7.38888888888889e-06, |
|
"loss": 0.2934, |
|
"step": 1675 |
|
}, |
|
{ |
|
"epoch": 0.44, |
|
"grad_norm": 12.559382438659668, |
|
"learning_rate": 7.333333333333333e-06, |
|
"loss": 0.3096, |
|
"step": 1700 |
|
}, |
|
{ |
|
"epoch": 0.44, |
|
"grad_norm": 10.66994857788086, |
|
"learning_rate": 7.277777777777778e-06, |
|
"loss": 0.3007, |
|
"step": 1725 |
|
}, |
|
{ |
|
"epoch": 0.45, |
|
"grad_norm": 11.067940711975098, |
|
"learning_rate": 7.222222222222223e-06, |
|
"loss": 0.2485, |
|
"step": 1750 |
|
}, |
|
{ |
|
"epoch": 0.46, |
|
"grad_norm": 10.337029457092285, |
|
"learning_rate": 7.166666666666667e-06, |
|
"loss": 0.3069, |
|
"step": 1775 |
|
}, |
|
{ |
|
"epoch": 0.46, |
|
"grad_norm": 10.672614097595215, |
|
"learning_rate": 7.111111111111112e-06, |
|
"loss": 0.3061, |
|
"step": 1800 |
|
}, |
|
{ |
|
"epoch": 0.47, |
|
"grad_norm": 15.886311531066895, |
|
"learning_rate": 7.055555555555557e-06, |
|
"loss": 0.3161, |
|
"step": 1825 |
|
}, |
|
{ |
|
"epoch": 0.48, |
|
"grad_norm": 8.174284934997559, |
|
"learning_rate": 7e-06, |
|
"loss": 0.2671, |
|
"step": 1850 |
|
}, |
|
{ |
|
"epoch": 0.48, |
|
"grad_norm": 10.332029342651367, |
|
"learning_rate": 6.944444444444445e-06, |
|
"loss": 0.2884, |
|
"step": 1875 |
|
}, |
|
{ |
|
"epoch": 0.49, |
|
"grad_norm": 8.286322593688965, |
|
"learning_rate": 6.88888888888889e-06, |
|
"loss": 0.2855, |
|
"step": 1900 |
|
}, |
|
{ |
|
"epoch": 0.5, |
|
"grad_norm": 12.562649726867676, |
|
"learning_rate": 6.833333333333334e-06, |
|
"loss": 0.2988, |
|
"step": 1925 |
|
}, |
|
{ |
|
"epoch": 0.5, |
|
"grad_norm": 11.962130546569824, |
|
"learning_rate": 6.777777777777779e-06, |
|
"loss": 0.3186, |
|
"step": 1950 |
|
}, |
|
{ |
|
"epoch": 0.51, |
|
"grad_norm": 6.070617198944092, |
|
"learning_rate": 6.7222222222222235e-06, |
|
"loss": 0.2963, |
|
"step": 1975 |
|
}, |
|
{ |
|
"epoch": 0.52, |
|
"grad_norm": 9.177543640136719, |
|
"learning_rate": 6.666666666666667e-06, |
|
"loss": 0.2801, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 0.52, |
|
"eval_loss": 0.28868672251701355, |
|
"eval_runtime": 2551.4189, |
|
"eval_samples_per_second": 3.042, |
|
"eval_steps_per_second": 0.381, |
|
"eval_wer": 0.5197722521240158, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 0.52, |
|
"grad_norm": 10.780439376831055, |
|
"learning_rate": 6.6111111111111115e-06, |
|
"loss": 0.2512, |
|
"step": 2025 |
|
}, |
|
{ |
|
"epoch": 0.53, |
|
"grad_norm": 11.412617683410645, |
|
"learning_rate": 6.555555555555556e-06, |
|
"loss": 0.2724, |
|
"step": 2050 |
|
}, |
|
{ |
|
"epoch": 0.53, |
|
"grad_norm": 10.96927547454834, |
|
"learning_rate": 6.5000000000000004e-06, |
|
"loss": 0.3125, |
|
"step": 2075 |
|
}, |
|
{ |
|
"epoch": 0.54, |
|
"grad_norm": 14.052020072937012, |
|
"learning_rate": 6.444444444444445e-06, |
|
"loss": 0.2899, |
|
"step": 2100 |
|
}, |
|
{ |
|
"epoch": 0.55, |
|
"grad_norm": 5.423408031463623, |
|
"learning_rate": 6.3888888888888885e-06, |
|
"loss": 0.2857, |
|
"step": 2125 |
|
}, |
|
{ |
|
"epoch": 0.55, |
|
"grad_norm": 10.999767303466797, |
|
"learning_rate": 6.333333333333333e-06, |
|
"loss": 0.2716, |
|
"step": 2150 |
|
}, |
|
{ |
|
"epoch": 0.56, |
|
"grad_norm": 7.399983882904053, |
|
"learning_rate": 6.277777777777778e-06, |
|
"loss": 0.2954, |
|
"step": 2175 |
|
}, |
|
{ |
|
"epoch": 0.57, |
|
"grad_norm": 11.7737455368042, |
|
"learning_rate": 6.222222222222223e-06, |
|
"loss": 0.2872, |
|
"step": 2200 |
|
}, |
|
{ |
|
"epoch": 0.57, |
|
"grad_norm": 10.09986686706543, |
|
"learning_rate": 6.166666666666667e-06, |
|
"loss": 0.2609, |
|
"step": 2225 |
|
}, |
|
{ |
|
"epoch": 0.58, |
|
"grad_norm": 6.141210079193115, |
|
"learning_rate": 6.111111111111112e-06, |
|
"loss": 0.2572, |
|
"step": 2250 |
|
}, |
|
{ |
|
"epoch": 0.59, |
|
"grad_norm": 12.591606140136719, |
|
"learning_rate": 6.055555555555555e-06, |
|
"loss": 0.2913, |
|
"step": 2275 |
|
}, |
|
{ |
|
"epoch": 0.59, |
|
"grad_norm": 9.1317720413208, |
|
"learning_rate": 6e-06, |
|
"loss": 0.2718, |
|
"step": 2300 |
|
}, |
|
{ |
|
"epoch": 0.6, |
|
"grad_norm": 13.439702033996582, |
|
"learning_rate": 5.944444444444445e-06, |
|
"loss": 0.2832, |
|
"step": 2325 |
|
}, |
|
{ |
|
"epoch": 0.61, |
|
"grad_norm": 7.997286796569824, |
|
"learning_rate": 5.88888888888889e-06, |
|
"loss": 0.2515, |
|
"step": 2350 |
|
}, |
|
{ |
|
"epoch": 0.61, |
|
"grad_norm": 14.81768798828125, |
|
"learning_rate": 5.833333333333334e-06, |
|
"loss": 0.2903, |
|
"step": 2375 |
|
}, |
|
{ |
|
"epoch": 0.62, |
|
"grad_norm": 8.115462303161621, |
|
"learning_rate": 5.777777777777778e-06, |
|
"loss": 0.2724, |
|
"step": 2400 |
|
}, |
|
{ |
|
"epoch": 0.62, |
|
"grad_norm": 9.983502388000488, |
|
"learning_rate": 5.722222222222222e-06, |
|
"loss": 0.3023, |
|
"step": 2425 |
|
}, |
|
{ |
|
"epoch": 0.63, |
|
"grad_norm": 5.400686740875244, |
|
"learning_rate": 5.666666666666667e-06, |
|
"loss": 0.2853, |
|
"step": 2450 |
|
}, |
|
{ |
|
"epoch": 0.64, |
|
"grad_norm": 7.283702373504639, |
|
"learning_rate": 5.611111111111112e-06, |
|
"loss": 0.2635, |
|
"step": 2475 |
|
}, |
|
{ |
|
"epoch": 0.64, |
|
"grad_norm": 8.558086395263672, |
|
"learning_rate": 5.555555555555557e-06, |
|
"loss": 0.2282, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 0.65, |
|
"grad_norm": 9.957533836364746, |
|
"learning_rate": 5.500000000000001e-06, |
|
"loss": 0.2873, |
|
"step": 2525 |
|
}, |
|
{ |
|
"epoch": 0.66, |
|
"grad_norm": 9.129793167114258, |
|
"learning_rate": 5.444444444444445e-06, |
|
"loss": 0.2651, |
|
"step": 2550 |
|
}, |
|
{ |
|
"epoch": 0.66, |
|
"grad_norm": 14.834382057189941, |
|
"learning_rate": 5.388888888888889e-06, |
|
"loss": 0.2867, |
|
"step": 2575 |
|
}, |
|
{ |
|
"epoch": 0.67, |
|
"grad_norm": 7.231227874755859, |
|
"learning_rate": 5.333333333333334e-06, |
|
"loss": 0.2546, |
|
"step": 2600 |
|
}, |
|
{ |
|
"epoch": 0.68, |
|
"grad_norm": 9.39655876159668, |
|
"learning_rate": 5.2777777777777785e-06, |
|
"loss": 0.2773, |
|
"step": 2625 |
|
}, |
|
{ |
|
"epoch": 0.68, |
|
"grad_norm": 13.442676544189453, |
|
"learning_rate": 5.2222222222222226e-06, |
|
"loss": 0.2761, |
|
"step": 2650 |
|
}, |
|
{ |
|
"epoch": 0.69, |
|
"grad_norm": 8.369064331054688, |
|
"learning_rate": 5.1666666666666675e-06, |
|
"loss": 0.2525, |
|
"step": 2675 |
|
}, |
|
{ |
|
"epoch": 0.7, |
|
"grad_norm": 10.61405086517334, |
|
"learning_rate": 5.1111111111111115e-06, |
|
"loss": 0.2906, |
|
"step": 2700 |
|
}, |
|
{ |
|
"epoch": 0.7, |
|
"grad_norm": 11.205185890197754, |
|
"learning_rate": 5.0555555555555555e-06, |
|
"loss": 0.2681, |
|
"step": 2725 |
|
}, |
|
{ |
|
"epoch": 0.71, |
|
"grad_norm": 7.688168048858643, |
|
"learning_rate": 5e-06, |
|
"loss": 0.2567, |
|
"step": 2750 |
|
}, |
|
{ |
|
"epoch": 0.72, |
|
"grad_norm": 9.2816162109375, |
|
"learning_rate": 4.944444444444445e-06, |
|
"loss": 0.2692, |
|
"step": 2775 |
|
}, |
|
{ |
|
"epoch": 0.72, |
|
"grad_norm": 11.392304420471191, |
|
"learning_rate": 4.888888888888889e-06, |
|
"loss": 0.2699, |
|
"step": 2800 |
|
}, |
|
{ |
|
"epoch": 0.73, |
|
"grad_norm": 7.2947211265563965, |
|
"learning_rate": 4.833333333333333e-06, |
|
"loss": 0.2547, |
|
"step": 2825 |
|
}, |
|
{ |
|
"epoch": 0.73, |
|
"grad_norm": 9.321547508239746, |
|
"learning_rate": 4.777777777777778e-06, |
|
"loss": 0.2542, |
|
"step": 2850 |
|
}, |
|
{ |
|
"epoch": 0.74, |
|
"grad_norm": 10.214452743530273, |
|
"learning_rate": 4.722222222222222e-06, |
|
"loss": 0.2629, |
|
"step": 2875 |
|
}, |
|
{ |
|
"epoch": 0.75, |
|
"grad_norm": 12.231490135192871, |
|
"learning_rate": 4.666666666666667e-06, |
|
"loss": 0.2816, |
|
"step": 2900 |
|
}, |
|
{ |
|
"epoch": 0.75, |
|
"grad_norm": 13.022787094116211, |
|
"learning_rate": 4.611111111111112e-06, |
|
"loss": 0.2688, |
|
"step": 2925 |
|
}, |
|
{ |
|
"epoch": 0.76, |
|
"grad_norm": 7.756204128265381, |
|
"learning_rate": 4.555555555555556e-06, |
|
"loss": 0.251, |
|
"step": 2950 |
|
}, |
|
{ |
|
"epoch": 0.77, |
|
"grad_norm": 5.8169145584106445, |
|
"learning_rate": 4.5e-06, |
|
"loss": 0.2485, |
|
"step": 2975 |
|
}, |
|
{ |
|
"epoch": 0.77, |
|
"grad_norm": 9.137019157409668, |
|
"learning_rate": 4.444444444444444e-06, |
|
"loss": 0.2804, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 0.77, |
|
"eval_loss": 0.2576030194759369, |
|
"eval_runtime": 2577.0627, |
|
"eval_samples_per_second": 3.012, |
|
"eval_steps_per_second": 0.377, |
|
"eval_wer": 0.48732262799697523, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 0.78, |
|
"grad_norm": 10.181511878967285, |
|
"learning_rate": 4.388888888888889e-06, |
|
"loss": 0.2726, |
|
"step": 3025 |
|
}, |
|
{ |
|
"epoch": 0.79, |
|
"grad_norm": 9.567985534667969, |
|
"learning_rate": 4.333333333333334e-06, |
|
"loss": 0.2648, |
|
"step": 3050 |
|
}, |
|
{ |
|
"epoch": 0.79, |
|
"grad_norm": 9.701679229736328, |
|
"learning_rate": 4.277777777777778e-06, |
|
"loss": 0.2331, |
|
"step": 3075 |
|
}, |
|
{ |
|
"epoch": 0.8, |
|
"grad_norm": 11.3832426071167, |
|
"learning_rate": 4.222222222222223e-06, |
|
"loss": 0.2313, |
|
"step": 3100 |
|
}, |
|
{ |
|
"epoch": 0.81, |
|
"grad_norm": 9.855692863464355, |
|
"learning_rate": 4.166666666666667e-06, |
|
"loss": 0.2713, |
|
"step": 3125 |
|
}, |
|
{ |
|
"epoch": 0.81, |
|
"grad_norm": 15.739023208618164, |
|
"learning_rate": 4.111111111111111e-06, |
|
"loss": 0.2597, |
|
"step": 3150 |
|
}, |
|
{ |
|
"epoch": 0.82, |
|
"grad_norm": 9.282655715942383, |
|
"learning_rate": 4.055555555555556e-06, |
|
"loss": 0.259, |
|
"step": 3175 |
|
}, |
|
{ |
|
"epoch": 0.82, |
|
"grad_norm": 7.786560535430908, |
|
"learning_rate": 4.000000000000001e-06, |
|
"loss": 0.2059, |
|
"step": 3200 |
|
}, |
|
{ |
|
"epoch": 0.83, |
|
"grad_norm": 10.460468292236328, |
|
"learning_rate": 3.944444444444445e-06, |
|
"loss": 0.2479, |
|
"step": 3225 |
|
}, |
|
{ |
|
"epoch": 0.84, |
|
"grad_norm": 6.100764751434326, |
|
"learning_rate": 3.88888888888889e-06, |
|
"loss": 0.2504, |
|
"step": 3250 |
|
}, |
|
{ |
|
"epoch": 0.84, |
|
"grad_norm": 8.303024291992188, |
|
"learning_rate": 3.833333333333334e-06, |
|
"loss": 0.2521, |
|
"step": 3275 |
|
}, |
|
{ |
|
"epoch": 0.85, |
|
"grad_norm": 11.755902290344238, |
|
"learning_rate": 3.777777777777778e-06, |
|
"loss": 0.2559, |
|
"step": 3300 |
|
}, |
|
{ |
|
"epoch": 0.86, |
|
"grad_norm": 6.604350566864014, |
|
"learning_rate": 3.7222222222222225e-06, |
|
"loss": 0.2345, |
|
"step": 3325 |
|
}, |
|
{ |
|
"epoch": 0.86, |
|
"grad_norm": 10.639260292053223, |
|
"learning_rate": 3.6666666666666666e-06, |
|
"loss": 0.2705, |
|
"step": 3350 |
|
}, |
|
{ |
|
"epoch": 0.87, |
|
"grad_norm": 11.366083145141602, |
|
"learning_rate": 3.6111111111111115e-06, |
|
"loss": 0.2138, |
|
"step": 3375 |
|
}, |
|
{ |
|
"epoch": 0.88, |
|
"grad_norm": 8.11689281463623, |
|
"learning_rate": 3.555555555555556e-06, |
|
"loss": 0.2414, |
|
"step": 3400 |
|
}, |
|
{ |
|
"epoch": 0.88, |
|
"grad_norm": 9.813558578491211, |
|
"learning_rate": 3.5e-06, |
|
"loss": 0.2522, |
|
"step": 3425 |
|
}, |
|
{ |
|
"epoch": 0.89, |
|
"grad_norm": 9.525871276855469, |
|
"learning_rate": 3.444444444444445e-06, |
|
"loss": 0.2537, |
|
"step": 3450 |
|
}, |
|
{ |
|
"epoch": 0.9, |
|
"grad_norm": 9.531010627746582, |
|
"learning_rate": 3.3888888888888893e-06, |
|
"loss": 0.2453, |
|
"step": 3475 |
|
}, |
|
{ |
|
"epoch": 0.9, |
|
"grad_norm": 12.289505958557129, |
|
"learning_rate": 3.3333333333333333e-06, |
|
"loss": 0.2243, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 0.91, |
|
"grad_norm": 7.548771858215332, |
|
"learning_rate": 3.277777777777778e-06, |
|
"loss": 0.2562, |
|
"step": 3525 |
|
}, |
|
{ |
|
"epoch": 0.91, |
|
"grad_norm": 10.303834915161133, |
|
"learning_rate": 3.2222222222222227e-06, |
|
"loss": 0.2427, |
|
"step": 3550 |
|
}, |
|
{ |
|
"epoch": 0.92, |
|
"grad_norm": 6.49224328994751, |
|
"learning_rate": 3.1666666666666667e-06, |
|
"loss": 0.2311, |
|
"step": 3575 |
|
}, |
|
{ |
|
"epoch": 0.93, |
|
"grad_norm": 7.009319305419922, |
|
"learning_rate": 3.1111111111111116e-06, |
|
"loss": 0.2164, |
|
"step": 3600 |
|
}, |
|
{ |
|
"epoch": 0.93, |
|
"grad_norm": 6.967477798461914, |
|
"learning_rate": 3.055555555555556e-06, |
|
"loss": 0.2315, |
|
"step": 3625 |
|
}, |
|
{ |
|
"epoch": 0.94, |
|
"grad_norm": 8.28270149230957, |
|
"learning_rate": 3e-06, |
|
"loss": 0.2758, |
|
"step": 3650 |
|
}, |
|
{ |
|
"epoch": 0.95, |
|
"grad_norm": 8.78381633758545, |
|
"learning_rate": 2.944444444444445e-06, |
|
"loss": 0.2611, |
|
"step": 3675 |
|
}, |
|
{ |
|
"epoch": 0.95, |
|
"grad_norm": 7.542636394500732, |
|
"learning_rate": 2.888888888888889e-06, |
|
"loss": 0.2703, |
|
"step": 3700 |
|
}, |
|
{ |
|
"epoch": 0.96, |
|
"grad_norm": 13.231207847595215, |
|
"learning_rate": 2.8333333333333335e-06, |
|
"loss": 0.2286, |
|
"step": 3725 |
|
}, |
|
{ |
|
"epoch": 0.97, |
|
"grad_norm": 11.00623607635498, |
|
"learning_rate": 2.7777777777777783e-06, |
|
"loss": 0.2745, |
|
"step": 3750 |
|
}, |
|
{ |
|
"epoch": 0.97, |
|
"grad_norm": 16.842548370361328, |
|
"learning_rate": 2.7222222222222224e-06, |
|
"loss": 0.2506, |
|
"step": 3775 |
|
}, |
|
{ |
|
"epoch": 0.98, |
|
"grad_norm": 15.040122032165527, |
|
"learning_rate": 2.666666666666667e-06, |
|
"loss": 0.2306, |
|
"step": 3800 |
|
}, |
|
{ |
|
"epoch": 0.99, |
|
"grad_norm": 4.292383193969727, |
|
"learning_rate": 2.6111111111111113e-06, |
|
"loss": 0.2446, |
|
"step": 3825 |
|
}, |
|
{ |
|
"epoch": 0.99, |
|
"grad_norm": 7.2162065505981445, |
|
"learning_rate": 2.5555555555555557e-06, |
|
"loss": 0.2368, |
|
"step": 3850 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"grad_norm": 8.940288543701172, |
|
"learning_rate": 2.5e-06, |
|
"loss": 0.2478, |
|
"step": 3875 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"grad_norm": 10.557374000549316, |
|
"learning_rate": 2.4444444444444447e-06, |
|
"loss": 0.1443, |
|
"step": 3900 |
|
}, |
|
{ |
|
"epoch": 1.01, |
|
"grad_norm": 6.950507640838623, |
|
"learning_rate": 2.388888888888889e-06, |
|
"loss": 0.1581, |
|
"step": 3925 |
|
}, |
|
{ |
|
"epoch": 1.02, |
|
"grad_norm": 7.295450687408447, |
|
"learning_rate": 2.3333333333333336e-06, |
|
"loss": 0.1544, |
|
"step": 3950 |
|
}, |
|
{ |
|
"epoch": 1.02, |
|
"grad_norm": 8.835938453674316, |
|
"learning_rate": 2.277777777777778e-06, |
|
"loss": 0.18, |
|
"step": 3975 |
|
}, |
|
{ |
|
"epoch": 1.03, |
|
"grad_norm": 8.587409019470215, |
|
"learning_rate": 2.222222222222222e-06, |
|
"loss": 0.1791, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 1.03, |
|
"eval_loss": 0.2372235357761383, |
|
"eval_runtime": 2357.7111, |
|
"eval_samples_per_second": 3.292, |
|
"eval_steps_per_second": 0.412, |
|
"eval_wer": 0.47144255148792313, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 1.04, |
|
"grad_norm": 5.195342540740967, |
|
"learning_rate": 2.166666666666667e-06, |
|
"loss": 0.1568, |
|
"step": 4025 |
|
}, |
|
{ |
|
"epoch": 1.04, |
|
"grad_norm": 5.360496520996094, |
|
"learning_rate": 2.1111111111111114e-06, |
|
"loss": 0.1615, |
|
"step": 4050 |
|
}, |
|
{ |
|
"epoch": 1.05, |
|
"grad_norm": 4.146684646606445, |
|
"learning_rate": 2.0555555555555555e-06, |
|
"loss": 0.1424, |
|
"step": 4075 |
|
}, |
|
{ |
|
"epoch": 1.06, |
|
"grad_norm": 8.721048355102539, |
|
"learning_rate": 2.0000000000000003e-06, |
|
"loss": 0.1152, |
|
"step": 4100 |
|
}, |
|
{ |
|
"epoch": 1.06, |
|
"grad_norm": 6.823887348175049, |
|
"learning_rate": 1.944444444444445e-06, |
|
"loss": 0.1571, |
|
"step": 4125 |
|
}, |
|
{ |
|
"epoch": 1.07, |
|
"grad_norm": 7.744790554046631, |
|
"learning_rate": 1.888888888888889e-06, |
|
"loss": 0.1454, |
|
"step": 4150 |
|
}, |
|
{ |
|
"epoch": 1.08, |
|
"grad_norm": 8.660394668579102, |
|
"learning_rate": 1.8333333333333333e-06, |
|
"loss": 0.1559, |
|
"step": 4175 |
|
}, |
|
{ |
|
"epoch": 1.08, |
|
"grad_norm": 6.369543552398682, |
|
"learning_rate": 1.777777777777778e-06, |
|
"loss": 0.1386, |
|
"step": 4200 |
|
}, |
|
{ |
|
"epoch": 1.09, |
|
"grad_norm": 6.839955806732178, |
|
"learning_rate": 1.7222222222222224e-06, |
|
"loss": 0.149, |
|
"step": 4225 |
|
}, |
|
{ |
|
"epoch": 1.1, |
|
"grad_norm": 5.007684230804443, |
|
"learning_rate": 1.6666666666666667e-06, |
|
"loss": 0.1547, |
|
"step": 4250 |
|
}, |
|
{ |
|
"epoch": 1.1, |
|
"grad_norm": 8.356603622436523, |
|
"learning_rate": 1.6111111111111113e-06, |
|
"loss": 0.1625, |
|
"step": 4275 |
|
}, |
|
{ |
|
"epoch": 1.11, |
|
"grad_norm": 4.997384548187256, |
|
"learning_rate": 1.5555555555555558e-06, |
|
"loss": 0.1568, |
|
"step": 4300 |
|
}, |
|
{ |
|
"epoch": 1.11, |
|
"grad_norm": 5.098149299621582, |
|
"learning_rate": 1.5e-06, |
|
"loss": 0.1621, |
|
"step": 4325 |
|
}, |
|
{ |
|
"epoch": 1.12, |
|
"grad_norm": 7.223897457122803, |
|
"learning_rate": 1.4444444444444445e-06, |
|
"loss": 0.1408, |
|
"step": 4350 |
|
}, |
|
{ |
|
"epoch": 1.13, |
|
"grad_norm": 7.733861446380615, |
|
"learning_rate": 1.3888888888888892e-06, |
|
"loss": 0.145, |
|
"step": 4375 |
|
}, |
|
{ |
|
"epoch": 1.13, |
|
"grad_norm": 10.624513626098633, |
|
"learning_rate": 1.3333333333333334e-06, |
|
"loss": 0.1783, |
|
"step": 4400 |
|
}, |
|
{ |
|
"epoch": 1.14, |
|
"grad_norm": 6.46937370300293, |
|
"learning_rate": 1.2777777777777779e-06, |
|
"loss": 0.1575, |
|
"step": 4425 |
|
}, |
|
{ |
|
"epoch": 1.15, |
|
"grad_norm": 6.646976947784424, |
|
"learning_rate": 1.2222222222222223e-06, |
|
"loss": 0.1416, |
|
"step": 4450 |
|
}, |
|
{ |
|
"epoch": 1.15, |
|
"grad_norm": 9.44258975982666, |
|
"learning_rate": 1.1666666666666668e-06, |
|
"loss": 0.136, |
|
"step": 4475 |
|
}, |
|
{ |
|
"epoch": 1.16, |
|
"grad_norm": 5.2407026290893555, |
|
"learning_rate": 1.111111111111111e-06, |
|
"loss": 0.1617, |
|
"step": 4500 |
|
}, |
|
{ |
|
"epoch": 1.17, |
|
"grad_norm": 6.014922142028809, |
|
"learning_rate": 1.0555555555555557e-06, |
|
"loss": 0.144, |
|
"step": 4525 |
|
}, |
|
{ |
|
"epoch": 1.17, |
|
"grad_norm": 6.770383358001709, |
|
"learning_rate": 1.0000000000000002e-06, |
|
"loss": 0.1655, |
|
"step": 4550 |
|
}, |
|
{ |
|
"epoch": 1.18, |
|
"grad_norm": 6.713871955871582, |
|
"learning_rate": 9.444444444444445e-07, |
|
"loss": 0.1619, |
|
"step": 4575 |
|
}, |
|
{ |
|
"epoch": 1.19, |
|
"grad_norm": 6.117084503173828, |
|
"learning_rate": 8.88888888888889e-07, |
|
"loss": 0.1577, |
|
"step": 4600 |
|
}, |
|
{ |
|
"epoch": 1.19, |
|
"grad_norm": 6.789885520935059, |
|
"learning_rate": 8.333333333333333e-07, |
|
"loss": 0.1503, |
|
"step": 4625 |
|
}, |
|
{ |
|
"epoch": 1.2, |
|
"grad_norm": 6.120697975158691, |
|
"learning_rate": 7.777777777777779e-07, |
|
"loss": 0.1752, |
|
"step": 4650 |
|
}, |
|
{ |
|
"epoch": 1.2, |
|
"grad_norm": 7.277804851531982, |
|
"learning_rate": 7.222222222222222e-07, |
|
"loss": 0.1331, |
|
"step": 4675 |
|
}, |
|
{ |
|
"epoch": 1.21, |
|
"grad_norm": 7.9337158203125, |
|
"learning_rate": 6.666666666666667e-07, |
|
"loss": 0.1495, |
|
"step": 4700 |
|
}, |
|
{ |
|
"epoch": 1.22, |
|
"grad_norm": 8.645359992980957, |
|
"learning_rate": 6.111111111111112e-07, |
|
"loss": 0.1416, |
|
"step": 4725 |
|
}, |
|
{ |
|
"epoch": 1.22, |
|
"grad_norm": 5.204892635345459, |
|
"learning_rate": 5.555555555555555e-07, |
|
"loss": 0.1363, |
|
"step": 4750 |
|
}, |
|
{ |
|
"epoch": 1.23, |
|
"grad_norm": 7.200200080871582, |
|
"learning_rate": 5.000000000000001e-07, |
|
"loss": 0.155, |
|
"step": 4775 |
|
}, |
|
{ |
|
"epoch": 1.24, |
|
"grad_norm": 3.2822766304016113, |
|
"learning_rate": 4.444444444444445e-07, |
|
"loss": 0.1556, |
|
"step": 4800 |
|
}, |
|
{ |
|
"epoch": 1.24, |
|
"grad_norm": 8.234893798828125, |
|
"learning_rate": 3.8888888888888895e-07, |
|
"loss": 0.1577, |
|
"step": 4825 |
|
}, |
|
{ |
|
"epoch": 1.25, |
|
"grad_norm": 6.150988578796387, |
|
"learning_rate": 3.3333333333333335e-07, |
|
"loss": 0.1666, |
|
"step": 4850 |
|
}, |
|
{ |
|
"epoch": 1.26, |
|
"grad_norm": 10.273090362548828, |
|
"learning_rate": 2.7777777777777776e-07, |
|
"loss": 0.157, |
|
"step": 4875 |
|
}, |
|
{ |
|
"epoch": 1.26, |
|
"grad_norm": 6.537949562072754, |
|
"learning_rate": 2.2222222222222224e-07, |
|
"loss": 0.1601, |
|
"step": 4900 |
|
}, |
|
{ |
|
"epoch": 1.27, |
|
"grad_norm": 8.629070281982422, |
|
"learning_rate": 1.6666666666666668e-07, |
|
"loss": 0.1381, |
|
"step": 4925 |
|
}, |
|
{ |
|
"epoch": 1.28, |
|
"grad_norm": 9.110799789428711, |
|
"learning_rate": 1.1111111111111112e-07, |
|
"loss": 0.1487, |
|
"step": 4950 |
|
}, |
|
{ |
|
"epoch": 1.28, |
|
"grad_norm": 3.492276191711426, |
|
"learning_rate": 5.555555555555556e-08, |
|
"loss": 0.1514, |
|
"step": 4975 |
|
}, |
|
{ |
|
"epoch": 1.29, |
|
"grad_norm": 8.911520004272461, |
|
"learning_rate": 0.0, |
|
"loss": 0.1529, |
|
"step": 5000 |
|
}, |
|
{ |
|
"epoch": 1.29, |
|
"eval_loss": 0.23124288022518158, |
|
"eval_runtime": 2607.326, |
|
"eval_samples_per_second": 2.977, |
|
"eval_steps_per_second": 0.372, |
|
"eval_wer": 0.46134513589253145, |
|
"step": 5000 |
|
} |
|
], |
|
"logging_steps": 25, |
|
"max_steps": 5000, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 2, |
|
"save_steps": 1000, |
|
"total_flos": 1.154226167119872e+19, |
|
"train_batch_size": 8, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|