|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 158.80044407438245, |
|
"global_step": 286000, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 1.11, |
|
"learning_rate": 9.968592592592593e-05, |
|
"loss": 1.9211, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 1.11, |
|
"eval_loss": 0.3011658489704132, |
|
"eval_runtime": 109.7461, |
|
"eval_samples_per_second": 116.214, |
|
"eval_steps_per_second": 3.636, |
|
"eval_wer": 0.4120684023643701, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 2.22, |
|
"learning_rate": 9.931555555555556e-05, |
|
"loss": 0.9582, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 2.22, |
|
"eval_loss": 0.24203644692897797, |
|
"eval_runtime": 109.3338, |
|
"eval_samples_per_second": 116.652, |
|
"eval_steps_per_second": 3.649, |
|
"eval_wer": 0.3336880590881792, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 3.33, |
|
"learning_rate": 9.894518518518519e-05, |
|
"loss": 0.8899, |
|
"step": 6000 |
|
}, |
|
{ |
|
"epoch": 3.33, |
|
"eval_loss": 0.21842752397060394, |
|
"eval_runtime": 109.0188, |
|
"eval_samples_per_second": 116.989, |
|
"eval_steps_per_second": 3.66, |
|
"eval_wer": 0.3000136973311277, |
|
"step": 6000 |
|
}, |
|
{ |
|
"epoch": 4.44, |
|
"learning_rate": 9.857500000000001e-05, |
|
"loss": 0.8488, |
|
"step": 8000 |
|
}, |
|
{ |
|
"epoch": 4.44, |
|
"eval_loss": 0.20210157334804535, |
|
"eval_runtime": 108.438, |
|
"eval_samples_per_second": 117.616, |
|
"eval_steps_per_second": 3.68, |
|
"eval_wer": 0.28011042156170646, |
|
"step": 8000 |
|
}, |
|
{ |
|
"epoch": 5.55, |
|
"learning_rate": 9.820481481481482e-05, |
|
"loss": 0.8243, |
|
"step": 10000 |
|
}, |
|
{ |
|
"epoch": 5.55, |
|
"eval_loss": 0.19518321752548218, |
|
"eval_runtime": 109.2218, |
|
"eval_samples_per_second": 116.772, |
|
"eval_steps_per_second": 3.653, |
|
"eval_wer": 0.2646956558387508, |
|
"step": 10000 |
|
}, |
|
{ |
|
"epoch": 6.66, |
|
"learning_rate": 9.783462962962964e-05, |
|
"loss": 0.8051, |
|
"step": 12000 |
|
}, |
|
{ |
|
"epoch": 6.66, |
|
"eval_loss": 0.18829651176929474, |
|
"eval_runtime": 110.5038, |
|
"eval_samples_per_second": 115.417, |
|
"eval_steps_per_second": 3.611, |
|
"eval_wer": 0.25592936391701526, |
|
"step": 12000 |
|
}, |
|
{ |
|
"epoch": 7.77, |
|
"learning_rate": 9.746444444444445e-05, |
|
"loss": 0.7875, |
|
"step": 14000 |
|
}, |
|
{ |
|
"epoch": 7.77, |
|
"eval_loss": 0.18309463560581207, |
|
"eval_runtime": 109.5923, |
|
"eval_samples_per_second": 116.377, |
|
"eval_steps_per_second": 3.641, |
|
"eval_wer": 0.24850119588237154, |
|
"step": 14000 |
|
}, |
|
{ |
|
"epoch": 8.88, |
|
"learning_rate": 9.709407407407408e-05, |
|
"loss": 0.772, |
|
"step": 16000 |
|
}, |
|
{ |
|
"epoch": 8.88, |
|
"eval_loss": 0.17918168008327484, |
|
"eval_runtime": 109.8993, |
|
"eval_samples_per_second": 116.052, |
|
"eval_steps_per_second": 3.631, |
|
"eval_wer": 0.24318030955968348, |
|
"step": 16000 |
|
}, |
|
{ |
|
"epoch": 9.99, |
|
"learning_rate": 9.672388888888889e-05, |
|
"loss": 0.7597, |
|
"step": 18000 |
|
}, |
|
{ |
|
"epoch": 9.99, |
|
"eval_loss": 0.17556481063365936, |
|
"eval_runtime": 109.4542, |
|
"eval_samples_per_second": 116.524, |
|
"eval_steps_per_second": 3.645, |
|
"eval_wer": 0.23213815338903582, |
|
"step": 18000 |
|
}, |
|
{ |
|
"epoch": 11.1, |
|
"learning_rate": 9.635351851851853e-05, |
|
"loss": 0.7462, |
|
"step": 20000 |
|
}, |
|
{ |
|
"epoch": 11.1, |
|
"eval_loss": 0.17337526381015778, |
|
"eval_runtime": 107.9125, |
|
"eval_samples_per_second": 118.188, |
|
"eval_steps_per_second": 3.697, |
|
"eval_wer": 0.22965156096892814, |
|
"step": 20000 |
|
}, |
|
{ |
|
"epoch": 12.22, |
|
"learning_rate": 9.598333333333334e-05, |
|
"loss": 0.7375, |
|
"step": 22000 |
|
}, |
|
{ |
|
"epoch": 12.22, |
|
"eval_loss": 0.16925489902496338, |
|
"eval_runtime": 109.8662, |
|
"eval_samples_per_second": 116.087, |
|
"eval_steps_per_second": 3.632, |
|
"eval_wer": 0.22406726443224562, |
|
"step": 22000 |
|
}, |
|
{ |
|
"epoch": 13.33, |
|
"learning_rate": 9.561314814814816e-05, |
|
"loss": 0.7266, |
|
"step": 24000 |
|
}, |
|
{ |
|
"epoch": 13.33, |
|
"eval_loss": 0.16615571081638336, |
|
"eval_runtime": 108.8823, |
|
"eval_samples_per_second": 117.136, |
|
"eval_steps_per_second": 3.665, |
|
"eval_wer": 0.22065346805887745, |
|
"step": 24000 |
|
}, |
|
{ |
|
"epoch": 14.44, |
|
"learning_rate": 9.524277777777778e-05, |
|
"loss": 0.7181, |
|
"step": 26000 |
|
}, |
|
{ |
|
"epoch": 14.44, |
|
"eval_loss": 0.16478531062602997, |
|
"eval_runtime": 108.8044, |
|
"eval_samples_per_second": 117.22, |
|
"eval_steps_per_second": 3.667, |
|
"eval_wer": 0.21787185619909596, |
|
"step": 26000 |
|
}, |
|
{ |
|
"epoch": 15.55, |
|
"learning_rate": 9.487277777777778e-05, |
|
"loss": 0.7143, |
|
"step": 28000 |
|
}, |
|
{ |
|
"epoch": 15.55, |
|
"eval_loss": 0.16348408162593842, |
|
"eval_runtime": 109.4918, |
|
"eval_samples_per_second": 116.484, |
|
"eval_steps_per_second": 3.644, |
|
"eval_wer": 0.21597530265833587, |
|
"step": 28000 |
|
}, |
|
{ |
|
"epoch": 16.66, |
|
"learning_rate": 9.45025925925926e-05, |
|
"loss": 0.7032, |
|
"step": 30000 |
|
}, |
|
{ |
|
"epoch": 16.66, |
|
"eval_loss": 0.16043423116207123, |
|
"eval_runtime": 109.1929, |
|
"eval_samples_per_second": 116.802, |
|
"eval_steps_per_second": 3.654, |
|
"eval_wer": 0.21385748453782044, |
|
"step": 30000 |
|
}, |
|
{ |
|
"epoch": 17.77, |
|
"learning_rate": 9.413222222222222e-05, |
|
"loss": 0.6988, |
|
"step": 32000 |
|
}, |
|
{ |
|
"epoch": 17.77, |
|
"eval_loss": 0.16011235117912292, |
|
"eval_runtime": 109.5116, |
|
"eval_samples_per_second": 116.463, |
|
"eval_steps_per_second": 3.643, |
|
"eval_wer": 0.21078085323836518, |
|
"step": 32000 |
|
}, |
|
{ |
|
"epoch": 18.88, |
|
"learning_rate": 9.376203703703704e-05, |
|
"loss": 0.6896, |
|
"step": 34000 |
|
}, |
|
{ |
|
"epoch": 18.88, |
|
"eval_loss": 0.15551629662513733, |
|
"eval_runtime": 109.5231, |
|
"eval_samples_per_second": 116.45, |
|
"eval_steps_per_second": 3.643, |
|
"eval_wer": 0.20503851057328598, |
|
"step": 34000 |
|
}, |
|
{ |
|
"epoch": 19.99, |
|
"learning_rate": 9.339166666666667e-05, |
|
"loss": 0.6837, |
|
"step": 36000 |
|
}, |
|
{ |
|
"epoch": 19.99, |
|
"eval_loss": 0.15630744397640228, |
|
"eval_runtime": 111.1508, |
|
"eval_samples_per_second": 114.745, |
|
"eval_steps_per_second": 3.59, |
|
"eval_wer": 0.20782012243306747, |
|
"step": 36000 |
|
}, |
|
{ |
|
"epoch": 21.1, |
|
"learning_rate": 9.302148148148149e-05, |
|
"loss": 0.6779, |
|
"step": 38000 |
|
}, |
|
{ |
|
"epoch": 21.1, |
|
"eval_loss": 0.15380945801734924, |
|
"eval_runtime": 108.7667, |
|
"eval_samples_per_second": 117.26, |
|
"eval_steps_per_second": 3.668, |
|
"eval_wer": 0.2020250977251894, |
|
"step": 38000 |
|
}, |
|
{ |
|
"epoch": 22.21, |
|
"learning_rate": 9.26512962962963e-05, |
|
"loss": 0.6749, |
|
"step": 40000 |
|
}, |
|
{ |
|
"epoch": 22.21, |
|
"eval_loss": 0.15313279628753662, |
|
"eval_runtime": 109.8109, |
|
"eval_samples_per_second": 116.145, |
|
"eval_steps_per_second": 3.634, |
|
"eval_wer": 0.20110843018048868, |
|
"step": 40000 |
|
}, |
|
{ |
|
"epoch": 23.32, |
|
"learning_rate": 9.228111111111112e-05, |
|
"loss": 0.6704, |
|
"step": 42000 |
|
}, |
|
{ |
|
"epoch": 23.32, |
|
"eval_loss": 0.15281446278095245, |
|
"eval_runtime": 109.3199, |
|
"eval_samples_per_second": 116.667, |
|
"eval_steps_per_second": 3.65, |
|
"eval_wer": 0.2010978937719289, |
|
"step": 42000 |
|
}, |
|
{ |
|
"epoch": 24.43, |
|
"learning_rate": 9.191074074074074e-05, |
|
"loss": 0.6669, |
|
"step": 44000 |
|
}, |
|
{ |
|
"epoch": 24.43, |
|
"eval_loss": 0.15204770863056183, |
|
"eval_runtime": 109.4873, |
|
"eval_samples_per_second": 116.488, |
|
"eval_steps_per_second": 3.644, |
|
"eval_wer": 0.19591398076051797, |
|
"step": 44000 |
|
}, |
|
{ |
|
"epoch": 25.54, |
|
"learning_rate": 9.154037037037038e-05, |
|
"loss": 0.6594, |
|
"step": 46000 |
|
}, |
|
{ |
|
"epoch": 25.54, |
|
"eval_loss": 0.14905157685279846, |
|
"eval_runtime": 108.5995, |
|
"eval_samples_per_second": 117.441, |
|
"eval_steps_per_second": 3.674, |
|
"eval_wer": 0.19542930596676816, |
|
"step": 46000 |
|
}, |
|
{ |
|
"epoch": 26.65, |
|
"learning_rate": 9.117018518518519e-05, |
|
"loss": 0.6542, |
|
"step": 48000 |
|
}, |
|
{ |
|
"epoch": 26.65, |
|
"eval_loss": 0.14924176037311554, |
|
"eval_runtime": 110.1651, |
|
"eval_samples_per_second": 115.772, |
|
"eval_steps_per_second": 3.622, |
|
"eval_wer": 0.19538716033252906, |
|
"step": 48000 |
|
}, |
|
{ |
|
"epoch": 27.76, |
|
"learning_rate": 9.079981481481482e-05, |
|
"loss": 0.652, |
|
"step": 50000 |
|
}, |
|
{ |
|
"epoch": 27.76, |
|
"eval_loss": 0.14801371097564697, |
|
"eval_runtime": 109.7651, |
|
"eval_samples_per_second": 116.194, |
|
"eval_steps_per_second": 3.635, |
|
"eval_wer": 0.19368026214584497, |
|
"step": 50000 |
|
}, |
|
{ |
|
"epoch": 28.87, |
|
"learning_rate": 9.042962962962963e-05, |
|
"loss": 0.6474, |
|
"step": 52000 |
|
}, |
|
{ |
|
"epoch": 28.87, |
|
"eval_loss": 0.14799948036670685, |
|
"eval_runtime": 110.6825, |
|
"eval_samples_per_second": 115.231, |
|
"eval_steps_per_second": 3.605, |
|
"eval_wer": 0.192394820301552, |
|
"step": 52000 |
|
}, |
|
{ |
|
"epoch": 29.98, |
|
"learning_rate": 9.005944444444445e-05, |
|
"loss": 0.6443, |
|
"step": 54000 |
|
}, |
|
{ |
|
"epoch": 29.98, |
|
"eval_loss": 0.14905065298080444, |
|
"eval_runtime": 110.5602, |
|
"eval_samples_per_second": 115.358, |
|
"eval_steps_per_second": 3.609, |
|
"eval_wer": 0.19057202162071035, |
|
"step": 54000 |
|
}, |
|
{ |
|
"epoch": 31.09, |
|
"learning_rate": 8.968907407407408e-05, |
|
"loss": 0.6407, |
|
"step": 56000 |
|
}, |
|
{ |
|
"epoch": 31.09, |
|
"eval_loss": 0.14624294638633728, |
|
"eval_runtime": 108.1653, |
|
"eval_samples_per_second": 117.912, |
|
"eval_steps_per_second": 3.689, |
|
"eval_wer": 0.19001359196704212, |
|
"step": 56000 |
|
}, |
|
{ |
|
"epoch": 32.2, |
|
"learning_rate": 8.931907407407408e-05, |
|
"loss": 0.6349, |
|
"step": 58000 |
|
}, |
|
{ |
|
"epoch": 32.2, |
|
"eval_loss": 0.14455664157867432, |
|
"eval_runtime": 109.1496, |
|
"eval_samples_per_second": 116.849, |
|
"eval_steps_per_second": 3.656, |
|
"eval_wer": 0.18883351420834693, |
|
"step": 58000 |
|
}, |
|
{ |
|
"epoch": 33.31, |
|
"learning_rate": 8.894870370370371e-05, |
|
"loss": 0.6337, |
|
"step": 60000 |
|
}, |
|
{ |
|
"epoch": 33.31, |
|
"eval_loss": 0.1445704698562622, |
|
"eval_runtime": 109.4136, |
|
"eval_samples_per_second": 116.567, |
|
"eval_steps_per_second": 3.647, |
|
"eval_wer": 0.18898102392818383, |
|
"step": 60000 |
|
}, |
|
{ |
|
"epoch": 34.43, |
|
"learning_rate": 8.85787037037037e-05, |
|
"loss": 0.6296, |
|
"step": 62000 |
|
}, |
|
{ |
|
"epoch": 34.43, |
|
"eval_loss": 0.14853893220424652, |
|
"eval_runtime": 108.9606, |
|
"eval_samples_per_second": 117.052, |
|
"eval_steps_per_second": 3.662, |
|
"eval_wer": 0.18811703842628202, |
|
"step": 62000 |
|
}, |
|
{ |
|
"epoch": 35.54, |
|
"learning_rate": 8.820833333333334e-05, |
|
"loss": 0.6249, |
|
"step": 64000 |
|
}, |
|
{ |
|
"epoch": 35.54, |
|
"eval_loss": 0.14442408084869385, |
|
"eval_runtime": 110.1343, |
|
"eval_samples_per_second": 115.804, |
|
"eval_steps_per_second": 3.623, |
|
"eval_wer": 0.18499826149258763, |
|
"step": 64000 |
|
}, |
|
{ |
|
"epoch": 36.65, |
|
"learning_rate": 8.783814814814815e-05, |
|
"loss": 0.6222, |
|
"step": 66000 |
|
}, |
|
{ |
|
"epoch": 36.65, |
|
"eval_loss": 0.14410334825515747, |
|
"eval_runtime": 109.3816, |
|
"eval_samples_per_second": 116.601, |
|
"eval_steps_per_second": 3.648, |
|
"eval_wer": 0.18493504304122896, |
|
"step": 66000 |
|
}, |
|
{ |
|
"epoch": 37.76, |
|
"learning_rate": 8.746796296296297e-05, |
|
"loss": 0.6173, |
|
"step": 68000 |
|
}, |
|
{ |
|
"epoch": 37.76, |
|
"eval_loss": 0.14380425214767456, |
|
"eval_runtime": 111.8352, |
|
"eval_samples_per_second": 114.043, |
|
"eval_steps_per_second": 3.568, |
|
"eval_wer": 0.18183733892465415, |
|
"step": 68000 |
|
}, |
|
{ |
|
"epoch": 38.87, |
|
"learning_rate": 8.709759259259259e-05, |
|
"loss": 0.6149, |
|
"step": 70000 |
|
}, |
|
{ |
|
"epoch": 38.87, |
|
"eval_loss": 0.14080771803855896, |
|
"eval_runtime": 110.5146, |
|
"eval_samples_per_second": 115.406, |
|
"eval_steps_per_second": 3.61, |
|
"eval_wer": 0.18231147730984418, |
|
"step": 70000 |
|
}, |
|
{ |
|
"epoch": 39.98, |
|
"learning_rate": 8.672740740740741e-05, |
|
"loss": 0.6114, |
|
"step": 72000 |
|
}, |
|
{ |
|
"epoch": 39.98, |
|
"eval_loss": 0.1435374915599823, |
|
"eval_runtime": 110.4812, |
|
"eval_samples_per_second": 115.44, |
|
"eval_steps_per_second": 3.611, |
|
"eval_wer": 0.18260649674951795, |
|
"step": 72000 |
|
}, |
|
{ |
|
"epoch": 41.09, |
|
"learning_rate": 8.635722222222222e-05, |
|
"loss": 0.6089, |
|
"step": 74000 |
|
}, |
|
{ |
|
"epoch": 41.09, |
|
"eval_loss": 0.14298424124717712, |
|
"eval_runtime": 107.7167, |
|
"eval_samples_per_second": 118.403, |
|
"eval_steps_per_second": 3.704, |
|
"eval_wer": 0.18134212772234456, |
|
"step": 74000 |
|
}, |
|
{ |
|
"epoch": 42.2, |
|
"learning_rate": 8.598722222222223e-05, |
|
"loss": 0.6066, |
|
"step": 76000 |
|
}, |
|
{ |
|
"epoch": 42.2, |
|
"eval_loss": 0.14788854122161865, |
|
"eval_runtime": 108.9238, |
|
"eval_samples_per_second": 117.091, |
|
"eval_steps_per_second": 3.663, |
|
"eval_wer": 0.18113139955114899, |
|
"step": 76000 |
|
}, |
|
{ |
|
"epoch": 43.31, |
|
"learning_rate": 8.561703703703704e-05, |
|
"loss": 0.6012, |
|
"step": 78000 |
|
}, |
|
{ |
|
"epoch": 43.31, |
|
"eval_loss": 0.1424262672662735, |
|
"eval_runtime": 109.6315, |
|
"eval_samples_per_second": 116.335, |
|
"eval_steps_per_second": 3.639, |
|
"eval_wer": 0.18278561569503418, |
|
"step": 78000 |
|
}, |
|
{ |
|
"epoch": 44.42, |
|
"learning_rate": 8.524666666666667e-05, |
|
"loss": 0.6009, |
|
"step": 80000 |
|
}, |
|
{ |
|
"epoch": 44.42, |
|
"eval_loss": 0.1408558338880539, |
|
"eval_runtime": 109.6234, |
|
"eval_samples_per_second": 116.344, |
|
"eval_steps_per_second": 3.64, |
|
"eval_wer": 0.17987756693253537, |
|
"step": 80000 |
|
}, |
|
{ |
|
"epoch": 45.53, |
|
"learning_rate": 8.487648148148148e-05, |
|
"loss": 0.5976, |
|
"step": 82000 |
|
}, |
|
{ |
|
"epoch": 45.53, |
|
"eval_loss": 0.14050035178661346, |
|
"eval_runtime": 112.9556, |
|
"eval_samples_per_second": 112.912, |
|
"eval_steps_per_second": 3.532, |
|
"eval_wer": 0.1795298654500627, |
|
"step": 82000 |
|
}, |
|
{ |
|
"epoch": 46.64, |
|
"learning_rate": 8.450611111111112e-05, |
|
"loss": 0.5949, |
|
"step": 84000 |
|
}, |
|
{ |
|
"epoch": 46.64, |
|
"eval_loss": 0.14021989703178406, |
|
"eval_runtime": 109.6871, |
|
"eval_samples_per_second": 116.276, |
|
"eval_steps_per_second": 3.638, |
|
"eval_wer": 0.1760317778082163, |
|
"step": 84000 |
|
}, |
|
{ |
|
"epoch": 47.75, |
|
"learning_rate": 8.413592592592593e-05, |
|
"loss": 0.5928, |
|
"step": 86000 |
|
}, |
|
{ |
|
"epoch": 47.75, |
|
"eval_loss": 0.14247484505176544, |
|
"eval_runtime": 109.7686, |
|
"eval_samples_per_second": 116.19, |
|
"eval_steps_per_second": 3.635, |
|
"eval_wer": 0.1776859939521015, |
|
"step": 86000 |
|
}, |
|
{ |
|
"epoch": 48.86, |
|
"learning_rate": 8.376574074074075e-05, |
|
"loss": 0.5886, |
|
"step": 88000 |
|
}, |
|
{ |
|
"epoch": 48.86, |
|
"eval_loss": 0.1418529599905014, |
|
"eval_runtime": 110.4954, |
|
"eval_samples_per_second": 115.426, |
|
"eval_steps_per_second": 3.611, |
|
"eval_wer": 0.17910840910767156, |
|
"step": 88000 |
|
}, |
|
{ |
|
"epoch": 49.97, |
|
"learning_rate": 8.339555555555556e-05, |
|
"loss": 0.5848, |
|
"step": 90000 |
|
}, |
|
{ |
|
"epoch": 49.97, |
|
"eval_loss": 0.14168688654899597, |
|
"eval_runtime": 111.002, |
|
"eval_samples_per_second": 114.899, |
|
"eval_steps_per_second": 3.595, |
|
"eval_wer": 0.1777281395863406, |
|
"step": 90000 |
|
}, |
|
{ |
|
"epoch": 51.08, |
|
"learning_rate": 8.302537037037037e-05, |
|
"loss": 0.5824, |
|
"step": 92000 |
|
}, |
|
{ |
|
"epoch": 51.08, |
|
"eval_loss": 0.1420871764421463, |
|
"eval_runtime": 113.8377, |
|
"eval_samples_per_second": 112.037, |
|
"eval_steps_per_second": 3.505, |
|
"eval_wer": 0.17755955704938414, |
|
"step": 92000 |
|
}, |
|
{ |
|
"epoch": 52.19, |
|
"learning_rate": 8.2655e-05, |
|
"loss": 0.5814, |
|
"step": 94000 |
|
}, |
|
{ |
|
"epoch": 52.19, |
|
"eval_loss": 0.14134614169597626, |
|
"eval_runtime": 108.8305, |
|
"eval_samples_per_second": 117.191, |
|
"eval_steps_per_second": 3.666, |
|
"eval_wer": 0.17652698901052588, |
|
"step": 94000 |
|
}, |
|
{ |
|
"epoch": 53.3, |
|
"learning_rate": 8.228481481481481e-05, |
|
"loss": 0.5774, |
|
"step": 96000 |
|
}, |
|
{ |
|
"epoch": 53.3, |
|
"eval_loss": 0.1393408626317978, |
|
"eval_runtime": 108.8087, |
|
"eval_samples_per_second": 117.215, |
|
"eval_steps_per_second": 3.667, |
|
"eval_wer": 0.1742721975787333, |
|
"step": 96000 |
|
}, |
|
{ |
|
"epoch": 54.41, |
|
"learning_rate": 8.191462962962963e-05, |
|
"loss": 0.573, |
|
"step": 98000 |
|
}, |
|
{ |
|
"epoch": 54.41, |
|
"eval_loss": 0.1408652812242508, |
|
"eval_runtime": 109.0156, |
|
"eval_samples_per_second": 116.992, |
|
"eval_steps_per_second": 3.66, |
|
"eval_wer": 0.1753574476603905, |
|
"step": 98000 |
|
}, |
|
{ |
|
"epoch": 55.52, |
|
"learning_rate": 8.154444444444444e-05, |
|
"loss": 0.573, |
|
"step": 100000 |
|
}, |
|
{ |
|
"epoch": 55.52, |
|
"eval_loss": 0.14187079668045044, |
|
"eval_runtime": 110.6508, |
|
"eval_samples_per_second": 115.264, |
|
"eval_steps_per_second": 3.606, |
|
"eval_wer": 0.17601070499109672, |
|
"step": 100000 |
|
}, |
|
{ |
|
"epoch": 56.64, |
|
"learning_rate": 8.117407407407408e-05, |
|
"loss": 0.5687, |
|
"step": 102000 |
|
}, |
|
{ |
|
"epoch": 56.64, |
|
"eval_loss": 0.1391436606645584, |
|
"eval_runtime": 109.8014, |
|
"eval_samples_per_second": 116.155, |
|
"eval_steps_per_second": 3.634, |
|
"eval_wer": 0.1731658746799566, |
|
"step": 102000 |
|
}, |
|
{ |
|
"epoch": 57.75, |
|
"learning_rate": 8.080388888888889e-05, |
|
"loss": 0.5648, |
|
"step": 104000 |
|
}, |
|
{ |
|
"epoch": 57.75, |
|
"eval_loss": 0.13951122760772705, |
|
"eval_runtime": 109.5333, |
|
"eval_samples_per_second": 116.44, |
|
"eval_steps_per_second": 3.643, |
|
"eval_wer": 0.1744197072985702, |
|
"step": 104000 |
|
}, |
|
{ |
|
"epoch": 58.86, |
|
"learning_rate": 8.043370370370371e-05, |
|
"loss": 0.5623, |
|
"step": 106000 |
|
}, |
|
{ |
|
"epoch": 58.86, |
|
"eval_loss": 0.141478031873703, |
|
"eval_runtime": 126.0697, |
|
"eval_samples_per_second": 101.166, |
|
"eval_steps_per_second": 3.165, |
|
"eval_wer": 0.17249154453213078, |
|
"step": 106000 |
|
}, |
|
{ |
|
"epoch": 59.97, |
|
"learning_rate": 8.006333333333333e-05, |
|
"loss": 0.5613, |
|
"step": 108000 |
|
}, |
|
{ |
|
"epoch": 59.97, |
|
"eval_loss": 0.140051931142807, |
|
"eval_runtime": 110.0609, |
|
"eval_samples_per_second": 115.881, |
|
"eval_steps_per_second": 3.625, |
|
"eval_wer": 0.17356625820522817, |
|
"step": 108000 |
|
}, |
|
{ |
|
"epoch": 61.08, |
|
"learning_rate": 7.969333333333335e-05, |
|
"loss": 0.5567, |
|
"step": 110000 |
|
}, |
|
{ |
|
"epoch": 61.08, |
|
"eval_loss": 0.1388338953256607, |
|
"eval_runtime": 111.9727, |
|
"eval_samples_per_second": 113.903, |
|
"eval_steps_per_second": 3.563, |
|
"eval_wer": 0.17155380417031052, |
|
"step": 110000 |
|
}, |
|
{ |
|
"epoch": 62.19, |
|
"learning_rate": 7.932296296296296e-05, |
|
"loss": 0.5558, |
|
"step": 112000 |
|
}, |
|
{ |
|
"epoch": 62.19, |
|
"eval_loss": 0.1424715220928192, |
|
"eval_runtime": 117.7299, |
|
"eval_samples_per_second": 108.333, |
|
"eval_steps_per_second": 3.389, |
|
"eval_wer": 0.1722386707266961, |
|
"step": 112000 |
|
}, |
|
{ |
|
"epoch": 63.3, |
|
"learning_rate": 7.895277777777778e-05, |
|
"loss": 0.5515, |
|
"step": 114000 |
|
}, |
|
{ |
|
"epoch": 63.3, |
|
"eval_loss": 0.14163948595523834, |
|
"eval_runtime": 110.1956, |
|
"eval_samples_per_second": 115.74, |
|
"eval_steps_per_second": 3.621, |
|
"eval_wer": 0.1728181731974839, |
|
"step": 114000 |
|
}, |
|
{ |
|
"epoch": 64.41, |
|
"learning_rate": 7.858240740740741e-05, |
|
"loss": 0.5517, |
|
"step": 116000 |
|
}, |
|
{ |
|
"epoch": 64.41, |
|
"eval_loss": 0.1408870965242386, |
|
"eval_runtime": 110.3371, |
|
"eval_samples_per_second": 115.591, |
|
"eval_steps_per_second": 3.616, |
|
"eval_wer": 0.17021568028321865, |
|
"step": 116000 |
|
}, |
|
{ |
|
"epoch": 65.52, |
|
"learning_rate": 7.821222222222223e-05, |
|
"loss": 0.5462, |
|
"step": 118000 |
|
}, |
|
{ |
|
"epoch": 65.52, |
|
"eval_loss": 0.14396576583385468, |
|
"eval_runtime": 110.3476, |
|
"eval_samples_per_second": 115.58, |
|
"eval_steps_per_second": 3.616, |
|
"eval_wer": 0.1711112750107998, |
|
"step": 118000 |
|
}, |
|
{ |
|
"epoch": 66.63, |
|
"learning_rate": 7.784203703703704e-05, |
|
"loss": 0.5468, |
|
"step": 120000 |
|
}, |
|
{ |
|
"epoch": 66.63, |
|
"eval_loss": 0.14202018082141876, |
|
"eval_runtime": 110.3323, |
|
"eval_samples_per_second": 115.596, |
|
"eval_steps_per_second": 3.616, |
|
"eval_wer": 0.16947813168403417, |
|
"step": 120000 |
|
}, |
|
{ |
|
"epoch": 67.74, |
|
"learning_rate": 7.747166666666667e-05, |
|
"loss": 0.5436, |
|
"step": 122000 |
|
}, |
|
{ |
|
"epoch": 67.74, |
|
"eval_loss": 0.1415812373161316, |
|
"eval_runtime": 111.4691, |
|
"eval_samples_per_second": 114.417, |
|
"eval_steps_per_second": 3.579, |
|
"eval_wer": 0.16994173366066442, |
|
"step": 122000 |
|
}, |
|
{ |
|
"epoch": 68.85, |
|
"learning_rate": 7.710166666666667e-05, |
|
"loss": 0.5391, |
|
"step": 124000 |
|
}, |
|
{ |
|
"epoch": 68.85, |
|
"eval_loss": 0.1431398242712021, |
|
"eval_runtime": 110.7164, |
|
"eval_samples_per_second": 115.195, |
|
"eval_steps_per_second": 3.604, |
|
"eval_wer": 0.1731131926371577, |
|
"step": 124000 |
|
}, |
|
{ |
|
"epoch": 69.96, |
|
"learning_rate": 7.673129629629629e-05, |
|
"loss": 0.5386, |
|
"step": 126000 |
|
}, |
|
{ |
|
"epoch": 69.96, |
|
"eval_loss": 0.13997778296470642, |
|
"eval_runtime": 115.7532, |
|
"eval_samples_per_second": 110.183, |
|
"eval_steps_per_second": 3.447, |
|
"eval_wer": 0.17011031619762088, |
|
"step": 126000 |
|
}, |
|
{ |
|
"epoch": 71.07, |
|
"learning_rate": 7.636111111111111e-05, |
|
"loss": 0.5355, |
|
"step": 128000 |
|
}, |
|
{ |
|
"epoch": 71.07, |
|
"eval_loss": 0.14404296875, |
|
"eval_runtime": 108.9746, |
|
"eval_samples_per_second": 117.036, |
|
"eval_steps_per_second": 3.661, |
|
"eval_wer": 0.17013138901474043, |
|
"step": 128000 |
|
}, |
|
{ |
|
"epoch": 72.18, |
|
"learning_rate": 7.599074074074074e-05, |
|
"loss": 0.5326, |
|
"step": 130000 |
|
}, |
|
{ |
|
"epoch": 72.18, |
|
"eval_loss": 0.14536485075950623, |
|
"eval_runtime": 110.0456, |
|
"eval_samples_per_second": 115.897, |
|
"eval_steps_per_second": 3.626, |
|
"eval_wer": 0.16906721175020284, |
|
"step": 130000 |
|
}, |
|
{ |
|
"epoch": 73.29, |
|
"learning_rate": 7.562074074074074e-05, |
|
"loss": 0.5312, |
|
"step": 132000 |
|
}, |
|
{ |
|
"epoch": 73.29, |
|
"eval_loss": 0.14173446595668793, |
|
"eval_runtime": 111.1917, |
|
"eval_samples_per_second": 114.703, |
|
"eval_steps_per_second": 3.588, |
|
"eval_wer": 0.1693306219641973, |
|
"step": 132000 |
|
}, |
|
{ |
|
"epoch": 74.4, |
|
"learning_rate": 7.525037037037037e-05, |
|
"loss": 0.5274, |
|
"step": 134000 |
|
}, |
|
{ |
|
"epoch": 74.4, |
|
"eval_loss": 0.1436368077993393, |
|
"eval_runtime": 109.3262, |
|
"eval_samples_per_second": 116.66, |
|
"eval_steps_per_second": 3.65, |
|
"eval_wer": 0.17104805655944114, |
|
"step": 134000 |
|
}, |
|
{ |
|
"epoch": 75.51, |
|
"learning_rate": 7.488e-05, |
|
"loss": 0.5227, |
|
"step": 136000 |
|
}, |
|
{ |
|
"epoch": 75.51, |
|
"eval_loss": 0.1463368982076645, |
|
"eval_runtime": 110.3294, |
|
"eval_samples_per_second": 115.599, |
|
"eval_steps_per_second": 3.616, |
|
"eval_wer": 0.17241778967221233, |
|
"step": 136000 |
|
}, |
|
{ |
|
"epoch": 76.62, |
|
"learning_rate": 7.450962962962963e-05, |
|
"loss": 0.5245, |
|
"step": 138000 |
|
}, |
|
{ |
|
"epoch": 76.62, |
|
"eval_loss": 0.14312437176704407, |
|
"eval_runtime": 109.8942, |
|
"eval_samples_per_second": 116.057, |
|
"eval_steps_per_second": 3.631, |
|
"eval_wer": 0.16633828193322026, |
|
"step": 138000 |
|
}, |
|
{ |
|
"epoch": 77.73, |
|
"learning_rate": 7.413962962962963e-05, |
|
"loss": 0.5216, |
|
"step": 140000 |
|
}, |
|
{ |
|
"epoch": 77.73, |
|
"eval_loss": 0.14303582906723022, |
|
"eval_runtime": 113.419, |
|
"eval_samples_per_second": 112.45, |
|
"eval_steps_per_second": 3.518, |
|
"eval_wer": 0.1681505442055021, |
|
"step": 140000 |
|
}, |
|
{ |
|
"epoch": 78.84, |
|
"learning_rate": 7.376925925925926e-05, |
|
"loss": 0.5163, |
|
"step": 142000 |
|
}, |
|
{ |
|
"epoch": 78.84, |
|
"eval_loss": 0.14348630607128143, |
|
"eval_runtime": 111.4429, |
|
"eval_samples_per_second": 114.444, |
|
"eval_steps_per_second": 3.58, |
|
"eval_wer": 0.1669915392639265, |
|
"step": 142000 |
|
}, |
|
{ |
|
"epoch": 79.96, |
|
"learning_rate": 7.339925925925926e-05, |
|
"loss": 0.5159, |
|
"step": 144000 |
|
}, |
|
{ |
|
"epoch": 79.96, |
|
"eval_loss": 0.14502613246440887, |
|
"eval_runtime": 110.5689, |
|
"eval_samples_per_second": 115.349, |
|
"eval_steps_per_second": 3.609, |
|
"eval_wer": 0.16607487171922578, |
|
"step": 144000 |
|
}, |
|
{ |
|
"epoch": 81.07, |
|
"learning_rate": 7.302888888888889e-05, |
|
"loss": 0.5152, |
|
"step": 146000 |
|
}, |
|
{ |
|
"epoch": 81.07, |
|
"eval_loss": 0.14235709607601166, |
|
"eval_runtime": 107.4013, |
|
"eval_samples_per_second": 118.751, |
|
"eval_steps_per_second": 3.715, |
|
"eval_wer": 0.16677027468417116, |
|
"step": 146000 |
|
}, |
|
{ |
|
"epoch": 82.18, |
|
"learning_rate": 7.265870370370371e-05, |
|
"loss": 0.5114, |
|
"step": 148000 |
|
}, |
|
{ |
|
"epoch": 82.18, |
|
"eval_loss": 0.14669308066368103, |
|
"eval_runtime": 108.9248, |
|
"eval_samples_per_second": 117.09, |
|
"eval_steps_per_second": 3.663, |
|
"eval_wer": 0.1700049521120231, |
|
"step": 148000 |
|
}, |
|
{ |
|
"epoch": 83.29, |
|
"learning_rate": 7.228833333333333e-05, |
|
"loss": 0.5102, |
|
"step": 150000 |
|
}, |
|
{ |
|
"epoch": 83.29, |
|
"eval_loss": 0.14800167083740234, |
|
"eval_runtime": 110.9546, |
|
"eval_samples_per_second": 114.948, |
|
"eval_steps_per_second": 3.596, |
|
"eval_wer": 0.16751835969191542, |
|
"step": 150000 |
|
}, |
|
{ |
|
"epoch": 84.4, |
|
"learning_rate": 7.191833333333333e-05, |
|
"loss": 0.5074, |
|
"step": 152000 |
|
}, |
|
{ |
|
"epoch": 84.4, |
|
"eval_loss": 0.14698636531829834, |
|
"eval_runtime": 109.1016, |
|
"eval_samples_per_second": 116.9, |
|
"eval_steps_per_second": 3.657, |
|
"eval_wer": 0.16448387402669926, |
|
"step": 152000 |
|
}, |
|
{ |
|
"epoch": 85.51, |
|
"learning_rate": 7.154796296296296e-05, |
|
"loss": 0.5048, |
|
"step": 154000 |
|
}, |
|
{ |
|
"epoch": 85.51, |
|
"eval_loss": 0.14937056601047516, |
|
"eval_runtime": 120.2598, |
|
"eval_samples_per_second": 106.054, |
|
"eval_steps_per_second": 3.318, |
|
"eval_wer": 0.16732870433783942, |
|
"step": 154000 |
|
}, |
|
{ |
|
"epoch": 86.62, |
|
"learning_rate": 7.117796296296296e-05, |
|
"loss": 0.5052, |
|
"step": 156000 |
|
}, |
|
{ |
|
"epoch": 86.62, |
|
"eval_loss": 0.14464369416236877, |
|
"eval_runtime": 109.6362, |
|
"eval_samples_per_second": 116.33, |
|
"eval_steps_per_second": 3.639, |
|
"eval_wer": 0.16599058045074755, |
|
"step": 156000 |
|
}, |
|
{ |
|
"epoch": 87.73, |
|
"learning_rate": 7.080759259259259e-05, |
|
"loss": 0.5017, |
|
"step": 158000 |
|
}, |
|
{ |
|
"epoch": 87.73, |
|
"eval_loss": 0.14508092403411865, |
|
"eval_runtime": 111.0679, |
|
"eval_samples_per_second": 114.831, |
|
"eval_steps_per_second": 3.592, |
|
"eval_wer": 0.1682559082910999, |
|
"step": 158000 |
|
}, |
|
{ |
|
"epoch": 88.84, |
|
"learning_rate": 7.04374074074074e-05, |
|
"loss": 0.5009, |
|
"step": 160000 |
|
}, |
|
{ |
|
"epoch": 88.84, |
|
"eval_loss": 0.15082955360412598, |
|
"eval_runtime": 117.4553, |
|
"eval_samples_per_second": 108.586, |
|
"eval_steps_per_second": 3.397, |
|
"eval_wer": 0.16449441043525903, |
|
"step": 160000 |
|
}, |
|
{ |
|
"epoch": 89.95, |
|
"learning_rate": 7.006703703703704e-05, |
|
"loss": 0.4991, |
|
"step": 162000 |
|
}, |
|
{ |
|
"epoch": 89.95, |
|
"eval_loss": 0.14789538085460663, |
|
"eval_runtime": 110.7552, |
|
"eval_samples_per_second": 115.155, |
|
"eval_steps_per_second": 3.603, |
|
"eval_wer": 0.16420992740414503, |
|
"step": 162000 |
|
}, |
|
{ |
|
"epoch": 91.06, |
|
"learning_rate": 6.969666666666666e-05, |
|
"loss": 0.4955, |
|
"step": 164000 |
|
}, |
|
{ |
|
"epoch": 91.06, |
|
"eval_loss": 0.15189045667648315, |
|
"eval_runtime": 110.2282, |
|
"eval_samples_per_second": 115.705, |
|
"eval_steps_per_second": 3.62, |
|
"eval_wer": 0.1638832987387919, |
|
"step": 164000 |
|
}, |
|
{ |
|
"epoch": 92.17, |
|
"learning_rate": 6.932666666666667e-05, |
|
"loss": 0.4954, |
|
"step": 166000 |
|
}, |
|
{ |
|
"epoch": 92.17, |
|
"eval_loss": 0.1472417414188385, |
|
"eval_runtime": 112.6524, |
|
"eval_samples_per_second": 113.216, |
|
"eval_steps_per_second": 3.542, |
|
"eval_wer": 0.16437850994110148, |
|
"step": 166000 |
|
}, |
|
{ |
|
"epoch": 93.28, |
|
"learning_rate": 6.895629629629629e-05, |
|
"loss": 0.4937, |
|
"step": 168000 |
|
}, |
|
{ |
|
"epoch": 93.28, |
|
"eval_loss": 0.14937187731266022, |
|
"eval_runtime": 111.7122, |
|
"eval_samples_per_second": 114.168, |
|
"eval_steps_per_second": 3.572, |
|
"eval_wer": 0.16531625030292174, |
|
"step": 168000 |
|
}, |
|
{ |
|
"epoch": 94.39, |
|
"learning_rate": 6.858611111111111e-05, |
|
"loss": 0.4913, |
|
"step": 170000 |
|
}, |
|
{ |
|
"epoch": 94.39, |
|
"eval_loss": 0.14994442462921143, |
|
"eval_runtime": 111.1265, |
|
"eval_samples_per_second": 114.77, |
|
"eval_steps_per_second": 3.591, |
|
"eval_wer": 0.16276643943145538, |
|
"step": 170000 |
|
}, |
|
{ |
|
"epoch": 95.5, |
|
"learning_rate": 6.821574074074074e-05, |
|
"loss": 0.4877, |
|
"step": 172000 |
|
}, |
|
{ |
|
"epoch": 95.5, |
|
"eval_loss": 0.15009111166000366, |
|
"eval_runtime": 110.8815, |
|
"eval_samples_per_second": 115.024, |
|
"eval_steps_per_second": 3.598, |
|
"eval_wer": 0.16477889346637306, |
|
"step": 172000 |
|
}, |
|
{ |
|
"epoch": 96.61, |
|
"learning_rate": 6.784555555555556e-05, |
|
"loss": 0.4855, |
|
"step": 174000 |
|
}, |
|
{ |
|
"epoch": 96.61, |
|
"eval_loss": 0.15036821365356445, |
|
"eval_runtime": 123.2578, |
|
"eval_samples_per_second": 103.474, |
|
"eval_steps_per_second": 3.237, |
|
"eval_wer": 0.16344076957928122, |
|
"step": 174000 |
|
}, |
|
{ |
|
"epoch": 97.72, |
|
"learning_rate": 6.747537037037037e-05, |
|
"loss": 0.4865, |
|
"step": 176000 |
|
}, |
|
{ |
|
"epoch": 97.72, |
|
"eval_loss": 0.15151378512382507, |
|
"eval_runtime": 110.1192, |
|
"eval_samples_per_second": 115.82, |
|
"eval_steps_per_second": 3.623, |
|
"eval_wer": 0.16542161438851952, |
|
"step": 176000 |
|
}, |
|
{ |
|
"epoch": 98.83, |
|
"learning_rate": 6.7105e-05, |
|
"loss": 0.4804, |
|
"step": 178000 |
|
}, |
|
{ |
|
"epoch": 98.83, |
|
"eval_loss": 0.1545591503381729, |
|
"eval_runtime": 109.0975, |
|
"eval_samples_per_second": 116.905, |
|
"eval_steps_per_second": 3.657, |
|
"eval_wer": 0.16173387139259712, |
|
"step": 178000 |
|
}, |
|
{ |
|
"epoch": 99.94, |
|
"learning_rate": 6.6735e-05, |
|
"loss": 0.4804, |
|
"step": 180000 |
|
}, |
|
{ |
|
"epoch": 99.94, |
|
"eval_loss": 0.14692111313343048, |
|
"eval_runtime": 110.3968, |
|
"eval_samples_per_second": 115.529, |
|
"eval_steps_per_second": 3.614, |
|
"eval_wer": 0.16394651719015058, |
|
"step": 180000 |
|
}, |
|
{ |
|
"epoch": 101.05, |
|
"learning_rate": 6.636481481481481e-05, |
|
"loss": 0.4784, |
|
"step": 182000 |
|
}, |
|
{ |
|
"epoch": 101.05, |
|
"eval_loss": 0.15785107016563416, |
|
"eval_runtime": 119.3969, |
|
"eval_samples_per_second": 106.82, |
|
"eval_steps_per_second": 3.342, |
|
"eval_wer": 0.16587467995659, |
|
"step": 182000 |
|
}, |
|
{ |
|
"epoch": 102.17, |
|
"learning_rate": 6.599462962962963e-05, |
|
"loss": 0.4754, |
|
"step": 184000 |
|
}, |
|
{ |
|
"epoch": 102.17, |
|
"eval_loss": 0.1541442722082138, |
|
"eval_runtime": 110.3894, |
|
"eval_samples_per_second": 115.536, |
|
"eval_steps_per_second": 3.614, |
|
"eval_wer": 0.16502123086324796, |
|
"step": 184000 |
|
}, |
|
{ |
|
"epoch": 103.28, |
|
"learning_rate": 6.562425925925926e-05, |
|
"loss": 0.4736, |
|
"step": 186000 |
|
}, |
|
{ |
|
"epoch": 103.28, |
|
"eval_loss": 0.15858229994773865, |
|
"eval_runtime": 111.9719, |
|
"eval_samples_per_second": 113.904, |
|
"eval_steps_per_second": 3.563, |
|
"eval_wer": 0.16279804865713474, |
|
"step": 186000 |
|
}, |
|
{ |
|
"epoch": 104.39, |
|
"learning_rate": 6.525407407407407e-05, |
|
"loss": 0.4733, |
|
"step": 188000 |
|
}, |
|
{ |
|
"epoch": 104.39, |
|
"eval_loss": 0.15577121078968048, |
|
"eval_runtime": 111.8331, |
|
"eval_samples_per_second": 114.045, |
|
"eval_steps_per_second": 3.568, |
|
"eval_wer": 0.16437850994110148, |
|
"step": 188000 |
|
}, |
|
{ |
|
"epoch": 105.5, |
|
"learning_rate": 6.48837037037037e-05, |
|
"loss": 0.4729, |
|
"step": 190000 |
|
}, |
|
{ |
|
"epoch": 105.5, |
|
"eval_loss": 0.15798160433769226, |
|
"eval_runtime": 110.7044, |
|
"eval_samples_per_second": 115.208, |
|
"eval_steps_per_second": 3.604, |
|
"eval_wer": 0.16497908522900884, |
|
"step": 190000 |
|
}, |
|
{ |
|
"epoch": 106.61, |
|
"learning_rate": 6.451351851851852e-05, |
|
"loss": 0.4708, |
|
"step": 192000 |
|
}, |
|
{ |
|
"epoch": 106.61, |
|
"eval_loss": 0.1557486653327942, |
|
"eval_runtime": 112.6241, |
|
"eval_samples_per_second": 113.244, |
|
"eval_steps_per_second": 3.543, |
|
"eval_wer": 0.16445226480101993, |
|
"step": 192000 |
|
}, |
|
{ |
|
"epoch": 107.72, |
|
"learning_rate": 6.414314814814814e-05, |
|
"loss": 0.4674, |
|
"step": 194000 |
|
}, |
|
{ |
|
"epoch": 107.72, |
|
"eval_loss": 0.15497049689292908, |
|
"eval_runtime": 111.8369, |
|
"eval_samples_per_second": 114.041, |
|
"eval_steps_per_second": 3.568, |
|
"eval_wer": 0.1644101191667808, |
|
"step": 194000 |
|
}, |
|
{ |
|
"epoch": 108.83, |
|
"learning_rate": 6.377314814814816e-05, |
|
"loss": 0.4645, |
|
"step": 196000 |
|
}, |
|
{ |
|
"epoch": 108.83, |
|
"eval_loss": 0.15774163603782654, |
|
"eval_runtime": 115.5016, |
|
"eval_samples_per_second": 110.423, |
|
"eval_steps_per_second": 3.454, |
|
"eval_wer": 0.16043789313974438, |
|
"step": 196000 |
|
}, |
|
{ |
|
"epoch": 109.94, |
|
"learning_rate": 6.340277777777777e-05, |
|
"loss": 0.4652, |
|
"step": 198000 |
|
}, |
|
{ |
|
"epoch": 109.94, |
|
"eval_loss": 0.1597561538219452, |
|
"eval_runtime": 112.242, |
|
"eval_samples_per_second": 113.63, |
|
"eval_steps_per_second": 3.555, |
|
"eval_wer": 0.16909882097588216, |
|
"step": 198000 |
|
}, |
|
{ |
|
"epoch": 111.05, |
|
"learning_rate": 6.303240740740741e-05, |
|
"loss": 0.4636, |
|
"step": 200000 |
|
}, |
|
{ |
|
"epoch": 111.05, |
|
"eval_loss": 0.15712451934814453, |
|
"eval_runtime": 116.0757, |
|
"eval_samples_per_second": 109.877, |
|
"eval_steps_per_second": 3.437, |
|
"eval_wer": 0.16316682295672696, |
|
"step": 200000 |
|
}, |
|
{ |
|
"epoch": 112.16, |
|
"learning_rate": 6.266222222222222e-05, |
|
"loss": 0.4615, |
|
"step": 202000 |
|
}, |
|
{ |
|
"epoch": 112.16, |
|
"eval_loss": 0.16025520861148834, |
|
"eval_runtime": 109.4143, |
|
"eval_samples_per_second": 116.566, |
|
"eval_steps_per_second": 3.647, |
|
"eval_wer": 0.15968980813200012, |
|
"step": 202000 |
|
}, |
|
{ |
|
"epoch": 113.27, |
|
"learning_rate": 6.229203703703704e-05, |
|
"loss": 0.4586, |
|
"step": 204000 |
|
}, |
|
{ |
|
"epoch": 113.27, |
|
"eval_loss": 0.16139160096645355, |
|
"eval_runtime": 110.5104, |
|
"eval_samples_per_second": 115.41, |
|
"eval_steps_per_second": 3.611, |
|
"eval_wer": 0.1633986239450421, |
|
"step": 204000 |
|
}, |
|
{ |
|
"epoch": 114.38, |
|
"learning_rate": 6.192185185185185e-05, |
|
"loss": 0.4546, |
|
"step": 206000 |
|
}, |
|
{ |
|
"epoch": 114.38, |
|
"eval_loss": 0.15872280299663544, |
|
"eval_runtime": 111.2156, |
|
"eval_samples_per_second": 114.678, |
|
"eval_steps_per_second": 3.588, |
|
"eval_wer": 0.1650001580461284, |
|
"step": 206000 |
|
}, |
|
{ |
|
"epoch": 115.49, |
|
"learning_rate": 6.155166666666668e-05, |
|
"loss": 0.4554, |
|
"step": 208000 |
|
}, |
|
{ |
|
"epoch": 115.49, |
|
"eval_loss": 0.1584981381893158, |
|
"eval_runtime": 109.0033, |
|
"eval_samples_per_second": 117.006, |
|
"eval_steps_per_second": 3.66, |
|
"eval_wer": 0.16186030829531445, |
|
"step": 208000 |
|
}, |
|
{ |
|
"epoch": 116.6, |
|
"learning_rate": 6.118148148148148e-05, |
|
"loss": 0.4534, |
|
"step": 210000 |
|
}, |
|
{ |
|
"epoch": 116.6, |
|
"eval_loss": 0.16507840156555176, |
|
"eval_runtime": 111.2078, |
|
"eval_samples_per_second": 114.686, |
|
"eval_steps_per_second": 3.588, |
|
"eval_wer": 0.16058540285958128, |
|
"step": 210000 |
|
}, |
|
{ |
|
"epoch": 117.71, |
|
"learning_rate": 6.081111111111112e-05, |
|
"loss": 0.452, |
|
"step": 212000 |
|
}, |
|
{ |
|
"epoch": 117.71, |
|
"eval_loss": 0.16277721524238586, |
|
"eval_runtime": 110.743, |
|
"eval_samples_per_second": 115.168, |
|
"eval_steps_per_second": 3.603, |
|
"eval_wer": 0.16030091982846728, |
|
"step": 212000 |
|
}, |
|
{ |
|
"epoch": 118.82, |
|
"learning_rate": 6.044092592592593e-05, |
|
"loss": 0.4498, |
|
"step": 214000 |
|
}, |
|
{ |
|
"epoch": 118.82, |
|
"eval_loss": 0.1640331745147705, |
|
"eval_runtime": 111.5522, |
|
"eval_samples_per_second": 114.332, |
|
"eval_steps_per_second": 3.577, |
|
"eval_wer": 0.1609752499762931, |
|
"step": 214000 |
|
}, |
|
{ |
|
"epoch": 119.93, |
|
"learning_rate": 6.0070555555555565e-05, |
|
"loss": 0.4475, |
|
"step": 216000 |
|
}, |
|
{ |
|
"epoch": 119.93, |
|
"eval_loss": 0.16334258019924164, |
|
"eval_runtime": 110.4588, |
|
"eval_samples_per_second": 115.464, |
|
"eval_steps_per_second": 3.612, |
|
"eval_wer": 0.16188138111243403, |
|
"step": 216000 |
|
}, |
|
{ |
|
"epoch": 121.04, |
|
"learning_rate": 5.9700370370370374e-05, |
|
"loss": 0.4456, |
|
"step": 218000 |
|
}, |
|
{ |
|
"epoch": 121.04, |
|
"eval_loss": 0.1660892814397812, |
|
"eval_runtime": 109.1417, |
|
"eval_samples_per_second": 116.857, |
|
"eval_steps_per_second": 3.656, |
|
"eval_wer": 0.15847812114762563, |
|
"step": 218000 |
|
}, |
|
{ |
|
"epoch": 122.15, |
|
"learning_rate": 5.933018518518518e-05, |
|
"loss": 0.4453, |
|
"step": 220000 |
|
}, |
|
{ |
|
"epoch": 122.15, |
|
"eval_loss": 0.16312651336193085, |
|
"eval_runtime": 109.8128, |
|
"eval_samples_per_second": 116.143, |
|
"eval_steps_per_second": 3.633, |
|
"eval_wer": 0.15943693432656544, |
|
"step": 220000 |
|
}, |
|
{ |
|
"epoch": 123.26, |
|
"learning_rate": 5.8960000000000005e-05, |
|
"loss": 0.443, |
|
"step": 222000 |
|
}, |
|
{ |
|
"epoch": 123.26, |
|
"eval_loss": 0.16088205575942993, |
|
"eval_runtime": 110.2399, |
|
"eval_samples_per_second": 115.693, |
|
"eval_steps_per_second": 3.619, |
|
"eval_wer": 0.15721375212045222, |
|
"step": 222000 |
|
}, |
|
{ |
|
"epoch": 124.38, |
|
"learning_rate": 5.8589814814814813e-05, |
|
"loss": 0.4437, |
|
"step": 224000 |
|
}, |
|
{ |
|
"epoch": 124.38, |
|
"eval_loss": 0.1613088697195053, |
|
"eval_runtime": 111.1211, |
|
"eval_samples_per_second": 114.776, |
|
"eval_steps_per_second": 3.591, |
|
"eval_wer": 0.15813041966515293, |
|
"step": 224000 |
|
}, |
|
{ |
|
"epoch": 125.49, |
|
"learning_rate": 5.821944444444445e-05, |
|
"loss": 0.4391, |
|
"step": 226000 |
|
}, |
|
{ |
|
"epoch": 125.49, |
|
"eval_loss": 0.17190876603126526, |
|
"eval_runtime": 125.7786, |
|
"eval_samples_per_second": 101.4, |
|
"eval_steps_per_second": 3.172, |
|
"eval_wer": 0.1587309949530603, |
|
"step": 226000 |
|
}, |
|
{ |
|
"epoch": 126.6, |
|
"learning_rate": 5.784925925925926e-05, |
|
"loss": 0.4372, |
|
"step": 228000 |
|
}, |
|
{ |
|
"epoch": 126.6, |
|
"eval_loss": 0.16242261230945587, |
|
"eval_runtime": 111.445, |
|
"eval_samples_per_second": 114.442, |
|
"eval_steps_per_second": 3.58, |
|
"eval_wer": 0.15744555310876734, |
|
"step": 228000 |
|
}, |
|
{ |
|
"epoch": 127.71, |
|
"learning_rate": 5.747907407407408e-05, |
|
"loss": 0.4377, |
|
"step": 230000 |
|
}, |
|
{ |
|
"epoch": 127.71, |
|
"eval_loss": 0.16486689448356628, |
|
"eval_runtime": 111.8011, |
|
"eval_samples_per_second": 114.078, |
|
"eval_steps_per_second": 3.569, |
|
"eval_wer": 0.15686605063797954, |
|
"step": 230000 |
|
}, |
|
{ |
|
"epoch": 128.82, |
|
"learning_rate": 5.71087037037037e-05, |
|
"loss": 0.4349, |
|
"step": 232000 |
|
}, |
|
{ |
|
"epoch": 128.82, |
|
"eval_loss": 0.1673368662595749, |
|
"eval_runtime": 112.5569, |
|
"eval_samples_per_second": 113.312, |
|
"eval_steps_per_second": 3.545, |
|
"eval_wer": 0.15947907996080457, |
|
"step": 232000 |
|
}, |
|
{ |
|
"epoch": 129.93, |
|
"learning_rate": 5.6738518518518525e-05, |
|
"loss": 0.4342, |
|
"step": 234000 |
|
}, |
|
{ |
|
"epoch": 129.93, |
|
"eval_loss": 0.16691875457763672, |
|
"eval_runtime": 118.5844, |
|
"eval_samples_per_second": 107.552, |
|
"eval_steps_per_second": 3.365, |
|
"eval_wer": 0.15731911620605002, |
|
"step": 234000 |
|
}, |
|
{ |
|
"epoch": 131.04, |
|
"learning_rate": 5.6368333333333334e-05, |
|
"loss": 0.4335, |
|
"step": 236000 |
|
}, |
|
{ |
|
"epoch": 131.04, |
|
"eval_loss": 0.17210978269577026, |
|
"eval_runtime": 110.1221, |
|
"eval_samples_per_second": 115.817, |
|
"eval_steps_per_second": 3.623, |
|
"eval_wer": 0.15840436628770718, |
|
"step": 236000 |
|
}, |
|
{ |
|
"epoch": 132.15, |
|
"learning_rate": 5.599796296296297e-05, |
|
"loss": 0.4302, |
|
"step": 238000 |
|
}, |
|
{ |
|
"epoch": 132.15, |
|
"eval_loss": 0.17578274011611938, |
|
"eval_runtime": 110.8686, |
|
"eval_samples_per_second": 115.037, |
|
"eval_steps_per_second": 3.599, |
|
"eval_wer": 0.1587309949530603, |
|
"step": 238000 |
|
}, |
|
{ |
|
"epoch": 133.26, |
|
"learning_rate": 5.562777777777778e-05, |
|
"loss": 0.4292, |
|
"step": 240000 |
|
}, |
|
{ |
|
"epoch": 133.26, |
|
"eval_loss": 0.173477903008461, |
|
"eval_runtime": 110.0323, |
|
"eval_samples_per_second": 115.911, |
|
"eval_steps_per_second": 3.626, |
|
"eval_wer": 0.16007965524871193, |
|
"step": 240000 |
|
}, |
|
{ |
|
"epoch": 134.37, |
|
"learning_rate": 5.5257407407407414e-05, |
|
"loss": 0.4269, |
|
"step": 242000 |
|
}, |
|
{ |
|
"epoch": 134.37, |
|
"eval_loss": 0.16973139345645905, |
|
"eval_runtime": 113.6422, |
|
"eval_samples_per_second": 112.229, |
|
"eval_steps_per_second": 3.511, |
|
"eval_wer": 0.1569819511321371, |
|
"step": 242000 |
|
}, |
|
{ |
|
"epoch": 135.48, |
|
"learning_rate": 5.488722222222222e-05, |
|
"loss": 0.4273, |
|
"step": 244000 |
|
}, |
|
{ |
|
"epoch": 135.48, |
|
"eval_loss": 0.17380733788013458, |
|
"eval_runtime": 110.4957, |
|
"eval_samples_per_second": 115.425, |
|
"eval_steps_per_second": 3.611, |
|
"eval_wer": 0.15647620352126773, |
|
"step": 244000 |
|
}, |
|
{ |
|
"epoch": 136.59, |
|
"learning_rate": 5.4517037037037045e-05, |
|
"loss": 0.4257, |
|
"step": 246000 |
|
}, |
|
{ |
|
"epoch": 136.59, |
|
"eval_loss": 0.16813333332538605, |
|
"eval_runtime": 111.9779, |
|
"eval_samples_per_second": 113.897, |
|
"eval_steps_per_second": 3.563, |
|
"eval_wer": 0.15652888556406663, |
|
"step": 246000 |
|
}, |
|
{ |
|
"epoch": 137.7, |
|
"learning_rate": 5.414666666666667e-05, |
|
"loss": 0.4221, |
|
"step": 248000 |
|
}, |
|
{ |
|
"epoch": 137.7, |
|
"eval_loss": 0.17320315539836884, |
|
"eval_runtime": 111.5731, |
|
"eval_samples_per_second": 114.311, |
|
"eval_steps_per_second": 3.576, |
|
"eval_wer": 0.15928942460672854, |
|
"step": 248000 |
|
}, |
|
{ |
|
"epoch": 138.81, |
|
"learning_rate": 5.377648148148149e-05, |
|
"loss": 0.422, |
|
"step": 250000 |
|
}, |
|
{ |
|
"epoch": 138.81, |
|
"eval_loss": 0.17005948722362518, |
|
"eval_runtime": 116.6285, |
|
"eval_samples_per_second": 109.356, |
|
"eval_steps_per_second": 3.421, |
|
"eval_wer": 0.15444267666923053, |
|
"step": 250000 |
|
}, |
|
{ |
|
"epoch": 139.92, |
|
"learning_rate": 5.34062962962963e-05, |
|
"loss": 0.4178, |
|
"step": 252000 |
|
}, |
|
{ |
|
"epoch": 139.92, |
|
"eval_loss": 0.17322979867458344, |
|
"eval_runtime": 117.6422, |
|
"eval_samples_per_second": 108.413, |
|
"eval_steps_per_second": 3.392, |
|
"eval_wer": 0.15422141208947518, |
|
"step": 252000 |
|
}, |
|
{ |
|
"epoch": 141.03, |
|
"learning_rate": 5.303611111111111e-05, |
|
"loss": 0.4182, |
|
"step": 254000 |
|
}, |
|
{ |
|
"epoch": 141.03, |
|
"eval_loss": 0.1738407015800476, |
|
"eval_runtime": 113.7677, |
|
"eval_samples_per_second": 112.106, |
|
"eval_steps_per_second": 3.507, |
|
"eval_wer": 0.15336796299613314, |
|
"step": 254000 |
|
}, |
|
{ |
|
"epoch": 142.14, |
|
"learning_rate": 5.266592592592593e-05, |
|
"loss": 0.4166, |
|
"step": 256000 |
|
}, |
|
{ |
|
"epoch": 142.14, |
|
"eval_loss": 0.1734342724084854, |
|
"eval_runtime": 112.0095, |
|
"eval_samples_per_second": 113.865, |
|
"eval_steps_per_second": 3.562, |
|
"eval_wer": 0.15277792411678556, |
|
"step": 256000 |
|
}, |
|
{ |
|
"epoch": 143.25, |
|
"learning_rate": 5.229574074074074e-05, |
|
"loss": 0.4163, |
|
"step": 258000 |
|
}, |
|
{ |
|
"epoch": 143.25, |
|
"eval_loss": 0.1752772331237793, |
|
"eval_runtime": 111.2622, |
|
"eval_samples_per_second": 114.63, |
|
"eval_steps_per_second": 3.586, |
|
"eval_wer": 0.15545417189096924, |
|
"step": 258000 |
|
}, |
|
{ |
|
"epoch": 144.36, |
|
"learning_rate": 5.1925370370370374e-05, |
|
"loss": 0.4144, |
|
"step": 260000 |
|
}, |
|
{ |
|
"epoch": 144.36, |
|
"eval_loss": 0.17757321894168854, |
|
"eval_runtime": 114.1149, |
|
"eval_samples_per_second": 111.765, |
|
"eval_steps_per_second": 3.496, |
|
"eval_wer": 0.15680283218662086, |
|
"step": 260000 |
|
}, |
|
{ |
|
"epoch": 145.47, |
|
"learning_rate": 5.1555e-05, |
|
"loss": 0.4103, |
|
"step": 262000 |
|
}, |
|
{ |
|
"epoch": 145.47, |
|
"eval_loss": 0.17166772484779358, |
|
"eval_runtime": 111.5777, |
|
"eval_samples_per_second": 114.306, |
|
"eval_steps_per_second": 3.576, |
|
"eval_wer": 0.15237754059151398, |
|
"step": 262000 |
|
}, |
|
{ |
|
"epoch": 146.59, |
|
"learning_rate": 5.1185000000000005e-05, |
|
"loss": 0.4117, |
|
"step": 264000 |
|
}, |
|
{ |
|
"epoch": 146.59, |
|
"eval_loss": 0.1743527501821518, |
|
"eval_runtime": 112.423, |
|
"eval_samples_per_second": 113.447, |
|
"eval_steps_per_second": 3.549, |
|
"eval_wer": 0.15553846315944747, |
|
"step": 264000 |
|
}, |
|
{ |
|
"epoch": 147.7, |
|
"learning_rate": 5.081462962962963e-05, |
|
"loss": 0.4085, |
|
"step": 266000 |
|
}, |
|
{ |
|
"epoch": 147.7, |
|
"eval_loss": 0.17306753993034363, |
|
"eval_runtime": 112.3063, |
|
"eval_samples_per_second": 113.564, |
|
"eval_steps_per_second": 3.553, |
|
"eval_wer": 0.15564382724504525, |
|
"step": 266000 |
|
}, |
|
{ |
|
"epoch": 148.81, |
|
"learning_rate": 5.0444259259259264e-05, |
|
"loss": 0.408, |
|
"step": 268000 |
|
}, |
|
{ |
|
"epoch": 148.81, |
|
"eval_loss": 0.17164543271064758, |
|
"eval_runtime": 113.1307, |
|
"eval_samples_per_second": 112.737, |
|
"eval_steps_per_second": 3.527, |
|
"eval_wer": 0.1530940163735789, |
|
"step": 268000 |
|
}, |
|
{ |
|
"epoch": 149.92, |
|
"learning_rate": 5.0073888888888886e-05, |
|
"loss": 0.4078, |
|
"step": 270000 |
|
}, |
|
{ |
|
"epoch": 149.92, |
|
"eval_loss": 0.17812077701091766, |
|
"eval_runtime": 114.4521, |
|
"eval_samples_per_second": 111.435, |
|
"eval_steps_per_second": 3.486, |
|
"eval_wer": 0.15219842164599776, |
|
"step": 270000 |
|
}, |
|
{ |
|
"epoch": 151.03, |
|
"learning_rate": 4.970425925925926e-05, |
|
"loss": 0.4051, |
|
"step": 272000 |
|
}, |
|
{ |
|
"epoch": 151.03, |
|
"eval_loss": 0.17345324158668518, |
|
"eval_runtime": 109.0896, |
|
"eval_samples_per_second": 116.913, |
|
"eval_steps_per_second": 3.658, |
|
"eval_wer": 0.15237754059151398, |
|
"step": 272000 |
|
}, |
|
{ |
|
"epoch": 152.14, |
|
"learning_rate": 4.933388888888889e-05, |
|
"loss": 0.4033, |
|
"step": 274000 |
|
}, |
|
{ |
|
"epoch": 152.14, |
|
"eval_loss": 0.18191710114479065, |
|
"eval_runtime": 111.0027, |
|
"eval_samples_per_second": 114.898, |
|
"eval_steps_per_second": 3.595, |
|
"eval_wer": 0.15185072016352505, |
|
"step": 274000 |
|
}, |
|
{ |
|
"epoch": 153.25, |
|
"learning_rate": 4.896351851851852e-05, |
|
"loss": 0.4037, |
|
"step": 276000 |
|
}, |
|
{ |
|
"epoch": 153.25, |
|
"eval_loss": 0.18181933462619781, |
|
"eval_runtime": 110.4212, |
|
"eval_samples_per_second": 115.503, |
|
"eval_steps_per_second": 3.613, |
|
"eval_wer": 0.15307294355645934, |
|
"step": 276000 |
|
}, |
|
{ |
|
"epoch": 154.36, |
|
"learning_rate": 4.8593333333333335e-05, |
|
"loss": 0.4002, |
|
"step": 278000 |
|
}, |
|
{ |
|
"epoch": 154.36, |
|
"eval_loss": 0.17859378457069397, |
|
"eval_runtime": 116.538, |
|
"eval_samples_per_second": 109.441, |
|
"eval_steps_per_second": 3.424, |
|
"eval_wer": 0.1548852058287412, |
|
"step": 278000 |
|
}, |
|
{ |
|
"epoch": 155.47, |
|
"learning_rate": 4.8222962962962964e-05, |
|
"loss": 0.4021, |
|
"step": 280000 |
|
}, |
|
{ |
|
"epoch": 155.47, |
|
"eval_loss": 0.17952340841293335, |
|
"eval_runtime": 112.6278, |
|
"eval_samples_per_second": 113.24, |
|
"eval_steps_per_second": 3.543, |
|
"eval_wer": 0.15223003087167708, |
|
"step": 280000 |
|
}, |
|
{ |
|
"epoch": 156.58, |
|
"learning_rate": 4.785277777777778e-05, |
|
"loss": 0.3985, |
|
"step": 282000 |
|
}, |
|
{ |
|
"epoch": 156.58, |
|
"eval_loss": 0.18241500854492188, |
|
"eval_runtime": 110.8984, |
|
"eval_samples_per_second": 115.006, |
|
"eval_steps_per_second": 3.598, |
|
"eval_wer": 0.1529570430623018, |
|
"step": 282000 |
|
}, |
|
{ |
|
"epoch": 157.69, |
|
"learning_rate": 4.7482592592592595e-05, |
|
"loss": 0.3987, |
|
"step": 284000 |
|
}, |
|
{ |
|
"epoch": 157.69, |
|
"eval_loss": 0.18239329755306244, |
|
"eval_runtime": 116.2871, |
|
"eval_samples_per_second": 109.677, |
|
"eval_steps_per_second": 3.431, |
|
"eval_wer": 0.15229324932303576, |
|
"step": 284000 |
|
}, |
|
{ |
|
"epoch": 158.8, |
|
"learning_rate": 4.7112222222222224e-05, |
|
"loss": 0.3957, |
|
"step": 286000 |
|
}, |
|
{ |
|
"epoch": 158.8, |
|
"eval_loss": 0.18385788798332214, |
|
"eval_runtime": 111.0629, |
|
"eval_samples_per_second": 114.836, |
|
"eval_steps_per_second": 3.593, |
|
"eval_wer": 0.15160838276665015, |
|
"step": 286000 |
|
} |
|
], |
|
"max_steps": 540300, |
|
"num_train_epochs": 300, |
|
"total_flos": 2.3864414600426045e+21, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|