| { | |
| "best_metric": 0.012459825724363327, | |
| "best_model_checkpoint": "./phase3-30-ep/checkpoint-473000", | |
| "epoch": 50.0, | |
| "eval_steps": 1000, | |
| "global_step": 487100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.10264832683227264, | |
| "grad_norm": 0.17517703771591187, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1055, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.10264832683227264, | |
| "eval_cer": 0.029768957345971563, | |
| "eval_loss": 0.07166381180286407, | |
| "eval_runtime": 21.4774, | |
| "eval_samples_per_second": 4.19, | |
| "eval_steps_per_second": 0.047, | |
| "eval_wer": 0.11796246648793565, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.20529665366454528, | |
| "grad_norm": 0.18113084137439728, | |
| "learning_rate": 0.0004989714050606871, | |
| "loss": 0.0829, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.20529665366454528, | |
| "eval_cer": 0.02754739336492891, | |
| "eval_loss": 0.06128456071019173, | |
| "eval_runtime": 24.2215, | |
| "eval_samples_per_second": 3.716, | |
| "eval_steps_per_second": 0.041, | |
| "eval_wer": 0.10455764075067024, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.3079449804968179, | |
| "grad_norm": 0.31132909655570984, | |
| "learning_rate": 0.0004979428101213742, | |
| "loss": 0.077, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.3079449804968179, | |
| "eval_cer": 0.026954976303317536, | |
| "eval_loss": 0.056063100695610046, | |
| "eval_runtime": 24.595, | |
| "eval_samples_per_second": 3.659, | |
| "eval_steps_per_second": 0.041, | |
| "eval_wer": 0.10187667560321716, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.41059330732909055, | |
| "grad_norm": 0.2209460735321045, | |
| "learning_rate": 0.0004969142151820613, | |
| "loss": 0.0746, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.41059330732909055, | |
| "eval_cer": 0.023548578199052133, | |
| "eval_loss": 0.05421268939971924, | |
| "eval_runtime": 39.13, | |
| "eval_samples_per_second": 2.3, | |
| "eval_steps_per_second": 0.026, | |
| "eval_wer": 0.09204647006255585, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5132416341613631, | |
| "grad_norm": 0.21206562221050262, | |
| "learning_rate": 0.0004958856202427484, | |
| "loss": 0.0723, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.5132416341613631, | |
| "eval_cer": 0.023548578199052133, | |
| "eval_loss": 0.053859543055295944, | |
| "eval_runtime": 39.8671, | |
| "eval_samples_per_second": 2.258, | |
| "eval_steps_per_second": 0.025, | |
| "eval_wer": 0.0902591599642538, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.6158899609936358, | |
| "grad_norm": 0.1705954223871231, | |
| "learning_rate": 0.0004948570253034355, | |
| "loss": 0.0705, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.6158899609936358, | |
| "eval_cer": 0.0231042654028436, | |
| "eval_loss": 0.05038898065686226, | |
| "eval_runtime": 23.9921, | |
| "eval_samples_per_second": 3.751, | |
| "eval_steps_per_second": 0.042, | |
| "eval_wer": 0.08936550491510277, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.7185382878259085, | |
| "grad_norm": 0.24641267955303192, | |
| "learning_rate": 0.0004938284303641226, | |
| "loss": 0.0693, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.7185382878259085, | |
| "eval_cer": 0.02428909952606635, | |
| "eval_loss": 0.04804808273911476, | |
| "eval_runtime": 25.4073, | |
| "eval_samples_per_second": 3.542, | |
| "eval_steps_per_second": 0.039, | |
| "eval_wer": 0.09204647006255585, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.8211866146581811, | |
| "grad_norm": 0.14618875086307526, | |
| "learning_rate": 0.0004927998354248098, | |
| "loss": 0.0687, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.8211866146581811, | |
| "eval_cer": 0.022067535545023696, | |
| "eval_loss": 0.046750105917453766, | |
| "eval_runtime": 23.0625, | |
| "eval_samples_per_second": 3.902, | |
| "eval_steps_per_second": 0.043, | |
| "eval_wer": 0.08310991957104558, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.9238349414904538, | |
| "grad_norm": 0.2242618203163147, | |
| "learning_rate": 0.0004917712404854969, | |
| "loss": 0.0668, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.9238349414904538, | |
| "eval_cer": 0.02177132701421801, | |
| "eval_loss": 0.046149224042892456, | |
| "eval_runtime": 23.9314, | |
| "eval_samples_per_second": 3.761, | |
| "eval_steps_per_second": 0.042, | |
| "eval_wer": 0.08489722966934764, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.0264832683227263, | |
| "grad_norm": 0.17396153509616852, | |
| "learning_rate": 0.000490742645546184, | |
| "loss": 0.066, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.0264832683227263, | |
| "eval_cer": 0.022363744075829382, | |
| "eval_loss": 0.044813916087150574, | |
| "eval_runtime": 28.3612, | |
| "eval_samples_per_second": 3.173, | |
| "eval_steps_per_second": 0.035, | |
| "eval_wer": 0.08579088471849866, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.129131595154999, | |
| "grad_norm": 0.24261055886745453, | |
| "learning_rate": 0.000489714050606871, | |
| "loss": 0.0632, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.129131595154999, | |
| "eval_cer": 0.02295616113744076, | |
| "eval_loss": 0.04648038372397423, | |
| "eval_runtime": 36.4007, | |
| "eval_samples_per_second": 2.472, | |
| "eval_steps_per_second": 0.027, | |
| "eval_wer": 0.08757819481680071, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.2317799219872716, | |
| "grad_norm": 0.20607537031173706, | |
| "learning_rate": 0.0004886854556675581, | |
| "loss": 0.0639, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.2317799219872716, | |
| "eval_cer": 0.021475118483412322, | |
| "eval_loss": 0.04516833648085594, | |
| "eval_runtime": 40.6247, | |
| "eval_samples_per_second": 2.215, | |
| "eval_steps_per_second": 0.025, | |
| "eval_wer": 0.0840035746201966, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.3344282488195442, | |
| "grad_norm": 0.2226237952709198, | |
| "learning_rate": 0.00048765686072824524, | |
| "loss": 0.0626, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.3344282488195442, | |
| "eval_cer": 0.022363744075829382, | |
| "eval_loss": 0.04331167787313461, | |
| "eval_runtime": 43.1244, | |
| "eval_samples_per_second": 2.087, | |
| "eval_steps_per_second": 0.023, | |
| "eval_wer": 0.08668453976764968, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.437076575651817, | |
| "grad_norm": 0.22998760640621185, | |
| "learning_rate": 0.00048662826578893233, | |
| "loss": 0.0617, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.437076575651817, | |
| "eval_cer": 0.020438388625592416, | |
| "eval_loss": 0.0439009889960289, | |
| "eval_runtime": 40.051, | |
| "eval_samples_per_second": 2.247, | |
| "eval_steps_per_second": 0.025, | |
| "eval_wer": 0.08132260947274352, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.5397249024840896, | |
| "grad_norm": 0.2044006586074829, | |
| "learning_rate": 0.0004855996708496194, | |
| "loss": 0.0612, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.5397249024840896, | |
| "eval_cer": 0.018364928909952605, | |
| "eval_loss": 0.039780329912900925, | |
| "eval_runtime": 40.1221, | |
| "eval_samples_per_second": 2.243, | |
| "eval_steps_per_second": 0.025, | |
| "eval_wer": 0.07149240393208221, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.642373229316362, | |
| "grad_norm": 0.25967568159103394, | |
| "learning_rate": 0.0004845710759103065, | |
| "loss": 0.0619, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.642373229316362, | |
| "eval_cer": 0.021178909952606635, | |
| "eval_loss": 0.04091305658221245, | |
| "eval_runtime": 39.0877, | |
| "eval_samples_per_second": 2.303, | |
| "eval_steps_per_second": 0.026, | |
| "eval_wer": 0.0777479892761394, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.7450215561486346, | |
| "grad_norm": 0.17572972178459167, | |
| "learning_rate": 0.0004835424809709936, | |
| "loss": 0.0617, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.7450215561486346, | |
| "eval_cer": 0.021475118483412322, | |
| "eval_loss": 0.04012183099985123, | |
| "eval_runtime": 39.7698, | |
| "eval_samples_per_second": 2.263, | |
| "eval_steps_per_second": 0.025, | |
| "eval_wer": 0.07864164432529044, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.8476698829809073, | |
| "grad_norm": 0.20715534687042236, | |
| "learning_rate": 0.0004825138860316807, | |
| "loss": 0.0607, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.8476698829809073, | |
| "eval_cer": 0.02177132701421801, | |
| "eval_loss": 0.04150845482945442, | |
| "eval_runtime": 39.4055, | |
| "eval_samples_per_second": 2.284, | |
| "eval_steps_per_second": 0.025, | |
| "eval_wer": 0.08132260947274352, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.95031820981318, | |
| "grad_norm": 0.3426735997200012, | |
| "learning_rate": 0.00048148529109236785, | |
| "loss": 0.0602, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.95031820981318, | |
| "eval_cer": 0.019105450236966824, | |
| "eval_loss": 0.03899341821670532, | |
| "eval_runtime": 39.6568, | |
| "eval_samples_per_second": 2.269, | |
| "eval_steps_per_second": 0.025, | |
| "eval_wer": 0.06881143878462913, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.0529665366454526, | |
| "grad_norm": 0.15929488837718964, | |
| "learning_rate": 0.00048045669615305494, | |
| "loss": 0.0585, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.0529665366454526, | |
| "eval_cer": 0.019994075829383885, | |
| "eval_loss": 0.03957120701670647, | |
| "eval_runtime": 24.5545, | |
| "eval_samples_per_second": 3.665, | |
| "eval_steps_per_second": 0.041, | |
| "eval_wer": 0.07506702412868632, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.1556148634777252, | |
| "grad_norm": 0.18062791228294373, | |
| "learning_rate": 0.00047942810121374204, | |
| "loss": 0.0579, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.1556148634777252, | |
| "eval_cer": 0.019105450236966824, | |
| "eval_loss": 0.039500825107097626, | |
| "eval_runtime": 21.8989, | |
| "eval_samples_per_second": 4.11, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.06970509383378017, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.258263190309998, | |
| "grad_norm": 0.20961548388004303, | |
| "learning_rate": 0.00047839950627442913, | |
| "loss": 0.0571, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 2.258263190309998, | |
| "eval_cer": 0.01925355450236967, | |
| "eval_loss": 0.04062485322356224, | |
| "eval_runtime": 23.0215, | |
| "eval_samples_per_second": 3.909, | |
| "eval_steps_per_second": 0.043, | |
| "eval_wer": 0.0741733690795353, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 2.3609115171422705, | |
| "grad_norm": 0.1982312947511673, | |
| "learning_rate": 0.0004773709113351162, | |
| "loss": 0.0574, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 2.3609115171422705, | |
| "eval_cer": 0.0173281990521327, | |
| "eval_loss": 0.03924456238746643, | |
| "eval_runtime": 26.6274, | |
| "eval_samples_per_second": 3.38, | |
| "eval_steps_per_second": 0.038, | |
| "eval_wer": 0.064343163538874, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 2.463559843974543, | |
| "grad_norm": 0.26111695170402527, | |
| "learning_rate": 0.0004763423163958033, | |
| "loss": 0.0568, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.463559843974543, | |
| "eval_cer": 0.017772511848341232, | |
| "eval_loss": 0.038703735917806625, | |
| "eval_runtime": 23.8875, | |
| "eval_samples_per_second": 3.768, | |
| "eval_steps_per_second": 0.042, | |
| "eval_wer": 0.0679177837354781, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.566208170806816, | |
| "grad_norm": 0.20790116488933563, | |
| "learning_rate": 0.0004753137214564904, | |
| "loss": 0.0571, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.566208170806816, | |
| "eval_cer": 0.017031990521327013, | |
| "eval_loss": 0.03755784407258034, | |
| "eval_runtime": 22.7427, | |
| "eval_samples_per_second": 3.957, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.064343163538874, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.6688564976390885, | |
| "grad_norm": 0.16015666723251343, | |
| "learning_rate": 0.00047428512651717756, | |
| "loss": 0.0572, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.6688564976390885, | |
| "eval_cer": 0.017920616113744077, | |
| "eval_loss": 0.03704160824418068, | |
| "eval_runtime": 22.6235, | |
| "eval_samples_per_second": 3.978, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.06613047363717604, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.771504824471361, | |
| "grad_norm": 0.17609256505966187, | |
| "learning_rate": 0.00047325653157786465, | |
| "loss": 0.0566, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.771504824471361, | |
| "eval_cer": 0.01851303317535545, | |
| "eval_loss": 0.03628876060247421, | |
| "eval_runtime": 22.4829, | |
| "eval_samples_per_second": 4.003, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.06702412868632708, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.874153151303634, | |
| "grad_norm": 0.19802771508693695, | |
| "learning_rate": 0.00047222793663855174, | |
| "loss": 0.0568, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 2.874153151303634, | |
| "eval_cer": 0.016291469194312798, | |
| "eval_loss": 0.03549469634890556, | |
| "eval_runtime": 22.5619, | |
| "eval_samples_per_second": 3.989, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.058981233243967826, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 2.9768014781359065, | |
| "grad_norm": 0.19432678818702698, | |
| "learning_rate": 0.00047119934169923884, | |
| "loss": 0.056, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 2.9768014781359065, | |
| "eval_cer": 0.017180094786729858, | |
| "eval_loss": 0.03459760919213295, | |
| "eval_runtime": 22.4771, | |
| "eval_samples_per_second": 4.004, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.06523681858802502, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 3.079449804968179, | |
| "grad_norm": 0.17181651294231415, | |
| "learning_rate": 0.00047017074675992593, | |
| "loss": 0.0546, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 3.079449804968179, | |
| "eval_cer": 0.018809241706161137, | |
| "eval_loss": 0.03514665365219116, | |
| "eval_runtime": 23.0403, | |
| "eval_samples_per_second": 3.906, | |
| "eval_steps_per_second": 0.043, | |
| "eval_wer": 0.06881143878462913, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 3.1820981318004518, | |
| "grad_norm": 0.29426151514053345, | |
| "learning_rate": 0.000469142151820613, | |
| "loss": 0.0545, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 3.1820981318004518, | |
| "eval_cer": 0.018216824644549764, | |
| "eval_loss": 0.03719107061624527, | |
| "eval_runtime": 24.2578, | |
| "eval_samples_per_second": 3.71, | |
| "eval_steps_per_second": 0.041, | |
| "eval_wer": 0.06881143878462913, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 3.2847464586327244, | |
| "grad_norm": 0.14310035109519958, | |
| "learning_rate": 0.0004681135568813001, | |
| "loss": 0.0533, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 3.2847464586327244, | |
| "eval_cer": 0.01688388625592417, | |
| "eval_loss": 0.03583410009741783, | |
| "eval_runtime": 22.3344, | |
| "eval_samples_per_second": 4.03, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06702412868632708, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 3.387394785464997, | |
| "grad_norm": 0.16393882036209106, | |
| "learning_rate": 0.00046708496194198726, | |
| "loss": 0.0537, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 3.387394785464997, | |
| "eval_cer": 0.017476303317535545, | |
| "eval_loss": 0.03669163957238197, | |
| "eval_runtime": 22.2397, | |
| "eval_samples_per_second": 4.047, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06523681858802502, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 3.4900431122972697, | |
| "grad_norm": 0.1863625943660736, | |
| "learning_rate": 0.00046605636700267436, | |
| "loss": 0.0542, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 3.4900431122972697, | |
| "eval_cer": 0.01762440758293839, | |
| "eval_loss": 0.03613027185201645, | |
| "eval_runtime": 22.4068, | |
| "eval_samples_per_second": 4.017, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.064343163538874, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 3.592691439129542, | |
| "grad_norm": 0.1313330978155136, | |
| "learning_rate": 0.00046502777206336145, | |
| "loss": 0.0536, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 3.592691439129542, | |
| "eval_cer": 0.018216824644549764, | |
| "eval_loss": 0.03634100779891014, | |
| "eval_runtime": 22.5099, | |
| "eval_samples_per_second": 3.998, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.06523681858802502, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 3.6953397659618146, | |
| "grad_norm": 0.15501771867275238, | |
| "learning_rate": 0.00046399917712404854, | |
| "loss": 0.0541, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 3.6953397659618146, | |
| "eval_cer": 0.016587677725118485, | |
| "eval_loss": 0.03412258252501488, | |
| "eval_runtime": 22.4232, | |
| "eval_samples_per_second": 4.014, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06166219839142091, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 3.7979880927940872, | |
| "grad_norm": 0.1870546042919159, | |
| "learning_rate": 0.00046297058218473564, | |
| "loss": 0.0538, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 3.7979880927940872, | |
| "eval_cer": 0.01762440758293839, | |
| "eval_loss": 0.03531961515545845, | |
| "eval_runtime": 22.4977, | |
| "eval_samples_per_second": 4.0, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.06523681858802502, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 3.90063641962636, | |
| "grad_norm": 0.1889723688364029, | |
| "learning_rate": 0.00046194198724542273, | |
| "loss": 0.054, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 3.90063641962636, | |
| "eval_cer": 0.014514218009478674, | |
| "eval_loss": 0.03371370583772659, | |
| "eval_runtime": 24.169, | |
| "eval_samples_per_second": 3.724, | |
| "eval_steps_per_second": 0.041, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 4.0032847464586325, | |
| "grad_norm": 0.17405888438224792, | |
| "learning_rate": 0.0004609133923061098, | |
| "loss": 0.0535, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 4.0032847464586325, | |
| "eval_cer": 0.015995260663507108, | |
| "eval_loss": 0.03488326445221901, | |
| "eval_runtime": 23.3428, | |
| "eval_samples_per_second": 3.856, | |
| "eval_steps_per_second": 0.043, | |
| "eval_wer": 0.058981233243967826, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 4.105933073290905, | |
| "grad_norm": 0.21957945823669434, | |
| "learning_rate": 0.00045988479736679697, | |
| "loss": 0.0519, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 4.105933073290905, | |
| "eval_cer": 0.015847156398104266, | |
| "eval_loss": 0.03299200162291527, | |
| "eval_runtime": 22.9291, | |
| "eval_samples_per_second": 3.925, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.06076854334226988, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 4.208581400123178, | |
| "grad_norm": 0.1985115259885788, | |
| "learning_rate": 0.00045885620242748406, | |
| "loss": 0.0513, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 4.208581400123178, | |
| "eval_cer": 0.015847156398104266, | |
| "eval_loss": 0.0352088063955307, | |
| "eval_runtime": 22.3573, | |
| "eval_samples_per_second": 4.026, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06344950848972297, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 4.3112297269554505, | |
| "grad_norm": 0.2313787192106247, | |
| "learning_rate": 0.00045782760748817116, | |
| "loss": 0.0515, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 4.3112297269554505, | |
| "eval_cer": 0.01688388625592417, | |
| "eval_loss": 0.03440188989043236, | |
| "eval_runtime": 23.1819, | |
| "eval_samples_per_second": 3.882, | |
| "eval_steps_per_second": 0.043, | |
| "eval_wer": 0.06344950848972297, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 4.413878053787723, | |
| "grad_norm": 0.14888563752174377, | |
| "learning_rate": 0.00045679901254885825, | |
| "loss": 0.0512, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 4.413878053787723, | |
| "eval_cer": 0.017031990521327013, | |
| "eval_loss": 0.03430500999093056, | |
| "eval_runtime": 22.7046, | |
| "eval_samples_per_second": 3.964, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.06344950848972297, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 4.516526380619996, | |
| "grad_norm": 0.1658962070941925, | |
| "learning_rate": 0.00045577041760954534, | |
| "loss": 0.0513, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 4.516526380619996, | |
| "eval_cer": 0.015106635071090047, | |
| "eval_loss": 0.03472462296485901, | |
| "eval_runtime": 22.7468, | |
| "eval_samples_per_second": 3.957, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05540661304736372, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 4.619174707452268, | |
| "grad_norm": 0.2193230837583542, | |
| "learning_rate": 0.00045474182267023244, | |
| "loss": 0.0516, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 4.619174707452268, | |
| "eval_cer": 0.015550947867298577, | |
| "eval_loss": 0.03404483199119568, | |
| "eval_runtime": 22.3387, | |
| "eval_samples_per_second": 4.029, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06166219839142091, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 4.721823034284541, | |
| "grad_norm": 0.2104436755180359, | |
| "learning_rate": 0.00045371322773091953, | |
| "loss": 0.0515, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 4.721823034284541, | |
| "eval_cer": 0.015847156398104266, | |
| "eval_loss": 0.033337026834487915, | |
| "eval_runtime": 22.8456, | |
| "eval_samples_per_second": 3.939, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.06344950848972297, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 4.824471361116814, | |
| "grad_norm": 0.18940144777297974, | |
| "learning_rate": 0.0004526846327916067, | |
| "loss": 0.0512, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 4.824471361116814, | |
| "eval_cer": 0.015995260663507108, | |
| "eval_loss": 0.03273012861609459, | |
| "eval_runtime": 22.7058, | |
| "eval_samples_per_second": 3.964, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.0580875781948168, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 4.927119687949086, | |
| "grad_norm": 0.1933116912841797, | |
| "learning_rate": 0.00045165603785229377, | |
| "loss": 0.0517, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 4.927119687949086, | |
| "eval_cer": 0.01525473933649289, | |
| "eval_loss": 0.03291744366288185, | |
| "eval_runtime": 22.6758, | |
| "eval_samples_per_second": 3.969, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05987488829311886, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 5.029768014781359, | |
| "grad_norm": 0.18987509608268738, | |
| "learning_rate": 0.00045062744291298086, | |
| "loss": 0.0508, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 5.029768014781359, | |
| "eval_cer": 0.017180094786729858, | |
| "eval_loss": 0.03271958604454994, | |
| "eval_runtime": 22.4963, | |
| "eval_samples_per_second": 4.001, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.06255585344057193, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 5.132416341613632, | |
| "grad_norm": 0.1620320975780487, | |
| "learning_rate": 0.00044959884797366796, | |
| "loss": 0.0491, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 5.132416341613632, | |
| "eval_cer": 0.014218009478672985, | |
| "eval_loss": 0.03121078759431839, | |
| "eval_runtime": 22.8979, | |
| "eval_samples_per_second": 3.93, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05451295799821269, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 5.235064668445904, | |
| "grad_norm": 0.1285402774810791, | |
| "learning_rate": 0.00044857025303435505, | |
| "loss": 0.0493, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 5.235064668445904, | |
| "eval_cer": 0.015550947867298577, | |
| "eval_loss": 0.03293353319168091, | |
| "eval_runtime": 22.5976, | |
| "eval_samples_per_second": 3.983, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.06166219839142091, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 5.337712995278177, | |
| "grad_norm": 0.24566827714443207, | |
| "learning_rate": 0.00044754165809504214, | |
| "loss": 0.0498, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 5.337712995278177, | |
| "eval_cer": 0.013181279620853081, | |
| "eval_loss": 0.030513830482959747, | |
| "eval_runtime": 22.5054, | |
| "eval_samples_per_second": 3.999, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05183199285075961, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 5.44036132211045, | |
| "grad_norm": 0.18935276567935944, | |
| "learning_rate": 0.00044651306315572923, | |
| "loss": 0.0498, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 5.44036132211045, | |
| "eval_cer": 0.013773696682464455, | |
| "eval_loss": 0.03185874596238136, | |
| "eval_runtime": 22.2794, | |
| "eval_samples_per_second": 4.04, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0580875781948168, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 5.543009648942722, | |
| "grad_norm": 0.3019377291202545, | |
| "learning_rate": 0.0004454844682164164, | |
| "loss": 0.0498, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 5.543009648942722, | |
| "eval_cer": 0.01643957345971564, | |
| "eval_loss": 0.03230896592140198, | |
| "eval_runtime": 22.4205, | |
| "eval_samples_per_second": 4.014, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.064343163538874, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 5.645657975774995, | |
| "grad_norm": 0.19573438167572021, | |
| "learning_rate": 0.0004444558732771035, | |
| "loss": 0.0499, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 5.645657975774995, | |
| "eval_cer": 0.014958530805687204, | |
| "eval_loss": 0.03094463422894478, | |
| "eval_runtime": 22.3739, | |
| "eval_samples_per_second": 4.023, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.058981233243967826, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 5.748306302607268, | |
| "grad_norm": 0.19702386856079102, | |
| "learning_rate": 0.00044342727833779057, | |
| "loss": 0.0496, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 5.748306302607268, | |
| "eval_cer": 0.01643957345971564, | |
| "eval_loss": 0.031046954914927483, | |
| "eval_runtime": 22.342, | |
| "eval_samples_per_second": 4.028, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.058981233243967826, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 5.85095462943954, | |
| "grad_norm": 0.21981871128082275, | |
| "learning_rate": 0.00044239868339847766, | |
| "loss": 0.0494, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 5.85095462943954, | |
| "eval_cer": 0.014069905213270142, | |
| "eval_loss": 0.03140529617667198, | |
| "eval_runtime": 22.392, | |
| "eval_samples_per_second": 4.019, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05719392314566577, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 5.953602956271813, | |
| "grad_norm": 0.1707638055086136, | |
| "learning_rate": 0.00044137008845916475, | |
| "loss": 0.0498, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 5.953602956271813, | |
| "eval_cer": 0.014218009478672985, | |
| "eval_loss": 0.031634390354156494, | |
| "eval_runtime": 22.7229, | |
| "eval_samples_per_second": 3.961, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05451295799821269, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 6.056251283104086, | |
| "grad_norm": 0.18458805978298187, | |
| "learning_rate": 0.00044034149351985185, | |
| "loss": 0.0481, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 6.056251283104086, | |
| "eval_cer": 0.013625592417061612, | |
| "eval_loss": 0.03125843033194542, | |
| "eval_runtime": 22.6851, | |
| "eval_samples_per_second": 3.967, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05540661304736372, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 6.158899609936358, | |
| "grad_norm": 0.176268070936203, | |
| "learning_rate": 0.00043931289858053894, | |
| "loss": 0.048, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 6.158899609936358, | |
| "eval_cer": 0.013477488151658768, | |
| "eval_loss": 0.0313909687101841, | |
| "eval_runtime": 22.4652, | |
| "eval_samples_per_second": 4.006, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05540661304736372, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 6.261547936768631, | |
| "grad_norm": 0.21893835067749023, | |
| "learning_rate": 0.0004382843036412261, | |
| "loss": 0.0481, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 6.261547936768631, | |
| "eval_cer": 0.015402843601895734, | |
| "eval_loss": 0.030466848984360695, | |
| "eval_runtime": 22.4624, | |
| "eval_samples_per_second": 4.007, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06255585344057193, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 6.3641962636009035, | |
| "grad_norm": 0.17575185000896454, | |
| "learning_rate": 0.0004372557087019132, | |
| "loss": 0.0481, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 6.3641962636009035, | |
| "eval_cer": 0.012588862559241706, | |
| "eval_loss": 0.029415711760520935, | |
| "eval_runtime": 22.6218, | |
| "eval_samples_per_second": 3.978, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 6.466844590433176, | |
| "grad_norm": 0.21119283139705658, | |
| "learning_rate": 0.0004362271137626003, | |
| "loss": 0.0484, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 6.466844590433176, | |
| "eval_cer": 0.013329383886255925, | |
| "eval_loss": 0.030311500653624535, | |
| "eval_runtime": 22.4678, | |
| "eval_samples_per_second": 4.006, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 6.569492917265449, | |
| "grad_norm": 0.20543061196804047, | |
| "learning_rate": 0.0004351985188232874, | |
| "loss": 0.0475, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 6.569492917265449, | |
| "eval_cer": 0.013625592417061612, | |
| "eval_loss": 0.030118942260742188, | |
| "eval_runtime": 22.5229, | |
| "eval_samples_per_second": 3.996, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 6.6721412440977215, | |
| "grad_norm": 0.4000137448310852, | |
| "learning_rate": 0.0004341699238839745, | |
| "loss": 0.0481, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 6.6721412440977215, | |
| "eval_cer": 0.015106635071090047, | |
| "eval_loss": 0.030896518379449844, | |
| "eval_runtime": 22.5328, | |
| "eval_samples_per_second": 3.994, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.0580875781948168, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 6.774789570929994, | |
| "grad_norm": 0.2505108118057251, | |
| "learning_rate": 0.0004331413289446616, | |
| "loss": 0.0486, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 6.774789570929994, | |
| "eval_cer": 0.012885071090047393, | |
| "eval_loss": 0.030706828460097313, | |
| "eval_runtime": 22.8491, | |
| "eval_samples_per_second": 3.939, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 6.877437897762267, | |
| "grad_norm": 0.1690637618303299, | |
| "learning_rate": 0.00043211273400534876, | |
| "loss": 0.0478, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 6.877437897762267, | |
| "eval_cer": 0.014514218009478674, | |
| "eval_loss": 0.02974248118698597, | |
| "eval_runtime": 22.6356, | |
| "eval_samples_per_second": 3.976, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05540661304736372, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 6.980086224594539, | |
| "grad_norm": 0.2266341745853424, | |
| "learning_rate": 0.00043108413906603585, | |
| "loss": 0.0481, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 6.980086224594539, | |
| "eval_cer": 0.014958530805687204, | |
| "eval_loss": 0.030792562291026115, | |
| "eval_runtime": 22.6232, | |
| "eval_samples_per_second": 3.978, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.0580875781948168, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 7.082734551426812, | |
| "grad_norm": 0.2072857916355133, | |
| "learning_rate": 0.00043005554412672294, | |
| "loss": 0.0462, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 7.082734551426812, | |
| "eval_cer": 0.013181279620853081, | |
| "eval_loss": 0.02916835993528366, | |
| "eval_runtime": 23.5494, | |
| "eval_samples_per_second": 3.822, | |
| "eval_steps_per_second": 0.042, | |
| "eval_wer": 0.05272564789991063, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 7.185382878259085, | |
| "grad_norm": 0.25637751817703247, | |
| "learning_rate": 0.00042902694918741004, | |
| "loss": 0.0466, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 7.185382878259085, | |
| "eval_cer": 0.013329383886255925, | |
| "eval_loss": 0.028768625110387802, | |
| "eval_runtime": 23.0948, | |
| "eval_samples_per_second": 3.897, | |
| "eval_steps_per_second": 0.043, | |
| "eval_wer": 0.05451295799821269, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 7.288031205091357, | |
| "grad_norm": 0.16115036606788635, | |
| "learning_rate": 0.00042799835424809713, | |
| "loss": 0.0465, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 7.288031205091357, | |
| "eval_cer": 0.013921800947867298, | |
| "eval_loss": 0.029431801289319992, | |
| "eval_runtime": 22.898, | |
| "eval_samples_per_second": 3.93, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05451295799821269, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 7.39067953192363, | |
| "grad_norm": 0.3072957396507263, | |
| "learning_rate": 0.0004269697593087842, | |
| "loss": 0.0464, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 7.39067953192363, | |
| "eval_cer": 0.013181279620853081, | |
| "eval_loss": 0.028679879382252693, | |
| "eval_runtime": 22.6337, | |
| "eval_samples_per_second": 3.976, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05183199285075961, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 7.493327858755903, | |
| "grad_norm": 0.3598809242248535, | |
| "learning_rate": 0.0004259411643694713, | |
| "loss": 0.0472, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 7.493327858755903, | |
| "eval_cer": 0.013181279620853081, | |
| "eval_loss": 0.030514726415276527, | |
| "eval_runtime": 22.6134, | |
| "eval_samples_per_second": 3.98, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 7.5959761855881744, | |
| "grad_norm": 0.24177242815494537, | |
| "learning_rate": 0.00042491256943015846, | |
| "loss": 0.0461, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 7.5959761855881744, | |
| "eval_cer": 0.013477488151658768, | |
| "eval_loss": 0.02993646450340748, | |
| "eval_runtime": 22.6695, | |
| "eval_samples_per_second": 3.97, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05183199285075961, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 7.698624512420448, | |
| "grad_norm": 0.14063900709152222, | |
| "learning_rate": 0.00042388397449084556, | |
| "loss": 0.0469, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 7.698624512420448, | |
| "eval_cer": 0.013773696682464455, | |
| "eval_loss": 0.02932840585708618, | |
| "eval_runtime": 22.5476, | |
| "eval_samples_per_second": 3.992, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05451295799821269, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 7.80127283925272, | |
| "grad_norm": 0.20371408760547638, | |
| "learning_rate": 0.00042285537955153265, | |
| "loss": 0.0473, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 7.80127283925272, | |
| "eval_cer": 0.014069905213270142, | |
| "eval_loss": 0.029385393485426903, | |
| "eval_runtime": 22.4143, | |
| "eval_samples_per_second": 4.015, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05719392314566577, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 7.903921166084993, | |
| "grad_norm": 0.17325064539909363, | |
| "learning_rate": 0.00042182678461221974, | |
| "loss": 0.047, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 7.903921166084993, | |
| "eval_cer": 0.012144549763033176, | |
| "eval_loss": 0.02776852808892727, | |
| "eval_runtime": 22.3852, | |
| "eval_samples_per_second": 4.021, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.049151027703306524, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 8.006569492917265, | |
| "grad_norm": 0.261836975812912, | |
| "learning_rate": 0.00042079818967290683, | |
| "loss": 0.0458, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 8.006569492917265, | |
| "eval_cer": 0.01229265402843602, | |
| "eval_loss": 0.02698938362300396, | |
| "eval_runtime": 22.4901, | |
| "eval_samples_per_second": 4.002, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 8.109217819749539, | |
| "grad_norm": 0.24788102507591248, | |
| "learning_rate": 0.00041976959473359393, | |
| "loss": 0.0448, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 8.109217819749539, | |
| "eval_cer": 0.014069905213270142, | |
| "eval_loss": 0.02948344498872757, | |
| "eval_runtime": 22.4348, | |
| "eval_samples_per_second": 4.012, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05719392314566577, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 8.21186614658181, | |
| "grad_norm": 0.22888223826885223, | |
| "learning_rate": 0.000418740999794281, | |
| "loss": 0.0455, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 8.21186614658181, | |
| "eval_cer": 0.012440758293838863, | |
| "eval_loss": 0.027837086468935013, | |
| "eval_runtime": 22.6805, | |
| "eval_samples_per_second": 3.968, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 8.314514473414084, | |
| "grad_norm": 0.15464870631694794, | |
| "learning_rate": 0.00041771240485496817, | |
| "loss": 0.0456, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 8.314514473414084, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.03025979734957218, | |
| "eval_runtime": 22.4815, | |
| "eval_samples_per_second": 4.003, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.06076854334226988, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 8.417162800246356, | |
| "grad_norm": 0.2563960552215576, | |
| "learning_rate": 0.00041668380991565526, | |
| "loss": 0.0449, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 8.417162800246356, | |
| "eval_cer": 0.013329383886255925, | |
| "eval_loss": 0.028577908873558044, | |
| "eval_runtime": 22.508, | |
| "eval_samples_per_second": 3.999, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 8.51981112707863, | |
| "grad_norm": 0.2178841084241867, | |
| "learning_rate": 0.00041565521497634235, | |
| "loss": 0.0453, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 8.51981112707863, | |
| "eval_cer": 0.011700236966824644, | |
| "eval_loss": 0.02777865342795849, | |
| "eval_runtime": 22.4852, | |
| "eval_samples_per_second": 4.003, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 8.622459453910901, | |
| "grad_norm": 0.16487497091293335, | |
| "learning_rate": 0.00041462662003702945, | |
| "loss": 0.0452, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 8.622459453910901, | |
| "eval_cer": 0.014218009478672985, | |
| "eval_loss": 0.02851984277367592, | |
| "eval_runtime": 22.544, | |
| "eval_samples_per_second": 3.992, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05451295799821269, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 8.725107780743175, | |
| "grad_norm": 0.1772727370262146, | |
| "learning_rate": 0.00041359802509771654, | |
| "loss": 0.0455, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 8.725107780743175, | |
| "eval_cer": 0.013477488151658768, | |
| "eval_loss": 0.027403153479099274, | |
| "eval_runtime": 22.2753, | |
| "eval_samples_per_second": 4.04, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05272564789991063, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 8.827756107575446, | |
| "grad_norm": 0.2657695710659027, | |
| "learning_rate": 0.00041256943015840363, | |
| "loss": 0.0454, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 8.827756107575446, | |
| "eval_cer": 0.014218009478672985, | |
| "eval_loss": 0.027919290587306023, | |
| "eval_runtime": 22.8782, | |
| "eval_samples_per_second": 3.934, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05630026809651475, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 8.93040443440772, | |
| "grad_norm": 0.18787504732608795, | |
| "learning_rate": 0.00041154083521909073, | |
| "loss": 0.0455, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 8.93040443440772, | |
| "eval_cer": 0.014958530805687204, | |
| "eval_loss": 0.027214767411351204, | |
| "eval_runtime": 22.7677, | |
| "eval_samples_per_second": 3.953, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05630026809651475, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 9.033052761239992, | |
| "grad_norm": 0.21097755432128906, | |
| "learning_rate": 0.0004105122402797779, | |
| "loss": 0.0451, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 9.033052761239992, | |
| "eval_cer": 0.01273696682464455, | |
| "eval_loss": 0.02636747434735298, | |
| "eval_runtime": 22.9582, | |
| "eval_samples_per_second": 3.92, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05183199285075961, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 9.135701088072265, | |
| "grad_norm": 0.17829887568950653, | |
| "learning_rate": 0.00040948364534046497, | |
| "loss": 0.0437, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 9.135701088072265, | |
| "eval_cer": 0.01110781990521327, | |
| "eval_loss": 0.026946688070893288, | |
| "eval_runtime": 22.62, | |
| "eval_samples_per_second": 3.979, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.045576407506702415, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 9.238349414904537, | |
| "grad_norm": 0.1892678588628769, | |
| "learning_rate": 0.00040845505040115206, | |
| "loss": 0.0436, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 9.238349414904537, | |
| "eval_cer": 0.013329383886255925, | |
| "eval_loss": 0.0261703971773386, | |
| "eval_runtime": 22.7469, | |
| "eval_samples_per_second": 3.957, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05183199285075961, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 9.34099774173681, | |
| "grad_norm": 0.1827981173992157, | |
| "learning_rate": 0.00040742645546183915, | |
| "loss": 0.0442, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 9.34099774173681, | |
| "eval_cer": 0.013625592417061612, | |
| "eval_loss": 0.026994889602065086, | |
| "eval_runtime": 22.3199, | |
| "eval_samples_per_second": 4.032, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 9.443646068569082, | |
| "grad_norm": 0.26229721307754517, | |
| "learning_rate": 0.00040639786052252625, | |
| "loss": 0.0442, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 9.443646068569082, | |
| "eval_cer": 0.012588862559241706, | |
| "eval_loss": 0.026131337508559227, | |
| "eval_runtime": 30.6242, | |
| "eval_samples_per_second": 2.939, | |
| "eval_steps_per_second": 0.033, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 9.546294395401356, | |
| "grad_norm": 0.31516391038894653, | |
| "learning_rate": 0.00040536926558321334, | |
| "loss": 0.0442, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 9.546294395401356, | |
| "eval_cer": 0.012588862559241706, | |
| "eval_loss": 0.023602332919836044, | |
| "eval_runtime": 27.5058, | |
| "eval_samples_per_second": 3.272, | |
| "eval_steps_per_second": 0.036, | |
| "eval_wer": 0.049151027703306524, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 9.648942722233627, | |
| "grad_norm": 0.19427119195461273, | |
| "learning_rate": 0.00040434067064390043, | |
| "loss": 0.0443, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 9.648942722233627, | |
| "eval_cer": 0.013033175355450236, | |
| "eval_loss": 0.02646990306675434, | |
| "eval_runtime": 27.6213, | |
| "eval_samples_per_second": 3.258, | |
| "eval_steps_per_second": 0.036, | |
| "eval_wer": 0.049151027703306524, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 9.751591049065901, | |
| "grad_norm": 0.26338282227516174, | |
| "learning_rate": 0.0004033120757045876, | |
| "loss": 0.0447, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 9.751591049065901, | |
| "eval_cer": 0.013625592417061612, | |
| "eval_loss": 0.025444395840168, | |
| "eval_runtime": 25.0834, | |
| "eval_samples_per_second": 3.588, | |
| "eval_steps_per_second": 0.04, | |
| "eval_wer": 0.05272564789991063, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 9.854239375898173, | |
| "grad_norm": 0.25808289647102356, | |
| "learning_rate": 0.0004022834807652747, | |
| "loss": 0.0445, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 9.854239375898173, | |
| "eval_cer": 0.014958530805687204, | |
| "eval_loss": 0.026209862902760506, | |
| "eval_runtime": 23.4245, | |
| "eval_samples_per_second": 3.842, | |
| "eval_steps_per_second": 0.043, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 9.956887702730446, | |
| "grad_norm": 0.18842875957489014, | |
| "learning_rate": 0.00040125488582596177, | |
| "loss": 0.0445, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 9.956887702730446, | |
| "eval_cer": 0.012440758293838863, | |
| "eval_loss": 0.027286237105727196, | |
| "eval_runtime": 22.6774, | |
| "eval_samples_per_second": 3.969, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 10.059536029562718, | |
| "grad_norm": 0.2742888331413269, | |
| "learning_rate": 0.00040022629088664886, | |
| "loss": 0.0432, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 10.059536029562718, | |
| "eval_cer": 0.011404028436018957, | |
| "eval_loss": 0.027185438200831413, | |
| "eval_runtime": 27.5005, | |
| "eval_samples_per_second": 3.273, | |
| "eval_steps_per_second": 0.036, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 10.162184356394992, | |
| "grad_norm": 0.20179295539855957, | |
| "learning_rate": 0.00039919769594733595, | |
| "loss": 0.0428, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 10.162184356394992, | |
| "eval_cer": 0.01481042654028436, | |
| "eval_loss": 0.02597665973007679, | |
| "eval_runtime": 25.1306, | |
| "eval_samples_per_second": 3.581, | |
| "eval_steps_per_second": 0.04, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 10.264832683227263, | |
| "grad_norm": 0.16221770644187927, | |
| "learning_rate": 0.00039816910100802305, | |
| "loss": 0.0429, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 10.264832683227263, | |
| "eval_cer": 0.011848341232227487, | |
| "eval_loss": 0.027527980506420135, | |
| "eval_runtime": 23.4033, | |
| "eval_samples_per_second": 3.846, | |
| "eval_steps_per_second": 0.043, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 10.367481010059535, | |
| "grad_norm": 0.17623300850391388, | |
| "learning_rate": 0.00039714050606871014, | |
| "loss": 0.0435, | |
| "step": 101000 | |
| }, | |
| { | |
| "epoch": 10.367481010059535, | |
| "eval_cer": 0.014218009478672985, | |
| "eval_loss": 0.026551930233836174, | |
| "eval_runtime": 23.1192, | |
| "eval_samples_per_second": 3.893, | |
| "eval_steps_per_second": 0.043, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 101000 | |
| }, | |
| { | |
| "epoch": 10.470129336891809, | |
| "grad_norm": 0.29380717873573303, | |
| "learning_rate": 0.0003961119111293973, | |
| "loss": 0.0431, | |
| "step": 102000 | |
| }, | |
| { | |
| "epoch": 10.470129336891809, | |
| "eval_cer": 0.017476303317535545, | |
| "eval_loss": 0.026500999927520752, | |
| "eval_runtime": 23.0878, | |
| "eval_samples_per_second": 3.898, | |
| "eval_steps_per_second": 0.043, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 102000 | |
| }, | |
| { | |
| "epoch": 10.572777663724082, | |
| "grad_norm": 0.19392183423042297, | |
| "learning_rate": 0.0003950833161900844, | |
| "loss": 0.043, | |
| "step": 103000 | |
| }, | |
| { | |
| "epoch": 10.572777663724082, | |
| "eval_cer": 0.013033175355450236, | |
| "eval_loss": 0.02609255537390709, | |
| "eval_runtime": 22.8011, | |
| "eval_samples_per_second": 3.947, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.049151027703306524, | |
| "step": 103000 | |
| }, | |
| { | |
| "epoch": 10.675425990556354, | |
| "grad_norm": 0.17925652861595154, | |
| "learning_rate": 0.0003940547212507715, | |
| "loss": 0.0433, | |
| "step": 104000 | |
| }, | |
| { | |
| "epoch": 10.675425990556354, | |
| "eval_cer": 0.01273696682464455, | |
| "eval_loss": 0.027248414233326912, | |
| "eval_runtime": 22.7524, | |
| "eval_samples_per_second": 3.956, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 104000 | |
| }, | |
| { | |
| "epoch": 10.778074317388626, | |
| "grad_norm": 0.21368491649627686, | |
| "learning_rate": 0.00039302612631145857, | |
| "loss": 0.0431, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 10.778074317388626, | |
| "eval_cer": 0.01273696682464455, | |
| "eval_loss": 0.025399256497621536, | |
| "eval_runtime": 22.4849, | |
| "eval_samples_per_second": 4.003, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 10.8807226442209, | |
| "grad_norm": 0.17027121782302856, | |
| "learning_rate": 0.00039199753137214566, | |
| "loss": 0.0435, | |
| "step": 106000 | |
| }, | |
| { | |
| "epoch": 10.8807226442209, | |
| "eval_cer": 0.01229265402843602, | |
| "eval_loss": 0.025961685925722122, | |
| "eval_runtime": 22.3736, | |
| "eval_samples_per_second": 4.023, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 106000 | |
| }, | |
| { | |
| "epoch": 10.983370971053171, | |
| "grad_norm": 0.3062898516654968, | |
| "learning_rate": 0.00039096893643283275, | |
| "loss": 0.0434, | |
| "step": 107000 | |
| }, | |
| { | |
| "epoch": 10.983370971053171, | |
| "eval_cer": 0.010219194312796208, | |
| "eval_loss": 0.0248806644231081, | |
| "eval_runtime": 22.4291, | |
| "eval_samples_per_second": 4.013, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 107000 | |
| }, | |
| { | |
| "epoch": 11.086019297885445, | |
| "grad_norm": 0.27476412057876587, | |
| "learning_rate": 0.00038994034149351985, | |
| "loss": 0.0418, | |
| "step": 108000 | |
| }, | |
| { | |
| "epoch": 11.086019297885445, | |
| "eval_cer": 0.011700236966824644, | |
| "eval_loss": 0.025811193510890007, | |
| "eval_runtime": 22.3629, | |
| "eval_samples_per_second": 4.025, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 108000 | |
| }, | |
| { | |
| "epoch": 11.188667624717716, | |
| "grad_norm": 0.18025143444538116, | |
| "learning_rate": 0.000388911746554207, | |
| "loss": 0.0419, | |
| "step": 109000 | |
| }, | |
| { | |
| "epoch": 11.188667624717716, | |
| "eval_cer": 0.01066350710900474, | |
| "eval_loss": 0.024703815579414368, | |
| "eval_runtime": 22.2656, | |
| "eval_samples_per_second": 4.042, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.044682752457551385, | |
| "step": 109000 | |
| }, | |
| { | |
| "epoch": 11.29131595154999, | |
| "grad_norm": 0.19146864116191864, | |
| "learning_rate": 0.0003878831516148941, | |
| "loss": 0.042, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 11.29131595154999, | |
| "eval_cer": 0.011404028436018957, | |
| "eval_loss": 0.025849131867289543, | |
| "eval_runtime": 22.6201, | |
| "eval_samples_per_second": 3.979, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 11.393964278382262, | |
| "grad_norm": 0.4587384760379791, | |
| "learning_rate": 0.0003868545566755812, | |
| "loss": 0.042, | |
| "step": 111000 | |
| }, | |
| { | |
| "epoch": 11.393964278382262, | |
| "eval_cer": 0.012440758293838863, | |
| "eval_loss": 0.025229139253497124, | |
| "eval_runtime": 22.5628, | |
| "eval_samples_per_second": 3.989, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 111000 | |
| }, | |
| { | |
| "epoch": 11.496612605214535, | |
| "grad_norm": 0.1752750277519226, | |
| "learning_rate": 0.0003858259617362683, | |
| "loss": 0.0428, | |
| "step": 112000 | |
| }, | |
| { | |
| "epoch": 11.496612605214535, | |
| "eval_cer": 0.01066350710900474, | |
| "eval_loss": 0.02614370547235012, | |
| "eval_runtime": 22.3268, | |
| "eval_samples_per_second": 4.031, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.045576407506702415, | |
| "step": 112000 | |
| }, | |
| { | |
| "epoch": 11.599260932046807, | |
| "grad_norm": 0.20651549100875854, | |
| "learning_rate": 0.00038479736679695537, | |
| "loss": 0.0428, | |
| "step": 113000 | |
| }, | |
| { | |
| "epoch": 11.599260932046807, | |
| "eval_cer": 0.01110781990521327, | |
| "eval_loss": 0.02591308392584324, | |
| "eval_runtime": 22.6162, | |
| "eval_samples_per_second": 3.979, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 113000 | |
| }, | |
| { | |
| "epoch": 11.70190925887908, | |
| "grad_norm": 0.1839723438024521, | |
| "learning_rate": 0.00038376877185764246, | |
| "loss": 0.0422, | |
| "step": 114000 | |
| }, | |
| { | |
| "epoch": 11.70190925887908, | |
| "eval_cer": 0.010515402843601895, | |
| "eval_loss": 0.026186056435108185, | |
| "eval_runtime": 22.64, | |
| "eval_samples_per_second": 3.975, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.044682752457551385, | |
| "step": 114000 | |
| }, | |
| { | |
| "epoch": 11.804557585711352, | |
| "grad_norm": 0.1559193879365921, | |
| "learning_rate": 0.00038274017691832955, | |
| "loss": 0.0426, | |
| "step": 115000 | |
| }, | |
| { | |
| "epoch": 11.804557585711352, | |
| "eval_cer": 0.011996445497630332, | |
| "eval_loss": 0.026291608810424805, | |
| "eval_runtime": 23.0438, | |
| "eval_samples_per_second": 3.906, | |
| "eval_steps_per_second": 0.043, | |
| "eval_wer": 0.045576407506702415, | |
| "step": 115000 | |
| }, | |
| { | |
| "epoch": 11.907205912543626, | |
| "grad_norm": 0.33867180347442627, | |
| "learning_rate": 0.0003817115819790167, | |
| "loss": 0.0422, | |
| "step": 116000 | |
| }, | |
| { | |
| "epoch": 11.907205912543626, | |
| "eval_cer": 0.011996445497630332, | |
| "eval_loss": 0.025052817538380623, | |
| "eval_runtime": 22.5162, | |
| "eval_samples_per_second": 3.997, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 116000 | |
| }, | |
| { | |
| "epoch": 12.009854239375898, | |
| "grad_norm": 0.2387935370206833, | |
| "learning_rate": 0.0003806829870397038, | |
| "loss": 0.042, | |
| "step": 117000 | |
| }, | |
| { | |
| "epoch": 12.009854239375898, | |
| "eval_cer": 0.011404028436018957, | |
| "eval_loss": 0.02570049837231636, | |
| "eval_runtime": 22.5888, | |
| "eval_samples_per_second": 3.984, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.044682752457551385, | |
| "step": 117000 | |
| }, | |
| { | |
| "epoch": 12.112502566208171, | |
| "grad_norm": 0.1758970469236374, | |
| "learning_rate": 0.0003796543921003909, | |
| "loss": 0.0406, | |
| "step": 118000 | |
| }, | |
| { | |
| "epoch": 12.112502566208171, | |
| "eval_cer": 0.01273696682464455, | |
| "eval_loss": 0.025880787521600723, | |
| "eval_runtime": 22.5099, | |
| "eval_samples_per_second": 3.998, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.049151027703306524, | |
| "step": 118000 | |
| }, | |
| { | |
| "epoch": 12.215150893040443, | |
| "grad_norm": 0.2268359512090683, | |
| "learning_rate": 0.000378625797161078, | |
| "loss": 0.0409, | |
| "step": 119000 | |
| }, | |
| { | |
| "epoch": 12.215150893040443, | |
| "eval_cer": 0.012588862559241706, | |
| "eval_loss": 0.024862516671419144, | |
| "eval_runtime": 22.7569, | |
| "eval_samples_per_second": 3.955, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 119000 | |
| }, | |
| { | |
| "epoch": 12.317799219872716, | |
| "grad_norm": 0.15519174933433533, | |
| "learning_rate": 0.0003775972022217651, | |
| "loss": 0.0412, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 12.317799219872716, | |
| "eval_cer": 0.0115521327014218, | |
| "eval_loss": 0.025986041873693466, | |
| "eval_runtime": 22.6794, | |
| "eval_samples_per_second": 3.968, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 12.420447546704988, | |
| "grad_norm": 0.3455216884613037, | |
| "learning_rate": 0.00037656860728245217, | |
| "loss": 0.0411, | |
| "step": 121000 | |
| }, | |
| { | |
| "epoch": 12.420447546704988, | |
| "eval_cer": 0.011700236966824644, | |
| "eval_loss": 0.025649528950452805, | |
| "eval_runtime": 22.7315, | |
| "eval_samples_per_second": 3.959, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 121000 | |
| }, | |
| { | |
| "epoch": 12.523095873537262, | |
| "grad_norm": 0.20411798357963562, | |
| "learning_rate": 0.00037554001234313926, | |
| "loss": 0.0418, | |
| "step": 122000 | |
| }, | |
| { | |
| "epoch": 12.523095873537262, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.02615606226027012, | |
| "eval_runtime": 22.7606, | |
| "eval_samples_per_second": 3.954, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 122000 | |
| }, | |
| { | |
| "epoch": 12.625744200369533, | |
| "grad_norm": 0.25552111864089966, | |
| "learning_rate": 0.0003745114174038264, | |
| "loss": 0.0414, | |
| "step": 123000 | |
| }, | |
| { | |
| "epoch": 12.625744200369533, | |
| "eval_cer": 0.01273696682464455, | |
| "eval_loss": 0.024597780779004097, | |
| "eval_runtime": 22.7241, | |
| "eval_samples_per_second": 3.961, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 123000 | |
| }, | |
| { | |
| "epoch": 12.728392527201807, | |
| "grad_norm": 0.24297872185707092, | |
| "learning_rate": 0.0003734828224645135, | |
| "loss": 0.0416, | |
| "step": 124000 | |
| }, | |
| { | |
| "epoch": 12.728392527201807, | |
| "eval_cer": 0.011996445497630332, | |
| "eval_loss": 0.0245036818087101, | |
| "eval_runtime": 22.3574, | |
| "eval_samples_per_second": 4.026, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 124000 | |
| }, | |
| { | |
| "epoch": 12.831040854034079, | |
| "grad_norm": 0.16708943247795105, | |
| "learning_rate": 0.0003724542275252006, | |
| "loss": 0.0421, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 12.831040854034079, | |
| "eval_cer": 0.010959715639810427, | |
| "eval_loss": 0.023803560063242912, | |
| "eval_runtime": 22.1003, | |
| "eval_samples_per_second": 4.072, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.044682752457551385, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 12.933689180866352, | |
| "grad_norm": 0.2480056881904602, | |
| "learning_rate": 0.0003714256325858877, | |
| "loss": 0.0417, | |
| "step": 126000 | |
| }, | |
| { | |
| "epoch": 12.933689180866352, | |
| "eval_cer": 0.010959715639810427, | |
| "eval_loss": 0.02339034155011177, | |
| "eval_runtime": 22.1887, | |
| "eval_samples_per_second": 4.056, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 126000 | |
| }, | |
| { | |
| "epoch": 13.036337507698624, | |
| "grad_norm": 0.15843307971954346, | |
| "learning_rate": 0.0003703970376465748, | |
| "loss": 0.041, | |
| "step": 127000 | |
| }, | |
| { | |
| "epoch": 13.036337507698624, | |
| "eval_cer": 0.010219194312796208, | |
| "eval_loss": 0.023780081421136856, | |
| "eval_runtime": 22.1307, | |
| "eval_samples_per_second": 4.067, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04289544235924933, | |
| "step": 127000 | |
| }, | |
| { | |
| "epoch": 13.138985834530898, | |
| "grad_norm": 0.17070743441581726, | |
| "learning_rate": 0.00036936844270726187, | |
| "loss": 0.0397, | |
| "step": 128000 | |
| }, | |
| { | |
| "epoch": 13.138985834530898, | |
| "eval_cer": 0.014514218009478674, | |
| "eval_loss": 0.0229303240776062, | |
| "eval_runtime": 22.379, | |
| "eval_samples_per_second": 4.022, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 128000 | |
| }, | |
| { | |
| "epoch": 13.24163416136317, | |
| "grad_norm": 0.1719464212656021, | |
| "learning_rate": 0.00036833984776794897, | |
| "loss": 0.0405, | |
| "step": 129000 | |
| }, | |
| { | |
| "epoch": 13.24163416136317, | |
| "eval_cer": 0.014218009478672985, | |
| "eval_loss": 0.023020587861537933, | |
| "eval_runtime": 22.2447, | |
| "eval_samples_per_second": 4.046, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.049151027703306524, | |
| "step": 129000 | |
| }, | |
| { | |
| "epoch": 13.344282488195443, | |
| "grad_norm": 0.16199146211147308, | |
| "learning_rate": 0.0003673112528286361, | |
| "loss": 0.0405, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 13.344282488195443, | |
| "eval_cer": 0.010367298578199052, | |
| "eval_loss": 0.023113010451197624, | |
| "eval_runtime": 22.549, | |
| "eval_samples_per_second": 3.991, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.043789097408400354, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 13.446930815027715, | |
| "grad_norm": 0.21660035848617554, | |
| "learning_rate": 0.0003662826578893232, | |
| "loss": 0.0406, | |
| "step": 131000 | |
| }, | |
| { | |
| "epoch": 13.446930815027715, | |
| "eval_cer": 0.01229265402843602, | |
| "eval_loss": 0.023945845663547516, | |
| "eval_runtime": 22.3889, | |
| "eval_samples_per_second": 4.02, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.049151027703306524, | |
| "step": 131000 | |
| }, | |
| { | |
| "epoch": 13.549579141859988, | |
| "grad_norm": 0.3124329447746277, | |
| "learning_rate": 0.0003652540629500103, | |
| "loss": 0.0406, | |
| "step": 132000 | |
| }, | |
| { | |
| "epoch": 13.549579141859988, | |
| "eval_cer": 0.01110781990521327, | |
| "eval_loss": 0.024438710883259773, | |
| "eval_runtime": 22.5799, | |
| "eval_samples_per_second": 3.986, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.043789097408400354, | |
| "step": 132000 | |
| }, | |
| { | |
| "epoch": 13.65222746869226, | |
| "grad_norm": 0.16738218069076538, | |
| "learning_rate": 0.0003642254680106974, | |
| "loss": 0.0406, | |
| "step": 133000 | |
| }, | |
| { | |
| "epoch": 13.65222746869226, | |
| "eval_cer": 0.011848341232227487, | |
| "eval_loss": 0.024815011769533157, | |
| "eval_runtime": 22.3814, | |
| "eval_samples_per_second": 4.021, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 133000 | |
| }, | |
| { | |
| "epoch": 13.754875795524534, | |
| "grad_norm": 0.1927761733531952, | |
| "learning_rate": 0.0003631968730713845, | |
| "loss": 0.0407, | |
| "step": 134000 | |
| }, | |
| { | |
| "epoch": 13.754875795524534, | |
| "eval_cer": 0.010959715639810427, | |
| "eval_loss": 0.023673338815569878, | |
| "eval_runtime": 22.7336, | |
| "eval_samples_per_second": 3.959, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.045576407506702415, | |
| "step": 134000 | |
| }, | |
| { | |
| "epoch": 13.857524122356805, | |
| "grad_norm": 0.17141355574131012, | |
| "learning_rate": 0.0003621682781320716, | |
| "loss": 0.0411, | |
| "step": 135000 | |
| }, | |
| { | |
| "epoch": 13.857524122356805, | |
| "eval_cer": 0.010959715639810427, | |
| "eval_loss": 0.02259986102581024, | |
| "eval_runtime": 22.6132, | |
| "eval_samples_per_second": 3.98, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.041108132260947276, | |
| "step": 135000 | |
| }, | |
| { | |
| "epoch": 13.960172449189079, | |
| "grad_norm": 0.24508166313171387, | |
| "learning_rate": 0.00036113968319275867, | |
| "loss": 0.0405, | |
| "step": 136000 | |
| }, | |
| { | |
| "epoch": 13.960172449189079, | |
| "eval_cer": 0.01066350710900474, | |
| "eval_loss": 0.022703783586621284, | |
| "eval_runtime": 22.7539, | |
| "eval_samples_per_second": 3.955, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.04289544235924933, | |
| "step": 136000 | |
| }, | |
| { | |
| "epoch": 14.06282077602135, | |
| "grad_norm": 0.16717751324176788, | |
| "learning_rate": 0.0003601110882534458, | |
| "loss": 0.0401, | |
| "step": 137000 | |
| }, | |
| { | |
| "epoch": 14.06282077602135, | |
| "eval_cer": 0.009922985781990521, | |
| "eval_loss": 0.02189534902572632, | |
| "eval_runtime": 22.4478, | |
| "eval_samples_per_second": 4.009, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.040214477211796246, | |
| "step": 137000 | |
| }, | |
| { | |
| "epoch": 14.165469102853624, | |
| "grad_norm": 0.42751288414001465, | |
| "learning_rate": 0.0003590824933141329, | |
| "loss": 0.0391, | |
| "step": 138000 | |
| }, | |
| { | |
| "epoch": 14.165469102853624, | |
| "eval_cer": 0.010367298578199052, | |
| "eval_loss": 0.022342221811413765, | |
| "eval_runtime": 22.5149, | |
| "eval_samples_per_second": 3.997, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.040214477211796246, | |
| "step": 138000 | |
| }, | |
| { | |
| "epoch": 14.268117429685896, | |
| "grad_norm": 0.2213069647550583, | |
| "learning_rate": 0.00035805389837482, | |
| "loss": 0.0394, | |
| "step": 139000 | |
| }, | |
| { | |
| "epoch": 14.268117429685896, | |
| "eval_cer": 0.010959715639810427, | |
| "eval_loss": 0.021898576989769936, | |
| "eval_runtime": 22.2749, | |
| "eval_samples_per_second": 4.04, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.043789097408400354, | |
| "step": 139000 | |
| }, | |
| { | |
| "epoch": 14.37076575651817, | |
| "grad_norm": 0.17157946527004242, | |
| "learning_rate": 0.0003570253034355071, | |
| "loss": 0.0395, | |
| "step": 140000 | |
| }, | |
| { | |
| "epoch": 14.37076575651817, | |
| "eval_cer": 0.010811611374407584, | |
| "eval_loss": 0.023508407175540924, | |
| "eval_runtime": 22.3534, | |
| "eval_samples_per_second": 4.026, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.043789097408400354, | |
| "step": 140000 | |
| }, | |
| { | |
| "epoch": 14.473414083350441, | |
| "grad_norm": 0.26436519622802734, | |
| "learning_rate": 0.0003559967084961942, | |
| "loss": 0.0398, | |
| "step": 141000 | |
| }, | |
| { | |
| "epoch": 14.473414083350441, | |
| "eval_cer": 0.011255924170616114, | |
| "eval_loss": 0.022889673709869385, | |
| "eval_runtime": 22.2676, | |
| "eval_samples_per_second": 4.042, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.045576407506702415, | |
| "step": 141000 | |
| }, | |
| { | |
| "epoch": 14.576062410182715, | |
| "grad_norm": 0.15638813376426697, | |
| "learning_rate": 0.0003549681135568813, | |
| "loss": 0.0399, | |
| "step": 142000 | |
| }, | |
| { | |
| "epoch": 14.576062410182715, | |
| "eval_cer": 0.009774881516587678, | |
| "eval_loss": 0.022171661257743835, | |
| "eval_runtime": 22.6709, | |
| "eval_samples_per_second": 3.97, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.0420017873100983, | |
| "step": 142000 | |
| }, | |
| { | |
| "epoch": 14.678710737014987, | |
| "grad_norm": 0.22069737315177917, | |
| "learning_rate": 0.0003539395186175684, | |
| "loss": 0.0397, | |
| "step": 143000 | |
| }, | |
| { | |
| "epoch": 14.678710737014987, | |
| "eval_cer": 0.014366113744075829, | |
| "eval_loss": 0.02216203510761261, | |
| "eval_runtime": 22.5252, | |
| "eval_samples_per_second": 3.996, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 143000 | |
| }, | |
| { | |
| "epoch": 14.78135906384726, | |
| "grad_norm": 0.25842490792274475, | |
| "learning_rate": 0.0003529109236782555, | |
| "loss": 0.0398, | |
| "step": 144000 | |
| }, | |
| { | |
| "epoch": 14.78135906384726, | |
| "eval_cer": 0.010515402843601895, | |
| "eval_loss": 0.023699576035141945, | |
| "eval_runtime": 22.5737, | |
| "eval_samples_per_second": 3.987, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.044682752457551385, | |
| "step": 144000 | |
| }, | |
| { | |
| "epoch": 14.884007390679532, | |
| "grad_norm": 0.2184634506702423, | |
| "learning_rate": 0.0003518823287389426, | |
| "loss": 0.0402, | |
| "step": 145000 | |
| }, | |
| { | |
| "epoch": 14.884007390679532, | |
| "eval_cer": 0.011255924170616114, | |
| "eval_loss": 0.022893035784363747, | |
| "eval_runtime": 22.3846, | |
| "eval_samples_per_second": 4.021, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 145000 | |
| }, | |
| { | |
| "epoch": 14.986655717511805, | |
| "grad_norm": 0.19810239970684052, | |
| "learning_rate": 0.0003508537337996297, | |
| "loss": 0.0403, | |
| "step": 146000 | |
| }, | |
| { | |
| "epoch": 14.986655717511805, | |
| "eval_cer": 0.011996445497630332, | |
| "eval_loss": 0.024159209802746773, | |
| "eval_runtime": 22.5604, | |
| "eval_samples_per_second": 3.989, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 146000 | |
| }, | |
| { | |
| "epoch": 15.089304044344077, | |
| "grad_norm": 0.2137177586555481, | |
| "learning_rate": 0.0003498251388603168, | |
| "loss": 0.0385, | |
| "step": 147000 | |
| }, | |
| { | |
| "epoch": 15.089304044344077, | |
| "eval_cer": 0.010811611374407584, | |
| "eval_loss": 0.022794917225837708, | |
| "eval_runtime": 22.3128, | |
| "eval_samples_per_second": 4.034, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.045576407506702415, | |
| "step": 147000 | |
| }, | |
| { | |
| "epoch": 15.19195237117635, | |
| "grad_norm": 0.1722225844860077, | |
| "learning_rate": 0.0003487965439210039, | |
| "loss": 0.0386, | |
| "step": 148000 | |
| }, | |
| { | |
| "epoch": 15.19195237117635, | |
| "eval_cer": 0.011848341232227487, | |
| "eval_loss": 0.02336839959025383, | |
| "eval_runtime": 22.3982, | |
| "eval_samples_per_second": 4.018, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 148000 | |
| }, | |
| { | |
| "epoch": 15.294600698008622, | |
| "grad_norm": 0.20236076414585114, | |
| "learning_rate": 0.000347767948981691, | |
| "loss": 0.0392, | |
| "step": 149000 | |
| }, | |
| { | |
| "epoch": 15.294600698008622, | |
| "eval_cer": 0.01229265402843602, | |
| "eval_loss": 0.02401108108460903, | |
| "eval_runtime": 22.0603, | |
| "eval_samples_per_second": 4.08, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 149000 | |
| }, | |
| { | |
| "epoch": 15.397249024840896, | |
| "grad_norm": 0.1955161690711975, | |
| "learning_rate": 0.0003467393540423781, | |
| "loss": 0.039, | |
| "step": 150000 | |
| }, | |
| { | |
| "epoch": 15.397249024840896, | |
| "eval_cer": 0.010811611374407584, | |
| "eval_loss": 0.022156517952680588, | |
| "eval_runtime": 22.3641, | |
| "eval_samples_per_second": 4.024, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.045576407506702415, | |
| "step": 150000 | |
| }, | |
| { | |
| "epoch": 15.499897351673168, | |
| "grad_norm": 0.24897447228431702, | |
| "learning_rate": 0.00034571075910306523, | |
| "loss": 0.0391, | |
| "step": 151000 | |
| }, | |
| { | |
| "epoch": 15.499897351673168, | |
| "eval_cer": 0.010219194312796208, | |
| "eval_loss": 0.022661181166768074, | |
| "eval_runtime": 22.3484, | |
| "eval_samples_per_second": 4.027, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04289544235924933, | |
| "step": 151000 | |
| }, | |
| { | |
| "epoch": 15.60254567850544, | |
| "grad_norm": 0.1920953094959259, | |
| "learning_rate": 0.0003446821641637523, | |
| "loss": 0.0394, | |
| "step": 152000 | |
| }, | |
| { | |
| "epoch": 15.60254567850544, | |
| "eval_cer": 0.010515402843601895, | |
| "eval_loss": 0.021091148257255554, | |
| "eval_runtime": 22.528, | |
| "eval_samples_per_second": 3.995, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.045576407506702415, | |
| "step": 152000 | |
| }, | |
| { | |
| "epoch": 15.705194005337713, | |
| "grad_norm": 0.20325519144535065, | |
| "learning_rate": 0.0003436535692244394, | |
| "loss": 0.0396, | |
| "step": 153000 | |
| }, | |
| { | |
| "epoch": 15.705194005337713, | |
| "eval_cer": 0.010811611374407584, | |
| "eval_loss": 0.023403970524668694, | |
| "eval_runtime": 22.2476, | |
| "eval_samples_per_second": 4.045, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.043789097408400354, | |
| "step": 153000 | |
| }, | |
| { | |
| "epoch": 15.807842332169985, | |
| "grad_norm": 0.15232166647911072, | |
| "learning_rate": 0.0003426249742851265, | |
| "loss": 0.0394, | |
| "step": 154000 | |
| }, | |
| { | |
| "epoch": 15.807842332169985, | |
| "eval_cer": 0.013625592417061612, | |
| "eval_loss": 0.022816922515630722, | |
| "eval_runtime": 22.7231, | |
| "eval_samples_per_second": 3.961, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 154000 | |
| }, | |
| { | |
| "epoch": 15.910490659002258, | |
| "grad_norm": 0.3241395056247711, | |
| "learning_rate": 0.0003415963793458136, | |
| "loss": 0.0392, | |
| "step": 155000 | |
| }, | |
| { | |
| "epoch": 15.910490659002258, | |
| "eval_cer": 0.010811611374407584, | |
| "eval_loss": 0.021291887387633324, | |
| "eval_runtime": 22.2235, | |
| "eval_samples_per_second": 4.05, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.044682752457551385, | |
| "step": 155000 | |
| }, | |
| { | |
| "epoch": 16.01313898583453, | |
| "grad_norm": 0.19988052546977997, | |
| "learning_rate": 0.0003405677844065007, | |
| "loss": 0.0393, | |
| "step": 156000 | |
| }, | |
| { | |
| "epoch": 16.01313898583453, | |
| "eval_cer": 0.011404028436018957, | |
| "eval_loss": 0.02226296253502369, | |
| "eval_runtime": 22.2614, | |
| "eval_samples_per_second": 4.043, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 156000 | |
| }, | |
| { | |
| "epoch": 16.115787312666804, | |
| "grad_norm": 0.23728616535663605, | |
| "learning_rate": 0.0003395391894671878, | |
| "loss": 0.0382, | |
| "step": 157000 | |
| }, | |
| { | |
| "epoch": 16.115787312666804, | |
| "eval_cer": 0.011255924170616114, | |
| "eval_loss": 0.021734587848186493, | |
| "eval_runtime": 22.1722, | |
| "eval_samples_per_second": 4.059, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 157000 | |
| }, | |
| { | |
| "epoch": 16.218435639499077, | |
| "grad_norm": 0.21486635506153107, | |
| "learning_rate": 0.00033851059452787494, | |
| "loss": 0.0379, | |
| "step": 158000 | |
| }, | |
| { | |
| "epoch": 16.218435639499077, | |
| "eval_cer": 0.009182464454976303, | |
| "eval_loss": 0.02133306674659252, | |
| "eval_runtime": 22.2489, | |
| "eval_samples_per_second": 4.045, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.040214477211796246, | |
| "step": 158000 | |
| }, | |
| { | |
| "epoch": 16.321083966331347, | |
| "grad_norm": 0.21918782591819763, | |
| "learning_rate": 0.00033748199958856203, | |
| "loss": 0.0382, | |
| "step": 159000 | |
| }, | |
| { | |
| "epoch": 16.321083966331347, | |
| "eval_cer": 0.009182464454976303, | |
| "eval_loss": 0.022134315222501755, | |
| "eval_runtime": 22.3537, | |
| "eval_samples_per_second": 4.026, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.040214477211796246, | |
| "step": 159000 | |
| }, | |
| { | |
| "epoch": 16.42373229316362, | |
| "grad_norm": 0.1927264928817749, | |
| "learning_rate": 0.0003364534046492491, | |
| "loss": 0.0383, | |
| "step": 160000 | |
| }, | |
| { | |
| "epoch": 16.42373229316362, | |
| "eval_cer": 0.0115521327014218, | |
| "eval_loss": 0.022146208211779594, | |
| "eval_runtime": 22.5984, | |
| "eval_samples_per_second": 3.983, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 160000 | |
| }, | |
| { | |
| "epoch": 16.526380619995894, | |
| "grad_norm": 0.19513466954231262, | |
| "learning_rate": 0.0003354248097099362, | |
| "loss": 0.0386, | |
| "step": 161000 | |
| }, | |
| { | |
| "epoch": 16.526380619995894, | |
| "eval_cer": 0.011700236966824644, | |
| "eval_loss": 0.021798642352223396, | |
| "eval_runtime": 22.2348, | |
| "eval_samples_per_second": 4.048, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 161000 | |
| }, | |
| { | |
| "epoch": 16.629028946828168, | |
| "grad_norm": 0.1991739124059677, | |
| "learning_rate": 0.0003343962147706233, | |
| "loss": 0.038, | |
| "step": 162000 | |
| }, | |
| { | |
| "epoch": 16.629028946828168, | |
| "eval_cer": 0.010515402843601895, | |
| "eval_loss": 0.021404601633548737, | |
| "eval_runtime": 22.5257, | |
| "eval_samples_per_second": 3.995, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.045576407506702415, | |
| "step": 162000 | |
| }, | |
| { | |
| "epoch": 16.731677273660438, | |
| "grad_norm": 0.19290116429328918, | |
| "learning_rate": 0.0003333676198313104, | |
| "loss": 0.0389, | |
| "step": 163000 | |
| }, | |
| { | |
| "epoch": 16.731677273660438, | |
| "eval_cer": 0.009922985781990521, | |
| "eval_loss": 0.0213669091463089, | |
| "eval_runtime": 22.5781, | |
| "eval_samples_per_second": 3.986, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.041108132260947276, | |
| "step": 163000 | |
| }, | |
| { | |
| "epoch": 16.83432560049271, | |
| "grad_norm": 0.29244861006736755, | |
| "learning_rate": 0.0003323390248919975, | |
| "loss": 0.0384, | |
| "step": 164000 | |
| }, | |
| { | |
| "epoch": 16.83432560049271, | |
| "eval_cer": 0.009478672985781991, | |
| "eval_loss": 0.02153705060482025, | |
| "eval_runtime": 22.335, | |
| "eval_samples_per_second": 4.03, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.040214477211796246, | |
| "step": 164000 | |
| }, | |
| { | |
| "epoch": 16.936973927324985, | |
| "grad_norm": 0.17148034274578094, | |
| "learning_rate": 0.00033131042995268465, | |
| "loss": 0.0381, | |
| "step": 165000 | |
| }, | |
| { | |
| "epoch": 16.936973927324985, | |
| "eval_cer": 0.010515402843601895, | |
| "eval_loss": 0.022320713847875595, | |
| "eval_runtime": 22.0009, | |
| "eval_samples_per_second": 4.091, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04289544235924933, | |
| "step": 165000 | |
| }, | |
| { | |
| "epoch": 17.03962225415726, | |
| "grad_norm": 0.31796592473983765, | |
| "learning_rate": 0.00033028183501337174, | |
| "loss": 0.0384, | |
| "step": 166000 | |
| }, | |
| { | |
| "epoch": 17.03962225415726, | |
| "eval_cer": 0.01110781990521327, | |
| "eval_loss": 0.02184494584798813, | |
| "eval_runtime": 22.2781, | |
| "eval_samples_per_second": 4.04, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.043789097408400354, | |
| "step": 166000 | |
| }, | |
| { | |
| "epoch": 17.14227058098953, | |
| "grad_norm": 0.2634246051311493, | |
| "learning_rate": 0.00032925324007405883, | |
| "loss": 0.0371, | |
| "step": 167000 | |
| }, | |
| { | |
| "epoch": 17.14227058098953, | |
| "eval_cer": 0.009922985781990521, | |
| "eval_loss": 0.022673843428492546, | |
| "eval_runtime": 22.5449, | |
| "eval_samples_per_second": 3.992, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.040214477211796246, | |
| "step": 167000 | |
| }, | |
| { | |
| "epoch": 17.244918907821802, | |
| "grad_norm": 0.21225817501544952, | |
| "learning_rate": 0.0003282246451347459, | |
| "loss": 0.0372, | |
| "step": 168000 | |
| }, | |
| { | |
| "epoch": 17.244918907821802, | |
| "eval_cer": 0.011848341232227487, | |
| "eval_loss": 0.021213963627815247, | |
| "eval_runtime": 22.3119, | |
| "eval_samples_per_second": 4.034, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 168000 | |
| }, | |
| { | |
| "epoch": 17.347567234654075, | |
| "grad_norm": 0.30099403858184814, | |
| "learning_rate": 0.000327196050195433, | |
| "loss": 0.0375, | |
| "step": 169000 | |
| }, | |
| { | |
| "epoch": 17.347567234654075, | |
| "eval_cer": 0.009478672985781991, | |
| "eval_loss": 0.021301671862602234, | |
| "eval_runtime": 22.2137, | |
| "eval_samples_per_second": 4.052, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03932082216264522, | |
| "step": 169000 | |
| }, | |
| { | |
| "epoch": 17.45021556148635, | |
| "grad_norm": 0.20359040796756744, | |
| "learning_rate": 0.0003261674552561201, | |
| "loss": 0.0378, | |
| "step": 170000 | |
| }, | |
| { | |
| "epoch": 17.45021556148635, | |
| "eval_cer": 0.010515402843601895, | |
| "eval_loss": 0.02227012813091278, | |
| "eval_runtime": 22.5554, | |
| "eval_samples_per_second": 3.99, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.04289544235924933, | |
| "step": 170000 | |
| }, | |
| { | |
| "epoch": 17.55286388831862, | |
| "grad_norm": 0.225717231631279, | |
| "learning_rate": 0.0003251388603168072, | |
| "loss": 0.0381, | |
| "step": 171000 | |
| }, | |
| { | |
| "epoch": 17.55286388831862, | |
| "eval_cer": 0.010515402843601895, | |
| "eval_loss": 0.02183985523879528, | |
| "eval_runtime": 22.3674, | |
| "eval_samples_per_second": 4.024, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04289544235924933, | |
| "step": 171000 | |
| }, | |
| { | |
| "epoch": 17.655512215150893, | |
| "grad_norm": 0.23642343282699585, | |
| "learning_rate": 0.00032411026537749435, | |
| "loss": 0.038, | |
| "step": 172000 | |
| }, | |
| { | |
| "epoch": 17.655512215150893, | |
| "eval_cer": 0.010071090047393365, | |
| "eval_loss": 0.02163875661790371, | |
| "eval_runtime": 22.3807, | |
| "eval_samples_per_second": 4.021, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0420017873100983, | |
| "step": 172000 | |
| }, | |
| { | |
| "epoch": 17.758160541983166, | |
| "grad_norm": 0.206275075674057, | |
| "learning_rate": 0.00032308167043818144, | |
| "loss": 0.0381, | |
| "step": 173000 | |
| }, | |
| { | |
| "epoch": 17.758160541983166, | |
| "eval_cer": 0.01110781990521327, | |
| "eval_loss": 0.021833743900060654, | |
| "eval_runtime": 22.1601, | |
| "eval_samples_per_second": 4.061, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.043789097408400354, | |
| "step": 173000 | |
| }, | |
| { | |
| "epoch": 17.86080886881544, | |
| "grad_norm": 0.1906212568283081, | |
| "learning_rate": 0.00032205307549886854, | |
| "loss": 0.0376, | |
| "step": 174000 | |
| }, | |
| { | |
| "epoch": 17.86080886881544, | |
| "eval_cer": 0.010219194312796208, | |
| "eval_loss": 0.0216918233782053, | |
| "eval_runtime": 22.5419, | |
| "eval_samples_per_second": 3.993, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.0420017873100983, | |
| "step": 174000 | |
| }, | |
| { | |
| "epoch": 17.96345719564771, | |
| "grad_norm": 0.2309373915195465, | |
| "learning_rate": 0.00032102448055955563, | |
| "loss": 0.0379, | |
| "step": 175000 | |
| }, | |
| { | |
| "epoch": 17.96345719564771, | |
| "eval_cer": 0.01229265402843602, | |
| "eval_loss": 0.022404534742236137, | |
| "eval_runtime": 22.4205, | |
| "eval_samples_per_second": 4.014, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 175000 | |
| }, | |
| { | |
| "epoch": 18.066105522479983, | |
| "grad_norm": 0.26210764050483704, | |
| "learning_rate": 0.0003199958856202427, | |
| "loss": 0.037, | |
| "step": 176000 | |
| }, | |
| { | |
| "epoch": 18.066105522479983, | |
| "eval_cer": 0.009626777251184835, | |
| "eval_loss": 0.02186032012104988, | |
| "eval_runtime": 22.3427, | |
| "eval_samples_per_second": 4.028, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0420017873100983, | |
| "step": 176000 | |
| }, | |
| { | |
| "epoch": 18.168753849312257, | |
| "grad_norm": 0.18146245181560516, | |
| "learning_rate": 0.0003189672906809298, | |
| "loss": 0.0366, | |
| "step": 177000 | |
| }, | |
| { | |
| "epoch": 18.168753849312257, | |
| "eval_cer": 0.009774881516587678, | |
| "eval_loss": 0.021789953112602234, | |
| "eval_runtime": 22.2303, | |
| "eval_samples_per_second": 4.049, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03842716711349419, | |
| "step": 177000 | |
| }, | |
| { | |
| "epoch": 18.27140217614453, | |
| "grad_norm": 0.21234826743602753, | |
| "learning_rate": 0.0003179386957416169, | |
| "loss": 0.0364, | |
| "step": 178000 | |
| }, | |
| { | |
| "epoch": 18.27140217614453, | |
| "eval_cer": 0.010071090047393365, | |
| "eval_loss": 0.023085610941052437, | |
| "eval_runtime": 22.5005, | |
| "eval_samples_per_second": 4.0, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.0420017873100983, | |
| "step": 178000 | |
| }, | |
| { | |
| "epoch": 18.3740505029768, | |
| "grad_norm": 0.20181190967559814, | |
| "learning_rate": 0.00031691010080230406, | |
| "loss": 0.037, | |
| "step": 179000 | |
| }, | |
| { | |
| "epoch": 18.3740505029768, | |
| "eval_cer": 0.009478672985781991, | |
| "eval_loss": 0.022074325010180473, | |
| "eval_runtime": 22.2124, | |
| "eval_samples_per_second": 4.052, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0420017873100983, | |
| "step": 179000 | |
| }, | |
| { | |
| "epoch": 18.476698829809074, | |
| "grad_norm": 0.20344142615795135, | |
| "learning_rate": 0.00031588150586299115, | |
| "loss": 0.0368, | |
| "step": 180000 | |
| }, | |
| { | |
| "epoch": 18.476698829809074, | |
| "eval_cer": 0.010219194312796208, | |
| "eval_loss": 0.02133142203092575, | |
| "eval_runtime": 22.4401, | |
| "eval_samples_per_second": 4.011, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.041108132260947276, | |
| "step": 180000 | |
| }, | |
| { | |
| "epoch": 18.579347156641347, | |
| "grad_norm": 0.26208797097206116, | |
| "learning_rate": 0.00031485291092367824, | |
| "loss": 0.0374, | |
| "step": 181000 | |
| }, | |
| { | |
| "epoch": 18.579347156641347, | |
| "eval_cer": 0.010811611374407584, | |
| "eval_loss": 0.022065425291657448, | |
| "eval_runtime": 22.2276, | |
| "eval_samples_per_second": 4.049, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.043789097408400354, | |
| "step": 181000 | |
| }, | |
| { | |
| "epoch": 18.68199548347362, | |
| "grad_norm": 0.21672701835632324, | |
| "learning_rate": 0.00031382431598436534, | |
| "loss": 0.0378, | |
| "step": 182000 | |
| }, | |
| { | |
| "epoch": 18.68199548347362, | |
| "eval_cer": 0.010367298578199052, | |
| "eval_loss": 0.022190110757946968, | |
| "eval_runtime": 22.4805, | |
| "eval_samples_per_second": 4.003, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.044682752457551385, | |
| "step": 182000 | |
| }, | |
| { | |
| "epoch": 18.78464381030589, | |
| "grad_norm": 0.26024818420410156, | |
| "learning_rate": 0.00031279572104505243, | |
| "loss": 0.0373, | |
| "step": 183000 | |
| }, | |
| { | |
| "epoch": 18.78464381030589, | |
| "eval_cer": 0.009330568720379146, | |
| "eval_loss": 0.0211643036454916, | |
| "eval_runtime": 22.2696, | |
| "eval_samples_per_second": 4.041, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.041108132260947276, | |
| "step": 183000 | |
| }, | |
| { | |
| "epoch": 18.887292137138164, | |
| "grad_norm": 0.15596991777420044, | |
| "learning_rate": 0.0003117671261057395, | |
| "loss": 0.0374, | |
| "step": 184000 | |
| }, | |
| { | |
| "epoch": 18.887292137138164, | |
| "eval_cer": 0.010367298578199052, | |
| "eval_loss": 0.02078518457710743, | |
| "eval_runtime": 22.3696, | |
| "eval_samples_per_second": 4.023, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0420017873100983, | |
| "step": 184000 | |
| }, | |
| { | |
| "epoch": 18.989940463970438, | |
| "grad_norm": 0.21325981616973877, | |
| "learning_rate": 0.0003107385311664266, | |
| "loss": 0.0377, | |
| "step": 185000 | |
| }, | |
| { | |
| "epoch": 18.989940463970438, | |
| "eval_cer": 0.011700236966824644, | |
| "eval_loss": 0.021274788305163383, | |
| "eval_runtime": 22.3681, | |
| "eval_samples_per_second": 4.024, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.045576407506702415, | |
| "step": 185000 | |
| }, | |
| { | |
| "epoch": 19.09258879080271, | |
| "grad_norm": 0.2881476581096649, | |
| "learning_rate": 0.00030970993622711376, | |
| "loss": 0.0362, | |
| "step": 186000 | |
| }, | |
| { | |
| "epoch": 19.09258879080271, | |
| "eval_cer": 0.009774881516587678, | |
| "eval_loss": 0.020506886765360832, | |
| "eval_runtime": 22.4918, | |
| "eval_samples_per_second": 4.001, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.040214477211796246, | |
| "step": 186000 | |
| }, | |
| { | |
| "epoch": 19.19523711763498, | |
| "grad_norm": 0.2128625214099884, | |
| "learning_rate": 0.00030868134128780086, | |
| "loss": 0.036, | |
| "step": 187000 | |
| }, | |
| { | |
| "epoch": 19.19523711763498, | |
| "eval_cer": 0.0115521327014218, | |
| "eval_loss": 0.021428626030683517, | |
| "eval_runtime": 22.4254, | |
| "eval_samples_per_second": 4.013, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 187000 | |
| }, | |
| { | |
| "epoch": 19.297885444467255, | |
| "grad_norm": 0.20976155996322632, | |
| "learning_rate": 0.000307652746348488, | |
| "loss": 0.0367, | |
| "step": 188000 | |
| }, | |
| { | |
| "epoch": 19.297885444467255, | |
| "eval_cer": 0.010959715639810427, | |
| "eval_loss": 0.020786074921488762, | |
| "eval_runtime": 22.3038, | |
| "eval_samples_per_second": 4.035, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.045576407506702415, | |
| "step": 188000 | |
| }, | |
| { | |
| "epoch": 19.40053377129953, | |
| "grad_norm": 0.21684007346630096, | |
| "learning_rate": 0.0003066241514091751, | |
| "loss": 0.0363, | |
| "step": 189000 | |
| }, | |
| { | |
| "epoch": 19.40053377129953, | |
| "eval_cer": 0.010367298578199052, | |
| "eval_loss": 0.020662952214479446, | |
| "eval_runtime": 22.499, | |
| "eval_samples_per_second": 4.0, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.041108132260947276, | |
| "step": 189000 | |
| }, | |
| { | |
| "epoch": 19.503182098131802, | |
| "grad_norm": 0.4317739009857178, | |
| "learning_rate": 0.0003055955564698622, | |
| "loss": 0.0365, | |
| "step": 190000 | |
| }, | |
| { | |
| "epoch": 19.503182098131802, | |
| "eval_cer": 0.009922985781990521, | |
| "eval_loss": 0.021092107519507408, | |
| "eval_runtime": 22.4144, | |
| "eval_samples_per_second": 4.015, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04289544235924933, | |
| "step": 190000 | |
| }, | |
| { | |
| "epoch": 19.605830424964072, | |
| "grad_norm": 0.23220385611057281, | |
| "learning_rate": 0.0003045669615305493, | |
| "loss": 0.0364, | |
| "step": 191000 | |
| }, | |
| { | |
| "epoch": 19.605830424964072, | |
| "eval_cer": 0.009922985781990521, | |
| "eval_loss": 0.020856238901615143, | |
| "eval_runtime": 22.4251, | |
| "eval_samples_per_second": 4.013, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.040214477211796246, | |
| "step": 191000 | |
| }, | |
| { | |
| "epoch": 19.708478751796346, | |
| "grad_norm": 0.20682792365550995, | |
| "learning_rate": 0.0003035383665912364, | |
| "loss": 0.0369, | |
| "step": 192000 | |
| }, | |
| { | |
| "epoch": 19.708478751796346, | |
| "eval_cer": 0.01066350710900474, | |
| "eval_loss": 0.020980246365070343, | |
| "eval_runtime": 22.3208, | |
| "eval_samples_per_second": 4.032, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.044682752457551385, | |
| "step": 192000 | |
| }, | |
| { | |
| "epoch": 19.81112707862862, | |
| "grad_norm": 0.160901740193367, | |
| "learning_rate": 0.0003025097716519235, | |
| "loss": 0.0366, | |
| "step": 193000 | |
| }, | |
| { | |
| "epoch": 19.81112707862862, | |
| "eval_cer": 0.009626777251184835, | |
| "eval_loss": 0.02076118066906929, | |
| "eval_runtime": 22.2895, | |
| "eval_samples_per_second": 4.038, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.040214477211796246, | |
| "step": 193000 | |
| }, | |
| { | |
| "epoch": 19.913775405460893, | |
| "grad_norm": 0.25278541445732117, | |
| "learning_rate": 0.0003014811767126106, | |
| "loss": 0.0369, | |
| "step": 194000 | |
| }, | |
| { | |
| "epoch": 19.913775405460893, | |
| "eval_cer": 0.009774881516587678, | |
| "eval_loss": 0.02101094461977482, | |
| "eval_runtime": 22.1128, | |
| "eval_samples_per_second": 4.07, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03932082216264522, | |
| "step": 194000 | |
| }, | |
| { | |
| "epoch": 20.016423732293163, | |
| "grad_norm": 0.18868520855903625, | |
| "learning_rate": 0.0003004525817732977, | |
| "loss": 0.0365, | |
| "step": 195000 | |
| }, | |
| { | |
| "epoch": 20.016423732293163, | |
| "eval_cer": 0.010071090047393365, | |
| "eval_loss": 0.021312745288014412, | |
| "eval_runtime": 22.2522, | |
| "eval_samples_per_second": 4.045, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.041108132260947276, | |
| "step": 195000 | |
| }, | |
| { | |
| "epoch": 20.119072059125436, | |
| "grad_norm": 0.16919797658920288, | |
| "learning_rate": 0.0002994239868339848, | |
| "loss": 0.0358, | |
| "step": 196000 | |
| }, | |
| { | |
| "epoch": 20.119072059125436, | |
| "eval_cer": 0.009626777251184835, | |
| "eval_loss": 0.020815536379814148, | |
| "eval_runtime": 22.1691, | |
| "eval_samples_per_second": 4.06, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03842716711349419, | |
| "step": 196000 | |
| }, | |
| { | |
| "epoch": 20.22172038595771, | |
| "grad_norm": 0.2387053221464157, | |
| "learning_rate": 0.0002983953918946719, | |
| "loss": 0.0357, | |
| "step": 197000 | |
| }, | |
| { | |
| "epoch": 20.22172038595771, | |
| "eval_cer": 0.009922985781990521, | |
| "eval_loss": 0.022103123366832733, | |
| "eval_runtime": 22.0878, | |
| "eval_samples_per_second": 4.075, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.041108132260947276, | |
| "step": 197000 | |
| }, | |
| { | |
| "epoch": 20.324368712789983, | |
| "grad_norm": 0.1803978830575943, | |
| "learning_rate": 0.000297366796955359, | |
| "loss": 0.036, | |
| "step": 198000 | |
| }, | |
| { | |
| "epoch": 20.324368712789983, | |
| "eval_cer": 0.010219194312796208, | |
| "eval_loss": 0.02122490108013153, | |
| "eval_runtime": 22.4345, | |
| "eval_samples_per_second": 4.012, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.040214477211796246, | |
| "step": 198000 | |
| }, | |
| { | |
| "epoch": 20.427017039622253, | |
| "grad_norm": 0.30070099234580994, | |
| "learning_rate": 0.0002963382020160461, | |
| "loss": 0.0358, | |
| "step": 199000 | |
| }, | |
| { | |
| "epoch": 20.427017039622253, | |
| "eval_cer": 0.010367298578199052, | |
| "eval_loss": 0.020686373114585876, | |
| "eval_runtime": 22.3744, | |
| "eval_samples_per_second": 4.022, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.041108132260947276, | |
| "step": 199000 | |
| }, | |
| { | |
| "epoch": 20.529665366454527, | |
| "grad_norm": 0.20007756352424622, | |
| "learning_rate": 0.00029530960707673323, | |
| "loss": 0.0361, | |
| "step": 200000 | |
| }, | |
| { | |
| "epoch": 20.529665366454527, | |
| "eval_cer": 0.008886255924170616, | |
| "eval_loss": 0.019611194729804993, | |
| "eval_runtime": 22.3448, | |
| "eval_samples_per_second": 4.028, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03485254691689008, | |
| "step": 200000 | |
| }, | |
| { | |
| "epoch": 20.6323136932868, | |
| "grad_norm": 0.2129463404417038, | |
| "learning_rate": 0.0002942810121374203, | |
| "loss": 0.0363, | |
| "step": 201000 | |
| }, | |
| { | |
| "epoch": 20.6323136932868, | |
| "eval_cer": 0.009182464454976303, | |
| "eval_loss": 0.021008532494306564, | |
| "eval_runtime": 22.292, | |
| "eval_samples_per_second": 4.037, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03842716711349419, | |
| "step": 201000 | |
| }, | |
| { | |
| "epoch": 20.73496202011907, | |
| "grad_norm": 0.16587744653224945, | |
| "learning_rate": 0.0002932524171981074, | |
| "loss": 0.0359, | |
| "step": 202000 | |
| }, | |
| { | |
| "epoch": 20.73496202011907, | |
| "eval_cer": 0.010367298578199052, | |
| "eval_loss": 0.02170945331454277, | |
| "eval_runtime": 22.2982, | |
| "eval_samples_per_second": 4.036, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03932082216264522, | |
| "step": 202000 | |
| }, | |
| { | |
| "epoch": 20.837610346951344, | |
| "grad_norm": 0.1517147719860077, | |
| "learning_rate": 0.0002922238222587945, | |
| "loss": 0.0362, | |
| "step": 203000 | |
| }, | |
| { | |
| "epoch": 20.837610346951344, | |
| "eval_cer": 0.01110781990521327, | |
| "eval_loss": 0.021431386470794678, | |
| "eval_runtime": 22.317, | |
| "eval_samples_per_second": 4.033, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.044682752457551385, | |
| "step": 203000 | |
| }, | |
| { | |
| "epoch": 20.940258673783617, | |
| "grad_norm": 0.33937105536460876, | |
| "learning_rate": 0.0002911952273194816, | |
| "loss": 0.0359, | |
| "step": 204000 | |
| }, | |
| { | |
| "epoch": 20.940258673783617, | |
| "eval_cer": 0.009330568720379146, | |
| "eval_loss": 0.01971680298447609, | |
| "eval_runtime": 22.523, | |
| "eval_samples_per_second": 3.996, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03842716711349419, | |
| "step": 204000 | |
| }, | |
| { | |
| "epoch": 21.04290700061589, | |
| "grad_norm": 0.19971835613250732, | |
| "learning_rate": 0.0002901666323801687, | |
| "loss": 0.0357, | |
| "step": 205000 | |
| }, | |
| { | |
| "epoch": 21.04290700061589, | |
| "eval_cer": 0.009774881516587678, | |
| "eval_loss": 0.020647110417485237, | |
| "eval_runtime": 22.1889, | |
| "eval_samples_per_second": 4.056, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0420017873100983, | |
| "step": 205000 | |
| }, | |
| { | |
| "epoch": 21.14555532744816, | |
| "grad_norm": 0.2110970914363861, | |
| "learning_rate": 0.0002891380374408558, | |
| "loss": 0.0346, | |
| "step": 206000 | |
| }, | |
| { | |
| "epoch": 21.14555532744816, | |
| "eval_cer": 0.010219194312796208, | |
| "eval_loss": 0.01989123784005642, | |
| "eval_runtime": 22.7934, | |
| "eval_samples_per_second": 3.949, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03932082216264522, | |
| "step": 206000 | |
| }, | |
| { | |
| "epoch": 21.248203654280434, | |
| "grad_norm": 0.20562046766281128, | |
| "learning_rate": 0.00028810944250154294, | |
| "loss": 0.0349, | |
| "step": 207000 | |
| }, | |
| { | |
| "epoch": 21.248203654280434, | |
| "eval_cer": 0.010515402843601895, | |
| "eval_loss": 0.020657481625676155, | |
| "eval_runtime": 22.7975, | |
| "eval_samples_per_second": 3.948, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.04289544235924933, | |
| "step": 207000 | |
| }, | |
| { | |
| "epoch": 21.350851981112708, | |
| "grad_norm": 0.16318105161190033, | |
| "learning_rate": 0.00028708084756223003, | |
| "loss": 0.0353, | |
| "step": 208000 | |
| }, | |
| { | |
| "epoch": 21.350851981112708, | |
| "eval_cer": 0.010367298578199052, | |
| "eval_loss": 0.01860020123422146, | |
| "eval_runtime": 22.3138, | |
| "eval_samples_per_second": 4.033, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04289544235924933, | |
| "step": 208000 | |
| }, | |
| { | |
| "epoch": 21.45350030794498, | |
| "grad_norm": 0.2521456480026245, | |
| "learning_rate": 0.0002860522526229171, | |
| "loss": 0.0356, | |
| "step": 209000 | |
| }, | |
| { | |
| "epoch": 21.45350030794498, | |
| "eval_cer": 0.009478672985781991, | |
| "eval_loss": 0.01932937651872635, | |
| "eval_runtime": 22.5041, | |
| "eval_samples_per_second": 3.999, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03932082216264522, | |
| "step": 209000 | |
| }, | |
| { | |
| "epoch": 21.55614863477725, | |
| "grad_norm": 0.15729331970214844, | |
| "learning_rate": 0.0002850236576836042, | |
| "loss": 0.0356, | |
| "step": 210000 | |
| }, | |
| { | |
| "epoch": 21.55614863477725, | |
| "eval_cer": 0.01066350710900474, | |
| "eval_loss": 0.01956385001540184, | |
| "eval_runtime": 22.4886, | |
| "eval_samples_per_second": 4.002, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.044682752457551385, | |
| "step": 210000 | |
| }, | |
| { | |
| "epoch": 21.658796961609525, | |
| "grad_norm": 0.3368454575538635, | |
| "learning_rate": 0.0002839950627442913, | |
| "loss": 0.0355, | |
| "step": 211000 | |
| }, | |
| { | |
| "epoch": 21.658796961609525, | |
| "eval_cer": 0.009330568720379146, | |
| "eval_loss": 0.017849326133728027, | |
| "eval_runtime": 22.2354, | |
| "eval_samples_per_second": 4.048, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03753351206434316, | |
| "step": 211000 | |
| }, | |
| { | |
| "epoch": 21.7614452884418, | |
| "grad_norm": 0.19548599421977997, | |
| "learning_rate": 0.0002829664678049784, | |
| "loss": 0.0356, | |
| "step": 212000 | |
| }, | |
| { | |
| "epoch": 21.7614452884418, | |
| "eval_cer": 0.010071090047393365, | |
| "eval_loss": 0.018993813544511795, | |
| "eval_runtime": 22.2405, | |
| "eval_samples_per_second": 4.047, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.040214477211796246, | |
| "step": 212000 | |
| }, | |
| { | |
| "epoch": 21.864093615274072, | |
| "grad_norm": 0.300447553396225, | |
| "learning_rate": 0.0002819378728656655, | |
| "loss": 0.0355, | |
| "step": 213000 | |
| }, | |
| { | |
| "epoch": 21.864093615274072, | |
| "eval_cer": 0.009478672985781991, | |
| "eval_loss": 0.018936272710561752, | |
| "eval_runtime": 22.3288, | |
| "eval_samples_per_second": 4.031, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.041108132260947276, | |
| "step": 213000 | |
| }, | |
| { | |
| "epoch": 21.966741942106342, | |
| "grad_norm": 0.2299223691225052, | |
| "learning_rate": 0.00028090927792635264, | |
| "loss": 0.0357, | |
| "step": 214000 | |
| }, | |
| { | |
| "epoch": 21.966741942106342, | |
| "eval_cer": 0.008738151658767773, | |
| "eval_loss": 0.017861895263195038, | |
| "eval_runtime": 22.4153, | |
| "eval_samples_per_second": 4.015, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 214000 | |
| }, | |
| { | |
| "epoch": 22.069390268938616, | |
| "grad_norm": 0.22763746976852417, | |
| "learning_rate": 0.00027988068298703974, | |
| "loss": 0.035, | |
| "step": 215000 | |
| }, | |
| { | |
| "epoch": 22.069390268938616, | |
| "eval_cer": 0.010219194312796208, | |
| "eval_loss": 0.018531307578086853, | |
| "eval_runtime": 22.5994, | |
| "eval_samples_per_second": 3.982, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.04289544235924933, | |
| "step": 215000 | |
| }, | |
| { | |
| "epoch": 22.17203859577089, | |
| "grad_norm": 0.17821846902370453, | |
| "learning_rate": 0.00027885208804772683, | |
| "loss": 0.0343, | |
| "step": 216000 | |
| }, | |
| { | |
| "epoch": 22.17203859577089, | |
| "eval_cer": 0.010219194312796208, | |
| "eval_loss": 0.01801004819571972, | |
| "eval_runtime": 22.6174, | |
| "eval_samples_per_second": 3.979, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.040214477211796246, | |
| "step": 216000 | |
| }, | |
| { | |
| "epoch": 22.274686922603163, | |
| "grad_norm": 0.20817448198795319, | |
| "learning_rate": 0.0002778234931084139, | |
| "loss": 0.0353, | |
| "step": 217000 | |
| }, | |
| { | |
| "epoch": 22.274686922603163, | |
| "eval_cer": 0.009330568720379146, | |
| "eval_loss": 0.01922653615474701, | |
| "eval_runtime": 22.3133, | |
| "eval_samples_per_second": 4.033, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0420017873100983, | |
| "step": 217000 | |
| }, | |
| { | |
| "epoch": 22.377335249435433, | |
| "grad_norm": 0.23358240723609924, | |
| "learning_rate": 0.000276794898169101, | |
| "loss": 0.0347, | |
| "step": 218000 | |
| }, | |
| { | |
| "epoch": 22.377335249435433, | |
| "eval_cer": 0.009774881516587678, | |
| "eval_loss": 0.019061286002397537, | |
| "eval_runtime": 22.516, | |
| "eval_samples_per_second": 3.997, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.041108132260947276, | |
| "step": 218000 | |
| }, | |
| { | |
| "epoch": 22.479983576267706, | |
| "grad_norm": 0.1586250215768814, | |
| "learning_rate": 0.0002757663032297881, | |
| "loss": 0.0344, | |
| "step": 219000 | |
| }, | |
| { | |
| "epoch": 22.479983576267706, | |
| "eval_cer": 0.008738151658767773, | |
| "eval_loss": 0.018482740968465805, | |
| "eval_runtime": 22.7441, | |
| "eval_samples_per_second": 3.957, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 219000 | |
| }, | |
| { | |
| "epoch": 22.58263190309998, | |
| "grad_norm": 0.18222519755363464, | |
| "learning_rate": 0.0002747377082904752, | |
| "loss": 0.0355, | |
| "step": 220000 | |
| }, | |
| { | |
| "epoch": 22.58263190309998, | |
| "eval_cer": 0.010071090047393365, | |
| "eval_loss": 0.01949753239750862, | |
| "eval_runtime": 22.5921, | |
| "eval_samples_per_second": 3.984, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.0420017873100983, | |
| "step": 220000 | |
| }, | |
| { | |
| "epoch": 22.685280229932253, | |
| "grad_norm": 0.24508023262023926, | |
| "learning_rate": 0.00027370911335116235, | |
| "loss": 0.0351, | |
| "step": 221000 | |
| }, | |
| { | |
| "epoch": 22.685280229932253, | |
| "eval_cer": 0.009330568720379146, | |
| "eval_loss": 0.019236860796809196, | |
| "eval_runtime": 22.1523, | |
| "eval_samples_per_second": 4.063, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.040214477211796246, | |
| "step": 221000 | |
| }, | |
| { | |
| "epoch": 22.787928556764523, | |
| "grad_norm": 0.20518304407596588, | |
| "learning_rate": 0.00027268051841184944, | |
| "loss": 0.0349, | |
| "step": 222000 | |
| }, | |
| { | |
| "epoch": 22.787928556764523, | |
| "eval_cer": 0.009182464454976303, | |
| "eval_loss": 0.01924031414091587, | |
| "eval_runtime": 22.4441, | |
| "eval_samples_per_second": 4.01, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.040214477211796246, | |
| "step": 222000 | |
| }, | |
| { | |
| "epoch": 22.890576883596797, | |
| "grad_norm": 0.34271785616874695, | |
| "learning_rate": 0.00027165192347253654, | |
| "loss": 0.0354, | |
| "step": 223000 | |
| }, | |
| { | |
| "epoch": 22.890576883596797, | |
| "eval_cer": 0.008738151658767773, | |
| "eval_loss": 0.018572239205241203, | |
| "eval_runtime": 22.2344, | |
| "eval_samples_per_second": 4.048, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03753351206434316, | |
| "step": 223000 | |
| }, | |
| { | |
| "epoch": 22.99322521042907, | |
| "grad_norm": 0.23716846108436584, | |
| "learning_rate": 0.00027062332853322363, | |
| "loss": 0.0351, | |
| "step": 224000 | |
| }, | |
| { | |
| "epoch": 22.99322521042907, | |
| "eval_cer": 0.008738151658767773, | |
| "eval_loss": 0.018431993201375008, | |
| "eval_runtime": 22.1175, | |
| "eval_samples_per_second": 4.069, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03753351206434316, | |
| "step": 224000 | |
| }, | |
| { | |
| "epoch": 23.095873537261344, | |
| "grad_norm": 0.20996840298175812, | |
| "learning_rate": 0.0002695947335939107, | |
| "loss": 0.0336, | |
| "step": 225000 | |
| }, | |
| { | |
| "epoch": 23.095873537261344, | |
| "eval_cer": 0.008886255924170616, | |
| "eval_loss": 0.01876632496714592, | |
| "eval_runtime": 22.1146, | |
| "eval_samples_per_second": 4.07, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03753351206434316, | |
| "step": 225000 | |
| }, | |
| { | |
| "epoch": 23.198521864093614, | |
| "grad_norm": 0.24316902458667755, | |
| "learning_rate": 0.0002685661386545978, | |
| "loss": 0.0342, | |
| "step": 226000 | |
| }, | |
| { | |
| "epoch": 23.198521864093614, | |
| "eval_cer": 0.009774881516587678, | |
| "eval_loss": 0.018810244277119637, | |
| "eval_runtime": 22.534, | |
| "eval_samples_per_second": 3.994, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.040214477211796246, | |
| "step": 226000 | |
| }, | |
| { | |
| "epoch": 23.301170190925887, | |
| "grad_norm": 0.21960946917533875, | |
| "learning_rate": 0.00026753754371528496, | |
| "loss": 0.0341, | |
| "step": 227000 | |
| }, | |
| { | |
| "epoch": 23.301170190925887, | |
| "eval_cer": 0.009330568720379146, | |
| "eval_loss": 0.01906524784862995, | |
| "eval_runtime": 22.5098, | |
| "eval_samples_per_second": 3.998, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03932082216264522, | |
| "step": 227000 | |
| }, | |
| { | |
| "epoch": 23.40381851775816, | |
| "grad_norm": 0.24521832168102264, | |
| "learning_rate": 0.00026650894877597206, | |
| "loss": 0.0346, | |
| "step": 228000 | |
| }, | |
| { | |
| "epoch": 23.40381851775816, | |
| "eval_cer": 0.008293838862559242, | |
| "eval_loss": 0.017732921987771988, | |
| "eval_runtime": 22.3346, | |
| "eval_samples_per_second": 4.03, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 228000 | |
| }, | |
| { | |
| "epoch": 23.506466844590435, | |
| "grad_norm": 0.20621488988399506, | |
| "learning_rate": 0.00026548035383665915, | |
| "loss": 0.0341, | |
| "step": 229000 | |
| }, | |
| { | |
| "epoch": 23.506466844590435, | |
| "eval_cer": 0.009330568720379146, | |
| "eval_loss": 0.018322736024856567, | |
| "eval_runtime": 22.307, | |
| "eval_samples_per_second": 4.035, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03842716711349419, | |
| "step": 229000 | |
| }, | |
| { | |
| "epoch": 23.609115171422705, | |
| "grad_norm": 0.22831113636493683, | |
| "learning_rate": 0.00026445175889734624, | |
| "loss": 0.0348, | |
| "step": 230000 | |
| }, | |
| { | |
| "epoch": 23.609115171422705, | |
| "eval_cer": 0.008886255924170616, | |
| "eval_loss": 0.018174562603235245, | |
| "eval_runtime": 22.2887, | |
| "eval_samples_per_second": 4.038, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03842716711349419, | |
| "step": 230000 | |
| }, | |
| { | |
| "epoch": 23.711763498254978, | |
| "grad_norm": 0.26975417137145996, | |
| "learning_rate": 0.00026342316395803334, | |
| "loss": 0.0343, | |
| "step": 231000 | |
| }, | |
| { | |
| "epoch": 23.711763498254978, | |
| "eval_cer": 0.008293838862559242, | |
| "eval_loss": 0.017916005104780197, | |
| "eval_runtime": 22.3158, | |
| "eval_samples_per_second": 4.033, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.035746201966041107, | |
| "step": 231000 | |
| }, | |
| { | |
| "epoch": 23.81441182508725, | |
| "grad_norm": 0.18927611410617828, | |
| "learning_rate": 0.00026239456901872043, | |
| "loss": 0.0347, | |
| "step": 232000 | |
| }, | |
| { | |
| "epoch": 23.81441182508725, | |
| "eval_cer": 0.009182464454976303, | |
| "eval_loss": 0.018648013472557068, | |
| "eval_runtime": 22.3793, | |
| "eval_samples_per_second": 4.022, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.040214477211796246, | |
| "step": 232000 | |
| }, | |
| { | |
| "epoch": 23.917060151919525, | |
| "grad_norm": 0.3485555350780487, | |
| "learning_rate": 0.0002613659740794075, | |
| "loss": 0.0349, | |
| "step": 233000 | |
| }, | |
| { | |
| "epoch": 23.917060151919525, | |
| "eval_cer": 0.00903436018957346, | |
| "eval_loss": 0.018734946846961975, | |
| "eval_runtime": 22.4017, | |
| "eval_samples_per_second": 4.018, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03842716711349419, | |
| "step": 233000 | |
| }, | |
| { | |
| "epoch": 24.019708478751795, | |
| "grad_norm": 0.19541509449481964, | |
| "learning_rate": 0.00026033737914009467, | |
| "loss": 0.0346, | |
| "step": 234000 | |
| }, | |
| { | |
| "epoch": 24.019708478751795, | |
| "eval_cer": 0.008441943127962086, | |
| "eval_loss": 0.01827438361942768, | |
| "eval_runtime": 22.4976, | |
| "eval_samples_per_second": 4.0, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 234000 | |
| }, | |
| { | |
| "epoch": 24.12235680558407, | |
| "grad_norm": 0.15688838064670563, | |
| "learning_rate": 0.00025930878420078176, | |
| "loss": 0.0335, | |
| "step": 235000 | |
| }, | |
| { | |
| "epoch": 24.12235680558407, | |
| "eval_cer": 0.00903436018957346, | |
| "eval_loss": 0.01861654222011566, | |
| "eval_runtime": 22.4918, | |
| "eval_samples_per_second": 4.001, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03932082216264522, | |
| "step": 235000 | |
| }, | |
| { | |
| "epoch": 24.225005132416342, | |
| "grad_norm": 0.15766377747058868, | |
| "learning_rate": 0.00025828018926146886, | |
| "loss": 0.0334, | |
| "step": 236000 | |
| }, | |
| { | |
| "epoch": 24.225005132416342, | |
| "eval_cer": 0.009478672985781991, | |
| "eval_loss": 0.017844926565885544, | |
| "eval_runtime": 22.542, | |
| "eval_samples_per_second": 3.993, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.040214477211796246, | |
| "step": 236000 | |
| }, | |
| { | |
| "epoch": 24.327653459248616, | |
| "grad_norm": 0.27199745178222656, | |
| "learning_rate": 0.00025725159432215595, | |
| "loss": 0.0335, | |
| "step": 237000 | |
| }, | |
| { | |
| "epoch": 24.327653459248616, | |
| "eval_cer": 0.010219194312796208, | |
| "eval_loss": 0.018539218232035637, | |
| "eval_runtime": 22.1807, | |
| "eval_samples_per_second": 4.058, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.043789097408400354, | |
| "step": 237000 | |
| }, | |
| { | |
| "epoch": 24.430301786080886, | |
| "grad_norm": 0.2251148819923401, | |
| "learning_rate": 0.00025622299938284304, | |
| "loss": 0.0339, | |
| "step": 238000 | |
| }, | |
| { | |
| "epoch": 24.430301786080886, | |
| "eval_cer": 0.009478672985781991, | |
| "eval_loss": 0.018467124551534653, | |
| "eval_runtime": 22.5004, | |
| "eval_samples_per_second": 4.0, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03753351206434316, | |
| "step": 238000 | |
| }, | |
| { | |
| "epoch": 24.53295011291316, | |
| "grad_norm": 0.1856725960969925, | |
| "learning_rate": 0.00025519440444353014, | |
| "loss": 0.0343, | |
| "step": 239000 | |
| }, | |
| { | |
| "epoch": 24.53295011291316, | |
| "eval_cer": 0.009774881516587678, | |
| "eval_loss": 0.01805044710636139, | |
| "eval_runtime": 22.2875, | |
| "eval_samples_per_second": 4.038, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.040214477211796246, | |
| "step": 239000 | |
| }, | |
| { | |
| "epoch": 24.635598439745433, | |
| "grad_norm": 0.23337046802043915, | |
| "learning_rate": 0.00025416580950421723, | |
| "loss": 0.0341, | |
| "step": 240000 | |
| }, | |
| { | |
| "epoch": 24.635598439745433, | |
| "eval_cer": 0.009182464454976303, | |
| "eval_loss": 0.018100356683135033, | |
| "eval_runtime": 22.3689, | |
| "eval_samples_per_second": 4.023, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03932082216264522, | |
| "step": 240000 | |
| }, | |
| { | |
| "epoch": 24.738246766577706, | |
| "grad_norm": 0.3009665012359619, | |
| "learning_rate": 0.0002531372145649044, | |
| "loss": 0.0342, | |
| "step": 241000 | |
| }, | |
| { | |
| "epoch": 24.738246766577706, | |
| "eval_cer": 0.008590047393364929, | |
| "eval_loss": 0.017376452684402466, | |
| "eval_runtime": 22.3829, | |
| "eval_samples_per_second": 4.021, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 241000 | |
| }, | |
| { | |
| "epoch": 24.840895093409976, | |
| "grad_norm": 0.20083321630954742, | |
| "learning_rate": 0.00025210861962559147, | |
| "loss": 0.0341, | |
| "step": 242000 | |
| }, | |
| { | |
| "epoch": 24.840895093409976, | |
| "eval_cer": 0.00903436018957346, | |
| "eval_loss": 0.018101993948221207, | |
| "eval_runtime": 22.6389, | |
| "eval_samples_per_second": 3.975, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03842716711349419, | |
| "step": 242000 | |
| }, | |
| { | |
| "epoch": 24.94354342024225, | |
| "grad_norm": 0.2611863315105438, | |
| "learning_rate": 0.00025108002468627856, | |
| "loss": 0.034, | |
| "step": 243000 | |
| }, | |
| { | |
| "epoch": 24.94354342024225, | |
| "eval_cer": 0.008590047393364929, | |
| "eval_loss": 0.01786319725215435, | |
| "eval_runtime": 22.5343, | |
| "eval_samples_per_second": 3.994, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03842716711349419, | |
| "step": 243000 | |
| }, | |
| { | |
| "epoch": 25.046191747074523, | |
| "grad_norm": 0.22002221643924713, | |
| "learning_rate": 0.00025005142974696566, | |
| "loss": 0.0337, | |
| "step": 244000 | |
| }, | |
| { | |
| "epoch": 25.046191747074523, | |
| "eval_cer": 0.00903436018957346, | |
| "eval_loss": 0.01813172735273838, | |
| "eval_runtime": 22.7449, | |
| "eval_samples_per_second": 3.957, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03932082216264522, | |
| "step": 244000 | |
| }, | |
| { | |
| "epoch": 25.148840073906797, | |
| "grad_norm": 0.22292836010456085, | |
| "learning_rate": 0.00024902283480765275, | |
| "loss": 0.0331, | |
| "step": 245000 | |
| }, | |
| { | |
| "epoch": 25.148840073906797, | |
| "eval_cer": 0.010071090047393365, | |
| "eval_loss": 0.018383100628852844, | |
| "eval_runtime": 22.3583, | |
| "eval_samples_per_second": 4.025, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.041108132260947276, | |
| "step": 245000 | |
| }, | |
| { | |
| "epoch": 25.251488400739067, | |
| "grad_norm": 0.22255383431911469, | |
| "learning_rate": 0.00024799423986833984, | |
| "loss": 0.0332, | |
| "step": 246000 | |
| }, | |
| { | |
| "epoch": 25.251488400739067, | |
| "eval_cer": 0.009922985781990521, | |
| "eval_loss": 0.018330469727516174, | |
| "eval_runtime": 22.5685, | |
| "eval_samples_per_second": 3.988, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.041108132260947276, | |
| "step": 246000 | |
| }, | |
| { | |
| "epoch": 25.35413672757134, | |
| "grad_norm": 0.16578556597232819, | |
| "learning_rate": 0.00024696564492902694, | |
| "loss": 0.0335, | |
| "step": 247000 | |
| }, | |
| { | |
| "epoch": 25.35413672757134, | |
| "eval_cer": 0.008441943127962086, | |
| "eval_loss": 0.017710883170366287, | |
| "eval_runtime": 22.2874, | |
| "eval_samples_per_second": 4.038, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.035746201966041107, | |
| "step": 247000 | |
| }, | |
| { | |
| "epoch": 25.456785054403614, | |
| "grad_norm": 0.24286945164203644, | |
| "learning_rate": 0.0002459370499897141, | |
| "loss": 0.0336, | |
| "step": 248000 | |
| }, | |
| { | |
| "epoch": 25.456785054403614, | |
| "eval_cer": 0.00784952606635071, | |
| "eval_loss": 0.017551274970173836, | |
| "eval_runtime": 22.5028, | |
| "eval_samples_per_second": 4.0, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 248000 | |
| }, | |
| { | |
| "epoch": 25.559433381235884, | |
| "grad_norm": 0.2211551070213318, | |
| "learning_rate": 0.0002449084550504012, | |
| "loss": 0.0334, | |
| "step": 249000 | |
| }, | |
| { | |
| "epoch": 25.559433381235884, | |
| "eval_cer": 0.00784952606635071, | |
| "eval_loss": 0.017883356660604477, | |
| "eval_runtime": 22.5913, | |
| "eval_samples_per_second": 3.984, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.032171581769437, | |
| "step": 249000 | |
| }, | |
| { | |
| "epoch": 25.662081708068158, | |
| "grad_norm": 0.22908490896224976, | |
| "learning_rate": 0.00024387986011108827, | |
| "loss": 0.0336, | |
| "step": 250000 | |
| }, | |
| { | |
| "epoch": 25.662081708068158, | |
| "eval_cer": 0.00903436018957346, | |
| "eval_loss": 0.01878712698817253, | |
| "eval_runtime": 22.5067, | |
| "eval_samples_per_second": 3.999, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03753351206434316, | |
| "step": 250000 | |
| }, | |
| { | |
| "epoch": 25.76473003490043, | |
| "grad_norm": 0.23008745908737183, | |
| "learning_rate": 0.00024285126517177536, | |
| "loss": 0.0337, | |
| "step": 251000 | |
| }, | |
| { | |
| "epoch": 25.76473003490043, | |
| "eval_cer": 0.009478672985781991, | |
| "eval_loss": 0.018499523401260376, | |
| "eval_runtime": 22.5329, | |
| "eval_samples_per_second": 3.994, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03753351206434316, | |
| "step": 251000 | |
| }, | |
| { | |
| "epoch": 25.867378361732705, | |
| "grad_norm": 0.197307750582695, | |
| "learning_rate": 0.00024182267023246246, | |
| "loss": 0.0337, | |
| "step": 252000 | |
| }, | |
| { | |
| "epoch": 25.867378361732705, | |
| "eval_cer": 0.008293838862559242, | |
| "eval_loss": 0.01780012436211109, | |
| "eval_runtime": 22.5224, | |
| "eval_samples_per_second": 3.996, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03485254691689008, | |
| "step": 252000 | |
| }, | |
| { | |
| "epoch": 25.970026688564978, | |
| "grad_norm": 0.24368754029273987, | |
| "learning_rate": 0.00024079407529314955, | |
| "loss": 0.0338, | |
| "step": 253000 | |
| }, | |
| { | |
| "epoch": 25.970026688564978, | |
| "eval_cer": 0.009626777251184835, | |
| "eval_loss": 0.017333028838038445, | |
| "eval_runtime": 22.4863, | |
| "eval_samples_per_second": 4.002, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03842716711349419, | |
| "step": 253000 | |
| }, | |
| { | |
| "epoch": 26.072675015397248, | |
| "grad_norm": 0.17760606110095978, | |
| "learning_rate": 0.00023976548035383667, | |
| "loss": 0.0328, | |
| "step": 254000 | |
| }, | |
| { | |
| "epoch": 26.072675015397248, | |
| "eval_cer": 0.008886255924170616, | |
| "eval_loss": 0.017533306032419205, | |
| "eval_runtime": 22.6165, | |
| "eval_samples_per_second": 3.979, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.035746201966041107, | |
| "step": 254000 | |
| }, | |
| { | |
| "epoch": 26.17532334222952, | |
| "grad_norm": 0.24285702407360077, | |
| "learning_rate": 0.00023873688541452376, | |
| "loss": 0.0325, | |
| "step": 255000 | |
| }, | |
| { | |
| "epoch": 26.17532334222952, | |
| "eval_cer": 0.009182464454976303, | |
| "eval_loss": 0.01821918785572052, | |
| "eval_runtime": 22.4982, | |
| "eval_samples_per_second": 4.0, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03842716711349419, | |
| "step": 255000 | |
| }, | |
| { | |
| "epoch": 26.277971669061795, | |
| "grad_norm": 0.216440811753273, | |
| "learning_rate": 0.00023770829047521086, | |
| "loss": 0.0328, | |
| "step": 256000 | |
| }, | |
| { | |
| "epoch": 26.277971669061795, | |
| "eval_cer": 0.00903436018957346, | |
| "eval_loss": 0.01876477338373661, | |
| "eval_runtime": 22.6643, | |
| "eval_samples_per_second": 3.971, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.035746201966041107, | |
| "step": 256000 | |
| }, | |
| { | |
| "epoch": 26.380619995894065, | |
| "grad_norm": 0.18202808499336243, | |
| "learning_rate": 0.00023667969553589798, | |
| "loss": 0.0327, | |
| "step": 257000 | |
| }, | |
| { | |
| "epoch": 26.380619995894065, | |
| "eval_cer": 0.008590047393364929, | |
| "eval_loss": 0.017915373668074608, | |
| "eval_runtime": 22.5776, | |
| "eval_samples_per_second": 3.986, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 257000 | |
| }, | |
| { | |
| "epoch": 26.48326832272634, | |
| "grad_norm": 0.183961883187294, | |
| "learning_rate": 0.00023565110059658507, | |
| "loss": 0.0331, | |
| "step": 258000 | |
| }, | |
| { | |
| "epoch": 26.48326832272634, | |
| "eval_cer": 0.009182464454976303, | |
| "eval_loss": 0.018149225041270256, | |
| "eval_runtime": 22.4894, | |
| "eval_samples_per_second": 4.002, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03753351206434316, | |
| "step": 258000 | |
| }, | |
| { | |
| "epoch": 26.585916649558612, | |
| "grad_norm": 0.2554219365119934, | |
| "learning_rate": 0.00023462250565727216, | |
| "loss": 0.033, | |
| "step": 259000 | |
| }, | |
| { | |
| "epoch": 26.585916649558612, | |
| "eval_cer": 0.008590047393364929, | |
| "eval_loss": 0.017295319586992264, | |
| "eval_runtime": 22.4744, | |
| "eval_samples_per_second": 4.005, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03753351206434316, | |
| "step": 259000 | |
| }, | |
| { | |
| "epoch": 26.688564976390886, | |
| "grad_norm": 0.3923017978668213, | |
| "learning_rate": 0.00023359391071795926, | |
| "loss": 0.0334, | |
| "step": 260000 | |
| }, | |
| { | |
| "epoch": 26.688564976390886, | |
| "eval_cer": 0.008441943127962086, | |
| "eval_loss": 0.018423665314912796, | |
| "eval_runtime": 22.6112, | |
| "eval_samples_per_second": 3.98, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03753351206434316, | |
| "step": 260000 | |
| }, | |
| { | |
| "epoch": 26.791213303223156, | |
| "grad_norm": 0.413510262966156, | |
| "learning_rate": 0.00023256531577864638, | |
| "loss": 0.0334, | |
| "step": 261000 | |
| }, | |
| { | |
| "epoch": 26.791213303223156, | |
| "eval_cer": 0.009330568720379146, | |
| "eval_loss": 0.017927566543221474, | |
| "eval_runtime": 22.5258, | |
| "eval_samples_per_second": 3.995, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.040214477211796246, | |
| "step": 261000 | |
| }, | |
| { | |
| "epoch": 26.89386163005543, | |
| "grad_norm": 0.25007760524749756, | |
| "learning_rate": 0.00023153672083933347, | |
| "loss": 0.0333, | |
| "step": 262000 | |
| }, | |
| { | |
| "epoch": 26.89386163005543, | |
| "eval_cer": 0.008738151658767773, | |
| "eval_loss": 0.017284687608480453, | |
| "eval_runtime": 22.5825, | |
| "eval_samples_per_second": 3.985, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03842716711349419, | |
| "step": 262000 | |
| }, | |
| { | |
| "epoch": 26.996509956887703, | |
| "grad_norm": 0.3856132924556732, | |
| "learning_rate": 0.00023050812590002056, | |
| "loss": 0.0336, | |
| "step": 263000 | |
| }, | |
| { | |
| "epoch": 26.996509956887703, | |
| "eval_cer": 0.007997630331753554, | |
| "eval_loss": 0.017093736678361893, | |
| "eval_runtime": 22.7688, | |
| "eval_samples_per_second": 3.953, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 263000 | |
| }, | |
| { | |
| "epoch": 27.099158283719976, | |
| "grad_norm": 0.3299943208694458, | |
| "learning_rate": 0.00022947953096070768, | |
| "loss": 0.0321, | |
| "step": 264000 | |
| }, | |
| { | |
| "epoch": 27.099158283719976, | |
| "eval_cer": 0.008886255924170616, | |
| "eval_loss": 0.017240121960639954, | |
| "eval_runtime": 22.5916, | |
| "eval_samples_per_second": 3.984, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03842716711349419, | |
| "step": 264000 | |
| }, | |
| { | |
| "epoch": 27.201806610552246, | |
| "grad_norm": 0.1730368584394455, | |
| "learning_rate": 0.00022845093602139478, | |
| "loss": 0.0327, | |
| "step": 265000 | |
| }, | |
| { | |
| "epoch": 27.201806610552246, | |
| "eval_cer": 0.008738151658767773, | |
| "eval_loss": 0.01783335767686367, | |
| "eval_runtime": 22.2378, | |
| "eval_samples_per_second": 4.047, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 265000 | |
| }, | |
| { | |
| "epoch": 27.30445493738452, | |
| "grad_norm": 0.3776761293411255, | |
| "learning_rate": 0.00022742234108208187, | |
| "loss": 0.0326, | |
| "step": 266000 | |
| }, | |
| { | |
| "epoch": 27.30445493738452, | |
| "eval_cer": 0.009182464454976303, | |
| "eval_loss": 0.017694596201181412, | |
| "eval_runtime": 22.5643, | |
| "eval_samples_per_second": 3.989, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.040214477211796246, | |
| "step": 266000 | |
| }, | |
| { | |
| "epoch": 27.407103264216794, | |
| "grad_norm": 0.17738159000873566, | |
| "learning_rate": 0.00022639374614276896, | |
| "loss": 0.0324, | |
| "step": 267000 | |
| }, | |
| { | |
| "epoch": 27.407103264216794, | |
| "eval_cer": 0.008590047393364929, | |
| "eval_loss": 0.017476912587881088, | |
| "eval_runtime": 22.7303, | |
| "eval_samples_per_second": 3.959, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 267000 | |
| }, | |
| { | |
| "epoch": 27.509751591049067, | |
| "grad_norm": 0.2550307512283325, | |
| "learning_rate": 0.00022536515120345608, | |
| "loss": 0.0325, | |
| "step": 268000 | |
| }, | |
| { | |
| "epoch": 27.509751591049067, | |
| "eval_cer": 0.008886255924170616, | |
| "eval_loss": 0.01743141934275627, | |
| "eval_runtime": 22.7564, | |
| "eval_samples_per_second": 3.955, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 268000 | |
| }, | |
| { | |
| "epoch": 27.612399917881337, | |
| "grad_norm": 0.16253231465816498, | |
| "learning_rate": 0.00022433655626414317, | |
| "loss": 0.0327, | |
| "step": 269000 | |
| }, | |
| { | |
| "epoch": 27.612399917881337, | |
| "eval_cer": 0.008590047393364929, | |
| "eval_loss": 0.017301015555858612, | |
| "eval_runtime": 23.1263, | |
| "eval_samples_per_second": 3.892, | |
| "eval_steps_per_second": 0.043, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 269000 | |
| }, | |
| { | |
| "epoch": 27.71504824471361, | |
| "grad_norm": 0.15195374190807343, | |
| "learning_rate": 0.00022330796132483027, | |
| "loss": 0.0322, | |
| "step": 270000 | |
| }, | |
| { | |
| "epoch": 27.71504824471361, | |
| "eval_cer": 0.007997630331753554, | |
| "eval_loss": 0.017138667404651642, | |
| "eval_runtime": 23.3844, | |
| "eval_samples_per_second": 3.849, | |
| "eval_steps_per_second": 0.043, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 270000 | |
| }, | |
| { | |
| "epoch": 27.817696571545884, | |
| "grad_norm": 0.2660813331604004, | |
| "learning_rate": 0.0002222793663855174, | |
| "loss": 0.0332, | |
| "step": 271000 | |
| }, | |
| { | |
| "epoch": 27.817696571545884, | |
| "eval_cer": 0.008441943127962086, | |
| "eval_loss": 0.016922015696763992, | |
| "eval_runtime": 22.6918, | |
| "eval_samples_per_second": 3.966, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03485254691689008, | |
| "step": 271000 | |
| }, | |
| { | |
| "epoch": 27.920344898378158, | |
| "grad_norm": 0.20602907240390778, | |
| "learning_rate": 0.00022125077144620448, | |
| "loss": 0.0328, | |
| "step": 272000 | |
| }, | |
| { | |
| "epoch": 27.920344898378158, | |
| "eval_cer": 0.008590047393364929, | |
| "eval_loss": 0.016544727608561516, | |
| "eval_runtime": 22.5504, | |
| "eval_samples_per_second": 3.991, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.035746201966041107, | |
| "step": 272000 | |
| }, | |
| { | |
| "epoch": 28.022993225210428, | |
| "grad_norm": 0.22045257687568665, | |
| "learning_rate": 0.00022022217650689157, | |
| "loss": 0.0324, | |
| "step": 273000 | |
| }, | |
| { | |
| "epoch": 28.022993225210428, | |
| "eval_cer": 0.007997630331753554, | |
| "eval_loss": 0.016568990424275398, | |
| "eval_runtime": 22.6458, | |
| "eval_samples_per_second": 3.974, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 273000 | |
| }, | |
| { | |
| "epoch": 28.1256415520427, | |
| "grad_norm": 0.18806225061416626, | |
| "learning_rate": 0.0002191935815675787, | |
| "loss": 0.0315, | |
| "step": 274000 | |
| }, | |
| { | |
| "epoch": 28.1256415520427, | |
| "eval_cer": 0.008145734597156399, | |
| "eval_loss": 0.016486881300807, | |
| "eval_runtime": 22.6614, | |
| "eval_samples_per_second": 3.972, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03485254691689008, | |
| "step": 274000 | |
| }, | |
| { | |
| "epoch": 28.228289878874975, | |
| "grad_norm": 0.1898849457502365, | |
| "learning_rate": 0.0002181649866282658, | |
| "loss": 0.0319, | |
| "step": 275000 | |
| }, | |
| { | |
| "epoch": 28.228289878874975, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.016169264912605286, | |
| "eval_runtime": 22.5647, | |
| "eval_samples_per_second": 3.989, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 275000 | |
| }, | |
| { | |
| "epoch": 28.33093820570725, | |
| "grad_norm": 0.27770882844924927, | |
| "learning_rate": 0.00021713639168895288, | |
| "loss": 0.0323, | |
| "step": 276000 | |
| }, | |
| { | |
| "epoch": 28.33093820570725, | |
| "eval_cer": 0.008293838862559242, | |
| "eval_loss": 0.01689663529396057, | |
| "eval_runtime": 22.5984, | |
| "eval_samples_per_second": 3.983, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03485254691689008, | |
| "step": 276000 | |
| }, | |
| { | |
| "epoch": 28.43358653253952, | |
| "grad_norm": 0.23435185849666595, | |
| "learning_rate": 0.00021610779674963997, | |
| "loss": 0.0322, | |
| "step": 277000 | |
| }, | |
| { | |
| "epoch": 28.43358653253952, | |
| "eval_cer": 0.009922985781990521, | |
| "eval_loss": 0.016887083649635315, | |
| "eval_runtime": 23.1898, | |
| "eval_samples_per_second": 3.881, | |
| "eval_steps_per_second": 0.043, | |
| "eval_wer": 0.03932082216264522, | |
| "step": 277000 | |
| }, | |
| { | |
| "epoch": 28.536234859371792, | |
| "grad_norm": 0.19802525639533997, | |
| "learning_rate": 0.0002150792018103271, | |
| "loss": 0.0319, | |
| "step": 278000 | |
| }, | |
| { | |
| "epoch": 28.536234859371792, | |
| "eval_cer": 0.00903436018957346, | |
| "eval_loss": 0.016642894595861435, | |
| "eval_runtime": 22.9038, | |
| "eval_samples_per_second": 3.929, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 278000 | |
| }, | |
| { | |
| "epoch": 28.638883186204065, | |
| "grad_norm": 0.33312317728996277, | |
| "learning_rate": 0.0002140506068710142, | |
| "loss": 0.0324, | |
| "step": 279000 | |
| }, | |
| { | |
| "epoch": 28.638883186204065, | |
| "eval_cer": 0.009774881516587678, | |
| "eval_loss": 0.016943588852882385, | |
| "eval_runtime": 22.6215, | |
| "eval_samples_per_second": 3.979, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03753351206434316, | |
| "step": 279000 | |
| }, | |
| { | |
| "epoch": 28.74153151303634, | |
| "grad_norm": 0.19455067813396454, | |
| "learning_rate": 0.00021302201193170128, | |
| "loss": 0.0326, | |
| "step": 280000 | |
| }, | |
| { | |
| "epoch": 28.74153151303634, | |
| "eval_cer": 0.008738151658767773, | |
| "eval_loss": 0.016804693266749382, | |
| "eval_runtime": 22.602, | |
| "eval_samples_per_second": 3.982, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03753351206434316, | |
| "step": 280000 | |
| }, | |
| { | |
| "epoch": 28.84417983986861, | |
| "grad_norm": 0.14672687649726868, | |
| "learning_rate": 0.0002119934169923884, | |
| "loss": 0.0328, | |
| "step": 281000 | |
| }, | |
| { | |
| "epoch": 28.84417983986861, | |
| "eval_cer": 0.008441943127962086, | |
| "eval_loss": 0.016962364315986633, | |
| "eval_runtime": 22.456, | |
| "eval_samples_per_second": 4.008, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 281000 | |
| }, | |
| { | |
| "epoch": 28.946828166700882, | |
| "grad_norm": 0.3485030233860016, | |
| "learning_rate": 0.00021096482205307552, | |
| "loss": 0.0322, | |
| "step": 282000 | |
| }, | |
| { | |
| "epoch": 28.946828166700882, | |
| "eval_cer": 0.009330568720379146, | |
| "eval_loss": 0.017243940383195877, | |
| "eval_runtime": 22.6149, | |
| "eval_samples_per_second": 3.98, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03842716711349419, | |
| "step": 282000 | |
| }, | |
| { | |
| "epoch": 29.049476493533156, | |
| "grad_norm": 0.20129866898059845, | |
| "learning_rate": 0.00020993622711376261, | |
| "loss": 0.0317, | |
| "step": 283000 | |
| }, | |
| { | |
| "epoch": 29.049476493533156, | |
| "eval_cer": 0.009330568720379146, | |
| "eval_loss": 0.01694132201373577, | |
| "eval_runtime": 22.5131, | |
| "eval_samples_per_second": 3.998, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03753351206434316, | |
| "step": 283000 | |
| }, | |
| { | |
| "epoch": 29.15212482036543, | |
| "grad_norm": 0.32001006603240967, | |
| "learning_rate": 0.0002089076321744497, | |
| "loss": 0.0315, | |
| "step": 284000 | |
| }, | |
| { | |
| "epoch": 29.15212482036543, | |
| "eval_cer": 0.007553317535545024, | |
| "eval_loss": 0.01657693088054657, | |
| "eval_runtime": 22.5678, | |
| "eval_samples_per_second": 3.988, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 284000 | |
| }, | |
| { | |
| "epoch": 29.2547731471977, | |
| "grad_norm": 0.21219150722026825, | |
| "learning_rate": 0.00020787903723513683, | |
| "loss": 0.0315, | |
| "step": 285000 | |
| }, | |
| { | |
| "epoch": 29.2547731471977, | |
| "eval_cer": 0.007997630331753554, | |
| "eval_loss": 0.016005953773856163, | |
| "eval_runtime": 22.3068, | |
| "eval_samples_per_second": 4.035, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03485254691689008, | |
| "step": 285000 | |
| }, | |
| { | |
| "epoch": 29.357421474029973, | |
| "grad_norm": 0.20887607336044312, | |
| "learning_rate": 0.00020685044229582392, | |
| "loss": 0.0319, | |
| "step": 286000 | |
| }, | |
| { | |
| "epoch": 29.357421474029973, | |
| "eval_cer": 0.008886255924170616, | |
| "eval_loss": 0.016784099861979485, | |
| "eval_runtime": 22.3044, | |
| "eval_samples_per_second": 4.035, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03485254691689008, | |
| "step": 286000 | |
| }, | |
| { | |
| "epoch": 29.460069800862247, | |
| "grad_norm": 0.28705254197120667, | |
| "learning_rate": 0.00020582184735651101, | |
| "loss": 0.0314, | |
| "step": 287000 | |
| }, | |
| { | |
| "epoch": 29.460069800862247, | |
| "eval_cer": 0.008441943127962086, | |
| "eval_loss": 0.016876235604286194, | |
| "eval_runtime": 22.2793, | |
| "eval_samples_per_second": 4.04, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 287000 | |
| }, | |
| { | |
| "epoch": 29.56271812769452, | |
| "grad_norm": 0.323476642370224, | |
| "learning_rate": 0.00020479325241719813, | |
| "loss": 0.032, | |
| "step": 288000 | |
| }, | |
| { | |
| "epoch": 29.56271812769452, | |
| "eval_cer": 0.007701421800947867, | |
| "eval_loss": 0.016845477744936943, | |
| "eval_runtime": 22.1977, | |
| "eval_samples_per_second": 4.054, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.032171581769437, | |
| "step": 288000 | |
| }, | |
| { | |
| "epoch": 29.66536645452679, | |
| "grad_norm": 0.1829458326101303, | |
| "learning_rate": 0.00020376465747788523, | |
| "loss": 0.0321, | |
| "step": 289000 | |
| }, | |
| { | |
| "epoch": 29.66536645452679, | |
| "eval_cer": 0.007553317535545024, | |
| "eval_loss": 0.015846768394112587, | |
| "eval_runtime": 22.1026, | |
| "eval_samples_per_second": 4.072, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 289000 | |
| }, | |
| { | |
| "epoch": 29.768014781359064, | |
| "grad_norm": 0.25482961535453796, | |
| "learning_rate": 0.00020273606253857232, | |
| "loss": 0.032, | |
| "step": 290000 | |
| }, | |
| { | |
| "epoch": 29.768014781359064, | |
| "eval_cer": 0.008441943127962086, | |
| "eval_loss": 0.015906278043985367, | |
| "eval_runtime": 22.3867, | |
| "eval_samples_per_second": 4.02, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 290000 | |
| }, | |
| { | |
| "epoch": 29.870663108191337, | |
| "grad_norm": 0.19813385605812073, | |
| "learning_rate": 0.00020170746759925941, | |
| "loss": 0.0319, | |
| "step": 291000 | |
| }, | |
| { | |
| "epoch": 29.870663108191337, | |
| "eval_cer": 0.009330568720379146, | |
| "eval_loss": 0.016582278534770012, | |
| "eval_runtime": 22.6644, | |
| "eval_samples_per_second": 3.971, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03932082216264522, | |
| "step": 291000 | |
| }, | |
| { | |
| "epoch": 29.97331143502361, | |
| "grad_norm": 0.23543916642665863, | |
| "learning_rate": 0.00020067887265994653, | |
| "loss": 0.0321, | |
| "step": 292000 | |
| }, | |
| { | |
| "epoch": 29.97331143502361, | |
| "eval_cer": 0.009774881516587678, | |
| "eval_loss": 0.01627834513783455, | |
| "eval_runtime": 22.7565, | |
| "eval_samples_per_second": 3.955, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03932082216264522, | |
| "step": 292000 | |
| }, | |
| { | |
| "epoch": 30.07595976185588, | |
| "grad_norm": 0.20127102732658386, | |
| "learning_rate": 0.00019965027772063363, | |
| "loss": 0.0309, | |
| "step": 293000 | |
| }, | |
| { | |
| "epoch": 30.07595976185588, | |
| "eval_cer": 0.008738151658767773, | |
| "eval_loss": 0.016053717583417892, | |
| "eval_runtime": 22.4512, | |
| "eval_samples_per_second": 4.009, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03842716711349419, | |
| "step": 293000 | |
| }, | |
| { | |
| "epoch": 30.178608088688154, | |
| "grad_norm": 0.19356395304203033, | |
| "learning_rate": 0.00019862168278132072, | |
| "loss": 0.031, | |
| "step": 294000 | |
| }, | |
| { | |
| "epoch": 30.178608088688154, | |
| "eval_cer": 0.008886255924170616, | |
| "eval_loss": 0.01641259714961052, | |
| "eval_runtime": 22.3968, | |
| "eval_samples_per_second": 4.018, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03753351206434316, | |
| "step": 294000 | |
| }, | |
| { | |
| "epoch": 30.281256415520428, | |
| "grad_norm": 0.3252677619457245, | |
| "learning_rate": 0.00019759308784200784, | |
| "loss": 0.0314, | |
| "step": 295000 | |
| }, | |
| { | |
| "epoch": 30.281256415520428, | |
| "eval_cer": 0.008293838862559242, | |
| "eval_loss": 0.015991076827049255, | |
| "eval_runtime": 22.4292, | |
| "eval_samples_per_second": 4.013, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.035746201966041107, | |
| "step": 295000 | |
| }, | |
| { | |
| "epoch": 30.3839047423527, | |
| "grad_norm": 0.1807209700345993, | |
| "learning_rate": 0.00019656449290269493, | |
| "loss": 0.0312, | |
| "step": 296000 | |
| }, | |
| { | |
| "epoch": 30.3839047423527, | |
| "eval_cer": 0.008145734597156399, | |
| "eval_loss": 0.015960650518536568, | |
| "eval_runtime": 22.2631, | |
| "eval_samples_per_second": 4.043, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.035746201966041107, | |
| "step": 296000 | |
| }, | |
| { | |
| "epoch": 30.48655306918497, | |
| "grad_norm": 0.2719903588294983, | |
| "learning_rate": 0.00019553589796338203, | |
| "loss": 0.0312, | |
| "step": 297000 | |
| }, | |
| { | |
| "epoch": 30.48655306918497, | |
| "eval_cer": 0.00903436018957346, | |
| "eval_loss": 0.015721548348665237, | |
| "eval_runtime": 22.1868, | |
| "eval_samples_per_second": 4.056, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03842716711349419, | |
| "step": 297000 | |
| }, | |
| { | |
| "epoch": 30.589201396017245, | |
| "grad_norm": 0.32360509037971497, | |
| "learning_rate": 0.00019450730302406912, | |
| "loss": 0.0314, | |
| "step": 298000 | |
| }, | |
| { | |
| "epoch": 30.589201396017245, | |
| "eval_cer": 0.008293838862559242, | |
| "eval_loss": 0.01612556166946888, | |
| "eval_runtime": 22.345, | |
| "eval_samples_per_second": 4.028, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 298000 | |
| }, | |
| { | |
| "epoch": 30.69184972284952, | |
| "grad_norm": 0.28737571835517883, | |
| "learning_rate": 0.00019347870808475624, | |
| "loss": 0.0317, | |
| "step": 299000 | |
| }, | |
| { | |
| "epoch": 30.69184972284952, | |
| "eval_cer": 0.00784952606635071, | |
| "eval_loss": 0.016010984778404236, | |
| "eval_runtime": 22.4982, | |
| "eval_samples_per_second": 4.0, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 299000 | |
| }, | |
| { | |
| "epoch": 30.794498049681792, | |
| "grad_norm": 0.16976721584796906, | |
| "learning_rate": 0.00019245011314544333, | |
| "loss": 0.0315, | |
| "step": 300000 | |
| }, | |
| { | |
| "epoch": 30.794498049681792, | |
| "eval_cer": 0.008590047393364929, | |
| "eval_loss": 0.01699880138039589, | |
| "eval_runtime": 22.4468, | |
| "eval_samples_per_second": 4.009, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.035746201966041107, | |
| "step": 300000 | |
| }, | |
| { | |
| "epoch": 30.897146376514062, | |
| "grad_norm": 0.2326597273349762, | |
| "learning_rate": 0.00019142151820613043, | |
| "loss": 0.0315, | |
| "step": 301000 | |
| }, | |
| { | |
| "epoch": 30.897146376514062, | |
| "eval_cer": 0.009330568720379146, | |
| "eval_loss": 0.016296546906232834, | |
| "eval_runtime": 22.4046, | |
| "eval_samples_per_second": 4.017, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03842716711349419, | |
| "step": 301000 | |
| }, | |
| { | |
| "epoch": 30.999794703346335, | |
| "grad_norm": 0.4060353934764862, | |
| "learning_rate": 0.00019039292326681755, | |
| "loss": 0.0318, | |
| "step": 302000 | |
| }, | |
| { | |
| "epoch": 30.999794703346335, | |
| "eval_cer": 0.008886255924170616, | |
| "eval_loss": 0.01701035536825657, | |
| "eval_runtime": 22.4393, | |
| "eval_samples_per_second": 4.011, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 302000 | |
| }, | |
| { | |
| "epoch": 31.10244303017861, | |
| "grad_norm": 0.19074688851833344, | |
| "learning_rate": 0.00018936432832750464, | |
| "loss": 0.0304, | |
| "step": 303000 | |
| }, | |
| { | |
| "epoch": 31.10244303017861, | |
| "eval_cer": 0.008886255924170616, | |
| "eval_loss": 0.016249870881438255, | |
| "eval_runtime": 22.6009, | |
| "eval_samples_per_second": 3.982, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.035746201966041107, | |
| "step": 303000 | |
| }, | |
| { | |
| "epoch": 31.20509135701088, | |
| "grad_norm": 0.1725562959909439, | |
| "learning_rate": 0.00018833573338819173, | |
| "loss": 0.0309, | |
| "step": 304000 | |
| }, | |
| { | |
| "epoch": 31.20509135701088, | |
| "eval_cer": 0.007997630331753554, | |
| "eval_loss": 0.01692904904484749, | |
| "eval_runtime": 22.3452, | |
| "eval_samples_per_second": 4.028, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 304000 | |
| }, | |
| { | |
| "epoch": 31.307739683843153, | |
| "grad_norm": 0.18745183944702148, | |
| "learning_rate": 0.00018730713844887883, | |
| "loss": 0.0308, | |
| "step": 305000 | |
| }, | |
| { | |
| "epoch": 31.307739683843153, | |
| "eval_cer": 0.008145734597156399, | |
| "eval_loss": 0.01693115197122097, | |
| "eval_runtime": 22.313, | |
| "eval_samples_per_second": 4.034, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.035746201966041107, | |
| "step": 305000 | |
| }, | |
| { | |
| "epoch": 31.410388010675426, | |
| "grad_norm": 0.2364443838596344, | |
| "learning_rate": 0.00018627854350956595, | |
| "loss": 0.0311, | |
| "step": 306000 | |
| }, | |
| { | |
| "epoch": 31.410388010675426, | |
| "eval_cer": 0.009626777251184835, | |
| "eval_loss": 0.016600091010332108, | |
| "eval_runtime": 22.2551, | |
| "eval_samples_per_second": 4.044, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03932082216264522, | |
| "step": 306000 | |
| }, | |
| { | |
| "epoch": 31.5130363375077, | |
| "grad_norm": 0.19486719369888306, | |
| "learning_rate": 0.00018524994857025304, | |
| "loss": 0.0309, | |
| "step": 307000 | |
| }, | |
| { | |
| "epoch": 31.5130363375077, | |
| "eval_cer": 0.008738151658767773, | |
| "eval_loss": 0.01619219221174717, | |
| "eval_runtime": 22.3592, | |
| "eval_samples_per_second": 4.025, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 307000 | |
| }, | |
| { | |
| "epoch": 31.61568466433997, | |
| "grad_norm": 0.20491057634353638, | |
| "learning_rate": 0.00018422135363094013, | |
| "loss": 0.0308, | |
| "step": 308000 | |
| }, | |
| { | |
| "epoch": 31.61568466433997, | |
| "eval_cer": 0.008590047393364929, | |
| "eval_loss": 0.016473235562443733, | |
| "eval_runtime": 22.4059, | |
| "eval_samples_per_second": 4.017, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 308000 | |
| }, | |
| { | |
| "epoch": 31.718332991172243, | |
| "grad_norm": 0.26360154151916504, | |
| "learning_rate": 0.00018319275869162725, | |
| "loss": 0.0312, | |
| "step": 309000 | |
| }, | |
| { | |
| "epoch": 31.718332991172243, | |
| "eval_cer": 0.008145734597156399, | |
| "eval_loss": 0.01562103908509016, | |
| "eval_runtime": 22.1604, | |
| "eval_samples_per_second": 4.061, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03485254691689008, | |
| "step": 309000 | |
| }, | |
| { | |
| "epoch": 31.820981318004517, | |
| "grad_norm": 0.15921452641487122, | |
| "learning_rate": 0.00018216416375231435, | |
| "loss": 0.0316, | |
| "step": 310000 | |
| }, | |
| { | |
| "epoch": 31.820981318004517, | |
| "eval_cer": 0.008738151658767773, | |
| "eval_loss": 0.016504855826497078, | |
| "eval_runtime": 22.5365, | |
| "eval_samples_per_second": 3.994, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 310000 | |
| }, | |
| { | |
| "epoch": 31.92362964483679, | |
| "grad_norm": 0.27201271057128906, | |
| "learning_rate": 0.00018113556881300144, | |
| "loss": 0.0314, | |
| "step": 311000 | |
| }, | |
| { | |
| "epoch": 31.92362964483679, | |
| "eval_cer": 0.008590047393364929, | |
| "eval_loss": 0.015464858151972294, | |
| "eval_runtime": 22.5378, | |
| "eval_samples_per_second": 3.993, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 311000 | |
| }, | |
| { | |
| "epoch": 32.02627797166906, | |
| "grad_norm": 0.3564852178096771, | |
| "learning_rate": 0.00018010697387368853, | |
| "loss": 0.0309, | |
| "step": 312000 | |
| }, | |
| { | |
| "epoch": 32.02627797166906, | |
| "eval_cer": 0.008145734597156399, | |
| "eval_loss": 0.01600516401231289, | |
| "eval_runtime": 22.3728, | |
| "eval_samples_per_second": 4.023, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 312000 | |
| }, | |
| { | |
| "epoch": 32.128926298501334, | |
| "grad_norm": 0.26071467995643616, | |
| "learning_rate": 0.00017907837893437565, | |
| "loss": 0.0307, | |
| "step": 313000 | |
| }, | |
| { | |
| "epoch": 32.128926298501334, | |
| "eval_cer": 0.008441943127962086, | |
| "eval_loss": 0.016347436234354973, | |
| "eval_runtime": 22.155, | |
| "eval_samples_per_second": 4.062, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 313000 | |
| }, | |
| { | |
| "epoch": 32.23157462533361, | |
| "grad_norm": 0.17180919647216797, | |
| "learning_rate": 0.00017804978399506275, | |
| "loss": 0.0307, | |
| "step": 314000 | |
| }, | |
| { | |
| "epoch": 32.23157462533361, | |
| "eval_cer": 0.009182464454976303, | |
| "eval_loss": 0.01613152027130127, | |
| "eval_runtime": 22.4889, | |
| "eval_samples_per_second": 4.002, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03753351206434316, | |
| "step": 314000 | |
| }, | |
| { | |
| "epoch": 32.33422295216588, | |
| "grad_norm": 0.16418492794036865, | |
| "learning_rate": 0.00017702118905574984, | |
| "loss": 0.0301, | |
| "step": 315000 | |
| }, | |
| { | |
| "epoch": 32.33422295216588, | |
| "eval_cer": 0.007701421800947867, | |
| "eval_loss": 0.015433421358466148, | |
| "eval_runtime": 22.2866, | |
| "eval_samples_per_second": 4.038, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 315000 | |
| }, | |
| { | |
| "epoch": 32.436871278998154, | |
| "grad_norm": 0.2582835853099823, | |
| "learning_rate": 0.00017599259411643696, | |
| "loss": 0.0307, | |
| "step": 316000 | |
| }, | |
| { | |
| "epoch": 32.436871278998154, | |
| "eval_cer": 0.00784952606635071, | |
| "eval_loss": 0.015836581587791443, | |
| "eval_runtime": 22.5371, | |
| "eval_samples_per_second": 3.993, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 316000 | |
| }, | |
| { | |
| "epoch": 32.53951960583043, | |
| "grad_norm": 0.2574046552181244, | |
| "learning_rate": 0.00017496399917712405, | |
| "loss": 0.0308, | |
| "step": 317000 | |
| }, | |
| { | |
| "epoch": 32.53951960583043, | |
| "eval_cer": 0.008441943127962086, | |
| "eval_loss": 0.015740592032670975, | |
| "eval_runtime": 22.3104, | |
| "eval_samples_per_second": 4.034, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03485254691689008, | |
| "step": 317000 | |
| }, | |
| { | |
| "epoch": 32.642167932662694, | |
| "grad_norm": 0.17120341956615448, | |
| "learning_rate": 0.00017393540423781115, | |
| "loss": 0.0304, | |
| "step": 318000 | |
| }, | |
| { | |
| "epoch": 32.642167932662694, | |
| "eval_cer": 0.008293838862559242, | |
| "eval_loss": 0.01630273461341858, | |
| "eval_runtime": 22.6719, | |
| "eval_samples_per_second": 3.97, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.035746201966041107, | |
| "step": 318000 | |
| }, | |
| { | |
| "epoch": 32.74481625949497, | |
| "grad_norm": 0.27363935112953186, | |
| "learning_rate": 0.00017290680929849824, | |
| "loss": 0.0309, | |
| "step": 319000 | |
| }, | |
| { | |
| "epoch": 32.74481625949497, | |
| "eval_cer": 0.009330568720379146, | |
| "eval_loss": 0.016513481736183167, | |
| "eval_runtime": 22.5102, | |
| "eval_samples_per_second": 3.998, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03932082216264522, | |
| "step": 319000 | |
| }, | |
| { | |
| "epoch": 32.84746458632724, | |
| "grad_norm": 0.29559651017189026, | |
| "learning_rate": 0.00017187821435918536, | |
| "loss": 0.0307, | |
| "step": 320000 | |
| }, | |
| { | |
| "epoch": 32.84746458632724, | |
| "eval_cer": 0.009330568720379146, | |
| "eval_loss": 0.01666964590549469, | |
| "eval_runtime": 22.527, | |
| "eval_samples_per_second": 3.995, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03932082216264522, | |
| "step": 320000 | |
| }, | |
| { | |
| "epoch": 32.950112913159515, | |
| "grad_norm": 0.1644178181886673, | |
| "learning_rate": 0.00017084961941987245, | |
| "loss": 0.0305, | |
| "step": 321000 | |
| }, | |
| { | |
| "epoch": 32.950112913159515, | |
| "eval_cer": 0.007701421800947867, | |
| "eval_loss": 0.01613970287144184, | |
| "eval_runtime": 22.6019, | |
| "eval_samples_per_second": 3.982, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 321000 | |
| }, | |
| { | |
| "epoch": 33.05276123999179, | |
| "grad_norm": 0.24239134788513184, | |
| "learning_rate": 0.00016982102448055955, | |
| "loss": 0.0301, | |
| "step": 322000 | |
| }, | |
| { | |
| "epoch": 33.05276123999179, | |
| "eval_cer": 0.008738151658767773, | |
| "eval_loss": 0.015571474097669125, | |
| "eval_runtime": 22.7263, | |
| "eval_samples_per_second": 3.96, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.035746201966041107, | |
| "step": 322000 | |
| }, | |
| { | |
| "epoch": 33.15540956682406, | |
| "grad_norm": 0.16490726172924042, | |
| "learning_rate": 0.00016879242954124667, | |
| "loss": 0.0296, | |
| "step": 323000 | |
| }, | |
| { | |
| "epoch": 33.15540956682406, | |
| "eval_cer": 0.008293838862559242, | |
| "eval_loss": 0.01544241514056921, | |
| "eval_runtime": 22.655, | |
| "eval_samples_per_second": 3.973, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.035746201966041107, | |
| "step": 323000 | |
| }, | |
| { | |
| "epoch": 33.258057893656336, | |
| "grad_norm": 0.19249847531318665, | |
| "learning_rate": 0.00016776383460193376, | |
| "loss": 0.0299, | |
| "step": 324000 | |
| }, | |
| { | |
| "epoch": 33.258057893656336, | |
| "eval_cer": 0.009478672985781991, | |
| "eval_loss": 0.015552397817373276, | |
| "eval_runtime": 22.3811, | |
| "eval_samples_per_second": 4.021, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03932082216264522, | |
| "step": 324000 | |
| }, | |
| { | |
| "epoch": 33.36070622048861, | |
| "grad_norm": 0.2540992200374603, | |
| "learning_rate": 0.00016673523966262085, | |
| "loss": 0.0298, | |
| "step": 325000 | |
| }, | |
| { | |
| "epoch": 33.36070622048861, | |
| "eval_cer": 0.008590047393364929, | |
| "eval_loss": 0.01545222382992506, | |
| "eval_runtime": 22.2235, | |
| "eval_samples_per_second": 4.05, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03485254691689008, | |
| "step": 325000 | |
| }, | |
| { | |
| "epoch": 33.463354547320876, | |
| "grad_norm": 0.17843221127986908, | |
| "learning_rate": 0.00016570664472330795, | |
| "loss": 0.0305, | |
| "step": 326000 | |
| }, | |
| { | |
| "epoch": 33.463354547320876, | |
| "eval_cer": 0.00784952606635071, | |
| "eval_loss": 0.015175366774201393, | |
| "eval_runtime": 22.1848, | |
| "eval_samples_per_second": 4.057, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.032171581769437, | |
| "step": 326000 | |
| }, | |
| { | |
| "epoch": 33.56600287415315, | |
| "grad_norm": 0.25464072823524475, | |
| "learning_rate": 0.00016467804978399507, | |
| "loss": 0.0306, | |
| "step": 327000 | |
| }, | |
| { | |
| "epoch": 33.56600287415315, | |
| "eval_cer": 0.008145734597156399, | |
| "eval_loss": 0.014621883630752563, | |
| "eval_runtime": 22.1184, | |
| "eval_samples_per_second": 4.069, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03485254691689008, | |
| "step": 327000 | |
| }, | |
| { | |
| "epoch": 33.66865120098542, | |
| "grad_norm": 0.16736038029193878, | |
| "learning_rate": 0.00016364945484468216, | |
| "loss": 0.0308, | |
| "step": 328000 | |
| }, | |
| { | |
| "epoch": 33.66865120098542, | |
| "eval_cer": 0.00903436018957346, | |
| "eval_loss": 0.01590174064040184, | |
| "eval_runtime": 22.2326, | |
| "eval_samples_per_second": 4.048, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.035746201966041107, | |
| "step": 328000 | |
| }, | |
| { | |
| "epoch": 33.771299527817696, | |
| "grad_norm": 0.24227890372276306, | |
| "learning_rate": 0.00016262085990536925, | |
| "loss": 0.0304, | |
| "step": 329000 | |
| }, | |
| { | |
| "epoch": 33.771299527817696, | |
| "eval_cer": 0.008145734597156399, | |
| "eval_loss": 0.015669850632548332, | |
| "eval_runtime": 22.3742, | |
| "eval_samples_per_second": 4.022, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.035746201966041107, | |
| "step": 329000 | |
| }, | |
| { | |
| "epoch": 33.87394785464997, | |
| "grad_norm": 0.18546319007873535, | |
| "learning_rate": 0.00016159226496605637, | |
| "loss": 0.0306, | |
| "step": 330000 | |
| }, | |
| { | |
| "epoch": 33.87394785464997, | |
| "eval_cer": 0.008145734597156399, | |
| "eval_loss": 0.015102799981832504, | |
| "eval_runtime": 22.5007, | |
| "eval_samples_per_second": 4.0, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 330000 | |
| }, | |
| { | |
| "epoch": 33.97659618148224, | |
| "grad_norm": 0.23194563388824463, | |
| "learning_rate": 0.00016056367002674347, | |
| "loss": 0.0305, | |
| "step": 331000 | |
| }, | |
| { | |
| "epoch": 33.97659618148224, | |
| "eval_cer": 0.007997630331753554, | |
| "eval_loss": 0.015225354582071304, | |
| "eval_runtime": 22.6228, | |
| "eval_samples_per_second": 3.978, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03485254691689008, | |
| "step": 331000 | |
| }, | |
| { | |
| "epoch": 34.07924450831452, | |
| "grad_norm": 0.18987098336219788, | |
| "learning_rate": 0.00015953507508743056, | |
| "loss": 0.0294, | |
| "step": 332000 | |
| }, | |
| { | |
| "epoch": 34.07924450831452, | |
| "eval_cer": 0.008145734597156399, | |
| "eval_loss": 0.015497728250920773, | |
| "eval_runtime": 22.4328, | |
| "eval_samples_per_second": 4.012, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 332000 | |
| }, | |
| { | |
| "epoch": 34.18189283514679, | |
| "grad_norm": 0.16452568769454956, | |
| "learning_rate": 0.00015850648014811765, | |
| "loss": 0.0299, | |
| "step": 333000 | |
| }, | |
| { | |
| "epoch": 34.18189283514679, | |
| "eval_cer": 0.008293838862559242, | |
| "eval_loss": 0.01510694995522499, | |
| "eval_runtime": 22.2639, | |
| "eval_samples_per_second": 4.042, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 333000 | |
| }, | |
| { | |
| "epoch": 34.28454116197906, | |
| "grad_norm": 0.2669082283973694, | |
| "learning_rate": 0.00015747788520880477, | |
| "loss": 0.0297, | |
| "step": 334000 | |
| }, | |
| { | |
| "epoch": 34.28454116197906, | |
| "eval_cer": 0.008293838862559242, | |
| "eval_loss": 0.015526418574154377, | |
| "eval_runtime": 22.4447, | |
| "eval_samples_per_second": 4.01, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03485254691689008, | |
| "step": 334000 | |
| }, | |
| { | |
| "epoch": 34.38718948881133, | |
| "grad_norm": 0.23023459315299988, | |
| "learning_rate": 0.00015644929026949187, | |
| "loss": 0.0298, | |
| "step": 335000 | |
| }, | |
| { | |
| "epoch": 34.38718948881133, | |
| "eval_cer": 0.008590047393364929, | |
| "eval_loss": 0.01542733982205391, | |
| "eval_runtime": 22.4734, | |
| "eval_samples_per_second": 4.005, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03753351206434316, | |
| "step": 335000 | |
| }, | |
| { | |
| "epoch": 34.489837815643604, | |
| "grad_norm": 0.3047637343406677, | |
| "learning_rate": 0.00015542069533017896, | |
| "loss": 0.0301, | |
| "step": 336000 | |
| }, | |
| { | |
| "epoch": 34.489837815643604, | |
| "eval_cer": 0.008886255924170616, | |
| "eval_loss": 0.0148982098326087, | |
| "eval_runtime": 22.2904, | |
| "eval_samples_per_second": 4.038, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03753351206434316, | |
| "step": 336000 | |
| }, | |
| { | |
| "epoch": 34.59248614247588, | |
| "grad_norm": 0.2189682275056839, | |
| "learning_rate": 0.00015439210039086608, | |
| "loss": 0.0299, | |
| "step": 337000 | |
| }, | |
| { | |
| "epoch": 34.59248614247588, | |
| "eval_cer": 0.00740521327014218, | |
| "eval_loss": 0.014870991930365562, | |
| "eval_runtime": 22.4248, | |
| "eval_samples_per_second": 4.013, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 337000 | |
| }, | |
| { | |
| "epoch": 34.69513446930815, | |
| "grad_norm": 0.1642763465642929, | |
| "learning_rate": 0.00015336350545155317, | |
| "loss": 0.03, | |
| "step": 338000 | |
| }, | |
| { | |
| "epoch": 34.69513446930815, | |
| "eval_cer": 0.008441943127962086, | |
| "eval_loss": 0.015146846882998943, | |
| "eval_runtime": 22.5811, | |
| "eval_samples_per_second": 3.986, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 338000 | |
| }, | |
| { | |
| "epoch": 34.797782796140424, | |
| "grad_norm": 0.21596594154834747, | |
| "learning_rate": 0.00015233491051224027, | |
| "loss": 0.03, | |
| "step": 339000 | |
| }, | |
| { | |
| "epoch": 34.797782796140424, | |
| "eval_cer": 0.007997630331753554, | |
| "eval_loss": 0.014487739652395248, | |
| "eval_runtime": 22.3193, | |
| "eval_samples_per_second": 4.032, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03485254691689008, | |
| "step": 339000 | |
| }, | |
| { | |
| "epoch": 34.9004311229727, | |
| "grad_norm": 0.15714465081691742, | |
| "learning_rate": 0.00015130631557292736, | |
| "loss": 0.0305, | |
| "step": 340000 | |
| }, | |
| { | |
| "epoch": 34.9004311229727, | |
| "eval_cer": 0.007997630331753554, | |
| "eval_loss": 0.014626013115048409, | |
| "eval_runtime": 22.3971, | |
| "eval_samples_per_second": 4.018, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03485254691689008, | |
| "step": 340000 | |
| }, | |
| { | |
| "epoch": 35.00307944980497, | |
| "grad_norm": 0.2639608383178711, | |
| "learning_rate": 0.00015027772063361448, | |
| "loss": 0.03, | |
| "step": 341000 | |
| }, | |
| { | |
| "epoch": 35.00307944980497, | |
| "eval_cer": 0.007997630331753554, | |
| "eval_loss": 0.014959813095629215, | |
| "eval_runtime": 22.4235, | |
| "eval_samples_per_second": 4.014, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 341000 | |
| }, | |
| { | |
| "epoch": 35.10572777663724, | |
| "grad_norm": 0.1461016684770584, | |
| "learning_rate": 0.00014924912569430157, | |
| "loss": 0.029, | |
| "step": 342000 | |
| }, | |
| { | |
| "epoch": 35.10572777663724, | |
| "eval_cer": 0.00784952606635071, | |
| "eval_loss": 0.015357970260083675, | |
| "eval_runtime": 22.4778, | |
| "eval_samples_per_second": 4.004, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 342000 | |
| }, | |
| { | |
| "epoch": 35.20837610346951, | |
| "grad_norm": 0.21027566492557526, | |
| "learning_rate": 0.00014822053075498867, | |
| "loss": 0.0294, | |
| "step": 343000 | |
| }, | |
| { | |
| "epoch": 35.20837610346951, | |
| "eval_cer": 0.008145734597156399, | |
| "eval_loss": 0.014925923198461533, | |
| "eval_runtime": 22.563, | |
| "eval_samples_per_second": 3.989, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 343000 | |
| }, | |
| { | |
| "epoch": 35.311024430301785, | |
| "grad_norm": 0.26173582673072815, | |
| "learning_rate": 0.0001471919358156758, | |
| "loss": 0.0294, | |
| "step": 344000 | |
| }, | |
| { | |
| "epoch": 35.311024430301785, | |
| "eval_cer": 0.007553317535545024, | |
| "eval_loss": 0.014718250371515751, | |
| "eval_runtime": 22.4858, | |
| "eval_samples_per_second": 4.003, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 344000 | |
| }, | |
| { | |
| "epoch": 35.41367275713406, | |
| "grad_norm": 0.21235166490077972, | |
| "learning_rate": 0.0001461633408763629, | |
| "loss": 0.0295, | |
| "step": 345000 | |
| }, | |
| { | |
| "epoch": 35.41367275713406, | |
| "eval_cer": 0.008293838862559242, | |
| "eval_loss": 0.014633492566645145, | |
| "eval_runtime": 22.4549, | |
| "eval_samples_per_second": 4.008, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 345000 | |
| }, | |
| { | |
| "epoch": 35.51632108396633, | |
| "grad_norm": 0.2826511263847351, | |
| "learning_rate": 0.00014513474593705, | |
| "loss": 0.0296, | |
| "step": 346000 | |
| }, | |
| { | |
| "epoch": 35.51632108396633, | |
| "eval_cer": 0.00903436018957346, | |
| "eval_loss": 0.015351605601608753, | |
| "eval_runtime": 22.6095, | |
| "eval_samples_per_second": 3.981, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 346000 | |
| }, | |
| { | |
| "epoch": 35.618969410798606, | |
| "grad_norm": 0.15846475958824158, | |
| "learning_rate": 0.00014410615099773712, | |
| "loss": 0.0298, | |
| "step": 347000 | |
| }, | |
| { | |
| "epoch": 35.618969410798606, | |
| "eval_cer": 0.007997630331753554, | |
| "eval_loss": 0.014954261481761932, | |
| "eval_runtime": 22.6016, | |
| "eval_samples_per_second": 3.982, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 347000 | |
| }, | |
| { | |
| "epoch": 35.72161773763088, | |
| "grad_norm": 0.30824708938598633, | |
| "learning_rate": 0.0001430775560584242, | |
| "loss": 0.0298, | |
| "step": 348000 | |
| }, | |
| { | |
| "epoch": 35.72161773763088, | |
| "eval_cer": 0.00784952606635071, | |
| "eval_loss": 0.015362138859927654, | |
| "eval_runtime": 22.4427, | |
| "eval_samples_per_second": 4.01, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 348000 | |
| }, | |
| { | |
| "epoch": 35.824266064463146, | |
| "grad_norm": 0.29468515515327454, | |
| "learning_rate": 0.0001420489611191113, | |
| "loss": 0.0297, | |
| "step": 349000 | |
| }, | |
| { | |
| "epoch": 35.824266064463146, | |
| "eval_cer": 0.008145734597156399, | |
| "eval_loss": 0.014831768348813057, | |
| "eval_runtime": 22.4167, | |
| "eval_samples_per_second": 4.015, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 349000 | |
| }, | |
| { | |
| "epoch": 35.92691439129542, | |
| "grad_norm": 0.3372247815132141, | |
| "learning_rate": 0.0001410203661797984, | |
| "loss": 0.0297, | |
| "step": 350000 | |
| }, | |
| { | |
| "epoch": 35.92691439129542, | |
| "eval_cer": 0.008293838862559242, | |
| "eval_loss": 0.015637291595339775, | |
| "eval_runtime": 22.6706, | |
| "eval_samples_per_second": 3.97, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03485254691689008, | |
| "step": 350000 | |
| }, | |
| { | |
| "epoch": 36.02956271812769, | |
| "grad_norm": 0.2713555693626404, | |
| "learning_rate": 0.00013999177124048552, | |
| "loss": 0.0295, | |
| "step": 351000 | |
| }, | |
| { | |
| "epoch": 36.02956271812769, | |
| "eval_cer": 0.008293838862559242, | |
| "eval_loss": 0.01471824198961258, | |
| "eval_runtime": 22.5633, | |
| "eval_samples_per_second": 3.989, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 351000 | |
| }, | |
| { | |
| "epoch": 36.132211044959966, | |
| "grad_norm": 0.2100389152765274, | |
| "learning_rate": 0.0001389631763011726, | |
| "loss": 0.0288, | |
| "step": 352000 | |
| }, | |
| { | |
| "epoch": 36.132211044959966, | |
| "eval_cer": 0.007553317535545024, | |
| "eval_loss": 0.014739991165697575, | |
| "eval_runtime": 22.6462, | |
| "eval_samples_per_second": 3.974, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 352000 | |
| }, | |
| { | |
| "epoch": 36.23485937179224, | |
| "grad_norm": 0.2058141529560089, | |
| "learning_rate": 0.0001379345813618597, | |
| "loss": 0.0294, | |
| "step": 353000 | |
| }, | |
| { | |
| "epoch": 36.23485937179224, | |
| "eval_cer": 0.007553317535545024, | |
| "eval_loss": 0.014310094527900219, | |
| "eval_runtime": 22.4136, | |
| "eval_samples_per_second": 4.015, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 353000 | |
| }, | |
| { | |
| "epoch": 36.33750769862451, | |
| "grad_norm": 0.21993263065814972, | |
| "learning_rate": 0.00013690598642254683, | |
| "loss": 0.0288, | |
| "step": 354000 | |
| }, | |
| { | |
| "epoch": 36.33750769862451, | |
| "eval_cer": 0.008293838862559242, | |
| "eval_loss": 0.01488146185874939, | |
| "eval_runtime": 22.7382, | |
| "eval_samples_per_second": 3.958, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 354000 | |
| }, | |
| { | |
| "epoch": 36.44015602545679, | |
| "grad_norm": 0.44507354497909546, | |
| "learning_rate": 0.00013587739148323392, | |
| "loss": 0.0291, | |
| "step": 355000 | |
| }, | |
| { | |
| "epoch": 36.44015602545679, | |
| "eval_cer": 0.008145734597156399, | |
| "eval_loss": 0.014914470724761486, | |
| "eval_runtime": 22.4894, | |
| "eval_samples_per_second": 4.002, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 355000 | |
| }, | |
| { | |
| "epoch": 36.54280435228906, | |
| "grad_norm": 0.29088860750198364, | |
| "learning_rate": 0.000134848796543921, | |
| "loss": 0.0297, | |
| "step": 356000 | |
| }, | |
| { | |
| "epoch": 36.54280435228906, | |
| "eval_cer": 0.008886255924170616, | |
| "eval_loss": 0.015177594497799873, | |
| "eval_runtime": 22.3664, | |
| "eval_samples_per_second": 4.024, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 356000 | |
| }, | |
| { | |
| "epoch": 36.64545267912133, | |
| "grad_norm": 0.28655806183815, | |
| "learning_rate": 0.0001338202016046081, | |
| "loss": 0.0291, | |
| "step": 357000 | |
| }, | |
| { | |
| "epoch": 36.64545267912133, | |
| "eval_cer": 0.008886255924170616, | |
| "eval_loss": 0.014854129403829575, | |
| "eval_runtime": 22.3549, | |
| "eval_samples_per_second": 4.026, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03485254691689008, | |
| "step": 357000 | |
| }, | |
| { | |
| "epoch": 36.7481010059536, | |
| "grad_norm": 0.16526740789413452, | |
| "learning_rate": 0.00013279160666529523, | |
| "loss": 0.0293, | |
| "step": 358000 | |
| }, | |
| { | |
| "epoch": 36.7481010059536, | |
| "eval_cer": 0.008293838862559242, | |
| "eval_loss": 0.015399602241814137, | |
| "eval_runtime": 22.5162, | |
| "eval_samples_per_second": 3.997, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.035746201966041107, | |
| "step": 358000 | |
| }, | |
| { | |
| "epoch": 36.850749332785874, | |
| "grad_norm": 0.16871041059494019, | |
| "learning_rate": 0.00013176301172598232, | |
| "loss": 0.0296, | |
| "step": 359000 | |
| }, | |
| { | |
| "epoch": 36.850749332785874, | |
| "eval_cer": 0.00740521327014218, | |
| "eval_loss": 0.015135127119719982, | |
| "eval_runtime": 22.5959, | |
| "eval_samples_per_second": 3.983, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 359000 | |
| }, | |
| { | |
| "epoch": 36.95339765961815, | |
| "grad_norm": 0.3259669542312622, | |
| "learning_rate": 0.0001307344167866694, | |
| "loss": 0.0296, | |
| "step": 360000 | |
| }, | |
| { | |
| "epoch": 36.95339765961815, | |
| "eval_cer": 0.007701421800947867, | |
| "eval_loss": 0.015445960685610771, | |
| "eval_runtime": 22.4067, | |
| "eval_samples_per_second": 4.017, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.032171581769437, | |
| "step": 360000 | |
| }, | |
| { | |
| "epoch": 37.05604598645042, | |
| "grad_norm": 0.2573050260543823, | |
| "learning_rate": 0.00012970582184735653, | |
| "loss": 0.0288, | |
| "step": 361000 | |
| }, | |
| { | |
| "epoch": 37.05604598645042, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.015334886498749256, | |
| "eval_runtime": 22.4559, | |
| "eval_samples_per_second": 4.008, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 361000 | |
| }, | |
| { | |
| "epoch": 37.158694313282695, | |
| "grad_norm": 0.3997851610183716, | |
| "learning_rate": 0.00012867722690804363, | |
| "loss": 0.0285, | |
| "step": 362000 | |
| }, | |
| { | |
| "epoch": 37.158694313282695, | |
| "eval_cer": 0.008293838862559242, | |
| "eval_loss": 0.015184923075139523, | |
| "eval_runtime": 22.5665, | |
| "eval_samples_per_second": 3.988, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03663985701519214, | |
| "step": 362000 | |
| }, | |
| { | |
| "epoch": 37.26134264011497, | |
| "grad_norm": 0.2915215790271759, | |
| "learning_rate": 0.00012764863196873072, | |
| "loss": 0.0287, | |
| "step": 363000 | |
| }, | |
| { | |
| "epoch": 37.26134264011497, | |
| "eval_cer": 0.007997630331753554, | |
| "eval_loss": 0.01511069294065237, | |
| "eval_runtime": 22.7402, | |
| "eval_samples_per_second": 3.958, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 363000 | |
| }, | |
| { | |
| "epoch": 37.36399096694724, | |
| "grad_norm": 0.1808217316865921, | |
| "learning_rate": 0.0001266200370294178, | |
| "loss": 0.0288, | |
| "step": 364000 | |
| }, | |
| { | |
| "epoch": 37.36399096694724, | |
| "eval_cer": 0.007997630331753554, | |
| "eval_loss": 0.015011030249297619, | |
| "eval_runtime": 22.3683, | |
| "eval_samples_per_second": 4.024, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 364000 | |
| }, | |
| { | |
| "epoch": 37.46663929377951, | |
| "grad_norm": 0.2159794569015503, | |
| "learning_rate": 0.00012559144209010493, | |
| "loss": 0.0289, | |
| "step": 365000 | |
| }, | |
| { | |
| "epoch": 37.46663929377951, | |
| "eval_cer": 0.008293838862559242, | |
| "eval_loss": 0.014206486754119396, | |
| "eval_runtime": 22.2265, | |
| "eval_samples_per_second": 4.049, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.035746201966041107, | |
| "step": 365000 | |
| }, | |
| { | |
| "epoch": 37.56928762061178, | |
| "grad_norm": 0.1461094170808792, | |
| "learning_rate": 0.00012456284715079203, | |
| "loss": 0.0291, | |
| "step": 366000 | |
| }, | |
| { | |
| "epoch": 37.56928762061178, | |
| "eval_cer": 0.007701421800947867, | |
| "eval_loss": 0.01417731773108244, | |
| "eval_runtime": 22.6758, | |
| "eval_samples_per_second": 3.969, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 366000 | |
| }, | |
| { | |
| "epoch": 37.671935947444055, | |
| "grad_norm": 0.18427909910678864, | |
| "learning_rate": 0.00012353425221147912, | |
| "loss": 0.0292, | |
| "step": 367000 | |
| }, | |
| { | |
| "epoch": 37.671935947444055, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.014333653263747692, | |
| "eval_runtime": 22.53, | |
| "eval_samples_per_second": 3.995, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 367000 | |
| }, | |
| { | |
| "epoch": 37.77458427427633, | |
| "grad_norm": 0.17020884156227112, | |
| "learning_rate": 0.00012250565727216624, | |
| "loss": 0.0292, | |
| "step": 368000 | |
| }, | |
| { | |
| "epoch": 37.77458427427633, | |
| "eval_cer": 0.007553317535545024, | |
| "eval_loss": 0.014181572012603283, | |
| "eval_runtime": 22.4523, | |
| "eval_samples_per_second": 4.008, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 368000 | |
| }, | |
| { | |
| "epoch": 37.8772326011086, | |
| "grad_norm": 0.15650001168251038, | |
| "learning_rate": 0.00012147706233285333, | |
| "loss": 0.0289, | |
| "step": 369000 | |
| }, | |
| { | |
| "epoch": 37.8772326011086, | |
| "eval_cer": 0.00784952606635071, | |
| "eval_loss": 0.014552351087331772, | |
| "eval_runtime": 22.33, | |
| "eval_samples_per_second": 4.03, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 369000 | |
| }, | |
| { | |
| "epoch": 37.979880927940876, | |
| "grad_norm": 0.25610971450805664, | |
| "learning_rate": 0.00012044846739354043, | |
| "loss": 0.0293, | |
| "step": 370000 | |
| }, | |
| { | |
| "epoch": 37.979880927940876, | |
| "eval_cer": 0.00740521327014218, | |
| "eval_loss": 0.01422215811908245, | |
| "eval_runtime": 22.9345, | |
| "eval_samples_per_second": 3.924, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.029490616621983913, | |
| "step": 370000 | |
| }, | |
| { | |
| "epoch": 38.08252925477315, | |
| "grad_norm": 0.25401365756988525, | |
| "learning_rate": 0.00011941987245422753, | |
| "loss": 0.0286, | |
| "step": 371000 | |
| }, | |
| { | |
| "epoch": 38.08252925477315, | |
| "eval_cer": 0.00740521327014218, | |
| "eval_loss": 0.014245335012674332, | |
| "eval_runtime": 22.3755, | |
| "eval_samples_per_second": 4.022, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 371000 | |
| }, | |
| { | |
| "epoch": 38.18517758160542, | |
| "grad_norm": 0.28409621119499207, | |
| "learning_rate": 0.00011839127751491462, | |
| "loss": 0.0281, | |
| "step": 372000 | |
| }, | |
| { | |
| "epoch": 38.18517758160542, | |
| "eval_cer": 0.00740521327014218, | |
| "eval_loss": 0.014165487140417099, | |
| "eval_runtime": 22.2381, | |
| "eval_samples_per_second": 4.047, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 372000 | |
| }, | |
| { | |
| "epoch": 38.28782590843769, | |
| "grad_norm": 0.20209653675556183, | |
| "learning_rate": 0.00011736268257560173, | |
| "loss": 0.0285, | |
| "step": 373000 | |
| }, | |
| { | |
| "epoch": 38.28782590843769, | |
| "eval_cer": 0.00740521327014218, | |
| "eval_loss": 0.014026058837771416, | |
| "eval_runtime": 22.4373, | |
| "eval_samples_per_second": 4.011, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.029490616621983913, | |
| "step": 373000 | |
| }, | |
| { | |
| "epoch": 38.39047423526996, | |
| "grad_norm": 0.2063857465982437, | |
| "learning_rate": 0.00011633408763628884, | |
| "loss": 0.029, | |
| "step": 374000 | |
| }, | |
| { | |
| "epoch": 38.39047423526996, | |
| "eval_cer": 0.007553317535545024, | |
| "eval_loss": 0.014216107316315174, | |
| "eval_runtime": 22.5613, | |
| "eval_samples_per_second": 3.989, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.032171581769437, | |
| "step": 374000 | |
| }, | |
| { | |
| "epoch": 38.493122562102236, | |
| "grad_norm": 0.1818021684885025, | |
| "learning_rate": 0.00011530549269697593, | |
| "loss": 0.0285, | |
| "step": 375000 | |
| }, | |
| { | |
| "epoch": 38.493122562102236, | |
| "eval_cer": 0.007997630331753554, | |
| "eval_loss": 0.014219836331903934, | |
| "eval_runtime": 22.5037, | |
| "eval_samples_per_second": 3.999, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 375000 | |
| }, | |
| { | |
| "epoch": 38.59577088893451, | |
| "grad_norm": 0.18460506200790405, | |
| "learning_rate": 0.00011427689775766304, | |
| "loss": 0.0285, | |
| "step": 376000 | |
| }, | |
| { | |
| "epoch": 38.59577088893451, | |
| "eval_cer": 0.008145734597156399, | |
| "eval_loss": 0.013656863942742348, | |
| "eval_runtime": 22.652, | |
| "eval_samples_per_second": 3.973, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 376000 | |
| }, | |
| { | |
| "epoch": 38.69841921576678, | |
| "grad_norm": 0.227520152926445, | |
| "learning_rate": 0.00011324830281835013, | |
| "loss": 0.0288, | |
| "step": 377000 | |
| }, | |
| { | |
| "epoch": 38.69841921576678, | |
| "eval_cer": 0.00784952606635071, | |
| "eval_loss": 0.01367582194507122, | |
| "eval_runtime": 22.645, | |
| "eval_samples_per_second": 3.974, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 377000 | |
| }, | |
| { | |
| "epoch": 38.80106754259906, | |
| "grad_norm": 0.1951012909412384, | |
| "learning_rate": 0.00011221970787903724, | |
| "loss": 0.0287, | |
| "step": 378000 | |
| }, | |
| { | |
| "epoch": 38.80106754259906, | |
| "eval_cer": 0.008738151658767773, | |
| "eval_loss": 0.014322535134851933, | |
| "eval_runtime": 22.4631, | |
| "eval_samples_per_second": 4.007, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03485254691689008, | |
| "step": 378000 | |
| }, | |
| { | |
| "epoch": 38.90371586943133, | |
| "grad_norm": 0.2870897352695465, | |
| "learning_rate": 0.00011119111293972433, | |
| "loss": 0.0288, | |
| "step": 379000 | |
| }, | |
| { | |
| "epoch": 38.90371586943133, | |
| "eval_cer": 0.00740521327014218, | |
| "eval_loss": 0.01387989055365324, | |
| "eval_runtime": 22.1509, | |
| "eval_samples_per_second": 4.063, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 379000 | |
| }, | |
| { | |
| "epoch": 39.006364196263604, | |
| "grad_norm": 0.1880166083574295, | |
| "learning_rate": 0.00011016251800041144, | |
| "loss": 0.0285, | |
| "step": 380000 | |
| }, | |
| { | |
| "epoch": 39.006364196263604, | |
| "eval_cer": 0.00784952606635071, | |
| "eval_loss": 0.014345232397317886, | |
| "eval_runtime": 22.2937, | |
| "eval_samples_per_second": 4.037, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 380000 | |
| }, | |
| { | |
| "epoch": 39.10901252309587, | |
| "grad_norm": 0.30459660291671753, | |
| "learning_rate": 0.00010913392306109854, | |
| "loss": 0.0279, | |
| "step": 381000 | |
| }, | |
| { | |
| "epoch": 39.10901252309587, | |
| "eval_cer": 0.008590047393364929, | |
| "eval_loss": 0.013818979263305664, | |
| "eval_runtime": 22.1693, | |
| "eval_samples_per_second": 4.06, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03485254691689008, | |
| "step": 381000 | |
| }, | |
| { | |
| "epoch": 39.211660849928144, | |
| "grad_norm": 0.2618006765842438, | |
| "learning_rate": 0.00010810532812178564, | |
| "loss": 0.0281, | |
| "step": 382000 | |
| }, | |
| { | |
| "epoch": 39.211660849928144, | |
| "eval_cer": 0.008145734597156399, | |
| "eval_loss": 0.013895167037844658, | |
| "eval_runtime": 22.6129, | |
| "eval_samples_per_second": 3.98, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 382000 | |
| }, | |
| { | |
| "epoch": 39.31430917676042, | |
| "grad_norm": 0.1827547401189804, | |
| "learning_rate": 0.00010707673318247274, | |
| "loss": 0.0281, | |
| "step": 383000 | |
| }, | |
| { | |
| "epoch": 39.31430917676042, | |
| "eval_cer": 0.006960900473933649, | |
| "eval_loss": 0.013730869628489017, | |
| "eval_runtime": 22.5617, | |
| "eval_samples_per_second": 3.989, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.028596961572832886, | |
| "step": 383000 | |
| }, | |
| { | |
| "epoch": 39.41695750359269, | |
| "grad_norm": 0.20615758001804352, | |
| "learning_rate": 0.00010604813824315984, | |
| "loss": 0.0283, | |
| "step": 384000 | |
| }, | |
| { | |
| "epoch": 39.41695750359269, | |
| "eval_cer": 0.0071090047393364926, | |
| "eval_loss": 0.014077040366828442, | |
| "eval_runtime": 22.4073, | |
| "eval_samples_per_second": 4.017, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 384000 | |
| }, | |
| { | |
| "epoch": 39.519605830424965, | |
| "grad_norm": 0.28781017661094666, | |
| "learning_rate": 0.00010501954330384694, | |
| "loss": 0.0284, | |
| "step": 385000 | |
| }, | |
| { | |
| "epoch": 39.519605830424965, | |
| "eval_cer": 0.0071090047393364926, | |
| "eval_loss": 0.013616991229355335, | |
| "eval_runtime": 22.4299, | |
| "eval_samples_per_second": 4.013, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 385000 | |
| }, | |
| { | |
| "epoch": 39.62225415725724, | |
| "grad_norm": 0.1402323842048645, | |
| "learning_rate": 0.00010399094836453404, | |
| "loss": 0.0287, | |
| "step": 386000 | |
| }, | |
| { | |
| "epoch": 39.62225415725724, | |
| "eval_cer": 0.007553317535545024, | |
| "eval_loss": 0.013439136557281017, | |
| "eval_runtime": 22.6508, | |
| "eval_samples_per_second": 3.973, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 386000 | |
| }, | |
| { | |
| "epoch": 39.72490248408951, | |
| "grad_norm": 0.22864773869514465, | |
| "learning_rate": 0.00010296235342522114, | |
| "loss": 0.0286, | |
| "step": 387000 | |
| }, | |
| { | |
| "epoch": 39.72490248408951, | |
| "eval_cer": 0.00784952606635071, | |
| "eval_loss": 0.01389585342258215, | |
| "eval_runtime": 22.2256, | |
| "eval_samples_per_second": 4.049, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 387000 | |
| }, | |
| { | |
| "epoch": 39.827550810921785, | |
| "grad_norm": 0.20083709061145782, | |
| "learning_rate": 0.00010193375848590825, | |
| "loss": 0.0287, | |
| "step": 388000 | |
| }, | |
| { | |
| "epoch": 39.827550810921785, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.014366346411406994, | |
| "eval_runtime": 22.4657, | |
| "eval_samples_per_second": 4.006, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 388000 | |
| }, | |
| { | |
| "epoch": 39.93019913775405, | |
| "grad_norm": 0.26983821392059326, | |
| "learning_rate": 0.00010090516354659534, | |
| "loss": 0.0285, | |
| "step": 389000 | |
| }, | |
| { | |
| "epoch": 39.93019913775405, | |
| "eval_cer": 0.008145734597156399, | |
| "eval_loss": 0.014381532557308674, | |
| "eval_runtime": 22.541, | |
| "eval_samples_per_second": 3.993, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 389000 | |
| }, | |
| { | |
| "epoch": 40.032847464586325, | |
| "grad_norm": 0.361680269241333, | |
| "learning_rate": 9.987656860728245e-05, | |
| "loss": 0.0278, | |
| "step": 390000 | |
| }, | |
| { | |
| "epoch": 40.032847464586325, | |
| "eval_cer": 0.007701421800947867, | |
| "eval_loss": 0.01420989166945219, | |
| "eval_runtime": 22.5652, | |
| "eval_samples_per_second": 3.988, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.032171581769437, | |
| "step": 390000 | |
| }, | |
| { | |
| "epoch": 40.1354957914186, | |
| "grad_norm": 0.17232652008533478, | |
| "learning_rate": 9.884797366796956e-05, | |
| "loss": 0.0276, | |
| "step": 391000 | |
| }, | |
| { | |
| "epoch": 40.1354957914186, | |
| "eval_cer": 0.00740521327014218, | |
| "eval_loss": 0.014762550592422485, | |
| "eval_runtime": 22.4968, | |
| "eval_samples_per_second": 4.001, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 391000 | |
| }, | |
| { | |
| "epoch": 40.23814411825087, | |
| "grad_norm": 0.35826170444488525, | |
| "learning_rate": 9.781937872865666e-05, | |
| "loss": 0.0278, | |
| "step": 392000 | |
| }, | |
| { | |
| "epoch": 40.23814411825087, | |
| "eval_cer": 0.007553317535545024, | |
| "eval_loss": 0.0146293630823493, | |
| "eval_runtime": 22.6001, | |
| "eval_samples_per_second": 3.982, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.032171581769437, | |
| "step": 392000 | |
| }, | |
| { | |
| "epoch": 40.340792445083146, | |
| "grad_norm": 0.41236042976379395, | |
| "learning_rate": 9.679078378934376e-05, | |
| "loss": 0.0279, | |
| "step": 393000 | |
| }, | |
| { | |
| "epoch": 40.340792445083146, | |
| "eval_cer": 0.00740521327014218, | |
| "eval_loss": 0.014173084869980812, | |
| "eval_runtime": 22.3588, | |
| "eval_samples_per_second": 4.025, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 393000 | |
| }, | |
| { | |
| "epoch": 40.44344077191542, | |
| "grad_norm": 0.13260366022586823, | |
| "learning_rate": 9.576218885003086e-05, | |
| "loss": 0.0281, | |
| "step": 394000 | |
| }, | |
| { | |
| "epoch": 40.44344077191542, | |
| "eval_cer": 0.006812796208530806, | |
| "eval_loss": 0.013978157192468643, | |
| "eval_runtime": 22.7638, | |
| "eval_samples_per_second": 3.954, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.028596961572832886, | |
| "step": 394000 | |
| }, | |
| { | |
| "epoch": 40.54608909874769, | |
| "grad_norm": 0.1822642832994461, | |
| "learning_rate": 9.473359391071797e-05, | |
| "loss": 0.0282, | |
| "step": 395000 | |
| }, | |
| { | |
| "epoch": 40.54608909874769, | |
| "eval_cer": 0.007701421800947867, | |
| "eval_loss": 0.014006286859512329, | |
| "eval_runtime": 23.5638, | |
| "eval_samples_per_second": 3.819, | |
| "eval_steps_per_second": 0.042, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 395000 | |
| }, | |
| { | |
| "epoch": 40.64873742557997, | |
| "grad_norm": 0.2515232563018799, | |
| "learning_rate": 9.370499897140506e-05, | |
| "loss": 0.0282, | |
| "step": 396000 | |
| }, | |
| { | |
| "epoch": 40.64873742557997, | |
| "eval_cer": 0.0071090047393364926, | |
| "eval_loss": 0.013893728144466877, | |
| "eval_runtime": 23.3851, | |
| "eval_samples_per_second": 3.849, | |
| "eval_steps_per_second": 0.043, | |
| "eval_wer": 0.029490616621983913, | |
| "step": 396000 | |
| }, | |
| { | |
| "epoch": 40.75138575241223, | |
| "grad_norm": 0.24263253808021545, | |
| "learning_rate": 9.267640403209217e-05, | |
| "loss": 0.0282, | |
| "step": 397000 | |
| }, | |
| { | |
| "epoch": 40.75138575241223, | |
| "eval_cer": 0.00740521327014218, | |
| "eval_loss": 0.013497700914740562, | |
| "eval_runtime": 22.831, | |
| "eval_samples_per_second": 3.942, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 397000 | |
| }, | |
| { | |
| "epoch": 40.85403407924451, | |
| "grad_norm": 0.3449910581111908, | |
| "learning_rate": 9.164780909277926e-05, | |
| "loss": 0.0281, | |
| "step": 398000 | |
| }, | |
| { | |
| "epoch": 40.85403407924451, | |
| "eval_cer": 0.007701421800947867, | |
| "eval_loss": 0.014030919410288334, | |
| "eval_runtime": 22.9536, | |
| "eval_samples_per_second": 3.921, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.032171581769437, | |
| "step": 398000 | |
| }, | |
| { | |
| "epoch": 40.95668240607678, | |
| "grad_norm": 0.34806039929389954, | |
| "learning_rate": 9.061921415346637e-05, | |
| "loss": 0.0283, | |
| "step": 399000 | |
| }, | |
| { | |
| "epoch": 40.95668240607678, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.013621107675135136, | |
| "eval_runtime": 22.7573, | |
| "eval_samples_per_second": 3.955, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 399000 | |
| }, | |
| { | |
| "epoch": 41.05933073290905, | |
| "grad_norm": 0.18996162712574005, | |
| "learning_rate": 8.959061921415348e-05, | |
| "loss": 0.0275, | |
| "step": 400000 | |
| }, | |
| { | |
| "epoch": 41.05933073290905, | |
| "eval_cer": 0.00740521327014218, | |
| "eval_loss": 0.013989981263875961, | |
| "eval_runtime": 22.345, | |
| "eval_samples_per_second": 4.028, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 400000 | |
| }, | |
| { | |
| "epoch": 41.16197905974133, | |
| "grad_norm": 0.22895024716854095, | |
| "learning_rate": 8.856202427484057e-05, | |
| "loss": 0.0278, | |
| "step": 401000 | |
| }, | |
| { | |
| "epoch": 41.16197905974133, | |
| "eval_cer": 0.007997630331753554, | |
| "eval_loss": 0.013726425357162952, | |
| "eval_runtime": 22.1918, | |
| "eval_samples_per_second": 4.056, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 401000 | |
| }, | |
| { | |
| "epoch": 41.2646273865736, | |
| "grad_norm": 0.20046815276145935, | |
| "learning_rate": 8.753342933552768e-05, | |
| "loss": 0.0276, | |
| "step": 402000 | |
| }, | |
| { | |
| "epoch": 41.2646273865736, | |
| "eval_cer": 0.007701421800947867, | |
| "eval_loss": 0.014126642607152462, | |
| "eval_runtime": 22.5351, | |
| "eval_samples_per_second": 3.994, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.032171581769437, | |
| "step": 402000 | |
| }, | |
| { | |
| "epoch": 41.367275713405874, | |
| "grad_norm": 0.2250887155532837, | |
| "learning_rate": 8.650483439621477e-05, | |
| "loss": 0.0277, | |
| "step": 403000 | |
| }, | |
| { | |
| "epoch": 41.367275713405874, | |
| "eval_cer": 0.007553317535545024, | |
| "eval_loss": 0.013941235840320587, | |
| "eval_runtime": 22.4208, | |
| "eval_samples_per_second": 4.014, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 403000 | |
| }, | |
| { | |
| "epoch": 41.46992404023814, | |
| "grad_norm": 0.20660826563835144, | |
| "learning_rate": 8.547623945690188e-05, | |
| "loss": 0.0278, | |
| "step": 404000 | |
| }, | |
| { | |
| "epoch": 41.46992404023814, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.013417248614132404, | |
| "eval_runtime": 22.349, | |
| "eval_samples_per_second": 4.027, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 404000 | |
| }, | |
| { | |
| "epoch": 41.572572367070414, | |
| "grad_norm": 0.20120590925216675, | |
| "learning_rate": 8.444764451758897e-05, | |
| "loss": 0.0278, | |
| "step": 405000 | |
| }, | |
| { | |
| "epoch": 41.572572367070414, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.01310759037733078, | |
| "eval_runtime": 22.3447, | |
| "eval_samples_per_second": 4.028, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 405000 | |
| }, | |
| { | |
| "epoch": 41.67522069390269, | |
| "grad_norm": 0.2711530327796936, | |
| "learning_rate": 8.341904957827608e-05, | |
| "loss": 0.0275, | |
| "step": 406000 | |
| }, | |
| { | |
| "epoch": 41.67522069390269, | |
| "eval_cer": 0.006664691943127962, | |
| "eval_loss": 0.012631074525415897, | |
| "eval_runtime": 22.6139, | |
| "eval_samples_per_second": 3.98, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.028596961572832886, | |
| "step": 406000 | |
| }, | |
| { | |
| "epoch": 41.77786902073496, | |
| "grad_norm": 0.16983696818351746, | |
| "learning_rate": 8.239045463896318e-05, | |
| "loss": 0.028, | |
| "step": 407000 | |
| }, | |
| { | |
| "epoch": 41.77786902073496, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.012849998660385609, | |
| "eval_runtime": 22.4766, | |
| "eval_samples_per_second": 4.004, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 407000 | |
| }, | |
| { | |
| "epoch": 41.880517347567235, | |
| "grad_norm": 0.1700281947851181, | |
| "learning_rate": 8.136185969965028e-05, | |
| "loss": 0.028, | |
| "step": 408000 | |
| }, | |
| { | |
| "epoch": 41.880517347567235, | |
| "eval_cer": 0.00740521327014218, | |
| "eval_loss": 0.012665662914514542, | |
| "eval_runtime": 22.4018, | |
| "eval_samples_per_second": 4.018, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 408000 | |
| }, | |
| { | |
| "epoch": 41.98316567439951, | |
| "grad_norm": 0.23489758372306824, | |
| "learning_rate": 8.033326476033738e-05, | |
| "loss": 0.0278, | |
| "step": 409000 | |
| }, | |
| { | |
| "epoch": 41.98316567439951, | |
| "eval_cer": 0.007553317535545024, | |
| "eval_loss": 0.013263718225061893, | |
| "eval_runtime": 22.3882, | |
| "eval_samples_per_second": 4.02, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 409000 | |
| }, | |
| { | |
| "epoch": 42.08581400123178, | |
| "grad_norm": 0.2607017159461975, | |
| "learning_rate": 7.930466982102448e-05, | |
| "loss": 0.0276, | |
| "step": 410000 | |
| }, | |
| { | |
| "epoch": 42.08581400123178, | |
| "eval_cer": 0.0071090047393364926, | |
| "eval_loss": 0.013156604953110218, | |
| "eval_runtime": 22.6758, | |
| "eval_samples_per_second": 3.969, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 410000 | |
| }, | |
| { | |
| "epoch": 42.188462328064055, | |
| "grad_norm": 0.20945683121681213, | |
| "learning_rate": 7.827607488171158e-05, | |
| "loss": 0.0272, | |
| "step": 411000 | |
| }, | |
| { | |
| "epoch": 42.188462328064055, | |
| "eval_cer": 0.006664691943127962, | |
| "eval_loss": 0.012946737930178642, | |
| "eval_runtime": 22.4875, | |
| "eval_samples_per_second": 4.002, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.028596961572832886, | |
| "step": 411000 | |
| }, | |
| { | |
| "epoch": 42.29111065489632, | |
| "grad_norm": 0.25923091173171997, | |
| "learning_rate": 7.724747994239868e-05, | |
| "loss": 0.0274, | |
| "step": 412000 | |
| }, | |
| { | |
| "epoch": 42.29111065489632, | |
| "eval_cer": 0.006960900473933649, | |
| "eval_loss": 0.012892471626400948, | |
| "eval_runtime": 22.5377, | |
| "eval_samples_per_second": 3.993, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.029490616621983913, | |
| "step": 412000 | |
| }, | |
| { | |
| "epoch": 42.393758981728595, | |
| "grad_norm": 0.2663125693798065, | |
| "learning_rate": 7.621888500308578e-05, | |
| "loss": 0.0273, | |
| "step": 413000 | |
| }, | |
| { | |
| "epoch": 42.393758981728595, | |
| "eval_cer": 0.006664691943127962, | |
| "eval_loss": 0.012838827446103096, | |
| "eval_runtime": 22.3627, | |
| "eval_samples_per_second": 4.025, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.029490616621983913, | |
| "step": 413000 | |
| }, | |
| { | |
| "epoch": 42.49640730856087, | |
| "grad_norm": 0.1551075577735901, | |
| "learning_rate": 7.519029006377289e-05, | |
| "loss": 0.0275, | |
| "step": 414000 | |
| }, | |
| { | |
| "epoch": 42.49640730856087, | |
| "eval_cer": 0.006664691943127962, | |
| "eval_loss": 0.01299965288490057, | |
| "eval_runtime": 22.7331, | |
| "eval_samples_per_second": 3.959, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.029490616621983913, | |
| "step": 414000 | |
| }, | |
| { | |
| "epoch": 42.59905563539314, | |
| "grad_norm": 0.16298335790634155, | |
| "learning_rate": 7.416169512445998e-05, | |
| "loss": 0.0275, | |
| "step": 415000 | |
| }, | |
| { | |
| "epoch": 42.59905563539314, | |
| "eval_cer": 0.0071090047393364926, | |
| "eval_loss": 0.013336721807718277, | |
| "eval_runtime": 22.767, | |
| "eval_samples_per_second": 3.953, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 415000 | |
| }, | |
| { | |
| "epoch": 42.701703962225416, | |
| "grad_norm": 0.20813694596290588, | |
| "learning_rate": 7.313310018514709e-05, | |
| "loss": 0.0278, | |
| "step": 416000 | |
| }, | |
| { | |
| "epoch": 42.701703962225416, | |
| "eval_cer": 0.007553317535545024, | |
| "eval_loss": 0.0134715735912323, | |
| "eval_runtime": 22.6511, | |
| "eval_samples_per_second": 3.973, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.032171581769437, | |
| "step": 416000 | |
| }, | |
| { | |
| "epoch": 42.80435228905769, | |
| "grad_norm": 0.16702768206596375, | |
| "learning_rate": 7.210450524583418e-05, | |
| "loss": 0.0279, | |
| "step": 417000 | |
| }, | |
| { | |
| "epoch": 42.80435228905769, | |
| "eval_cer": 0.007701421800947867, | |
| "eval_loss": 0.013460041023790836, | |
| "eval_runtime": 22.4175, | |
| "eval_samples_per_second": 4.015, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 417000 | |
| }, | |
| { | |
| "epoch": 42.90700061588996, | |
| "grad_norm": 0.2635751962661743, | |
| "learning_rate": 7.107591030652129e-05, | |
| "loss": 0.0273, | |
| "step": 418000 | |
| }, | |
| { | |
| "epoch": 42.90700061588996, | |
| "eval_cer": 0.00740521327014218, | |
| "eval_loss": 0.013223396614193916, | |
| "eval_runtime": 22.8666, | |
| "eval_samples_per_second": 3.936, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 418000 | |
| }, | |
| { | |
| "epoch": 43.00964894272224, | |
| "grad_norm": 0.1874234676361084, | |
| "learning_rate": 7.004731536720838e-05, | |
| "loss": 0.0276, | |
| "step": 419000 | |
| }, | |
| { | |
| "epoch": 43.00964894272224, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.013210386037826538, | |
| "eval_runtime": 22.5237, | |
| "eval_samples_per_second": 3.996, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.032171581769437, | |
| "step": 419000 | |
| }, | |
| { | |
| "epoch": 43.1122972695545, | |
| "grad_norm": 0.16425500810146332, | |
| "learning_rate": 6.901872042789549e-05, | |
| "loss": 0.0269, | |
| "step": 420000 | |
| }, | |
| { | |
| "epoch": 43.1122972695545, | |
| "eval_cer": 0.0071090047393364926, | |
| "eval_loss": 0.013268969021737576, | |
| "eval_runtime": 22.5906, | |
| "eval_samples_per_second": 3.984, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 420000 | |
| }, | |
| { | |
| "epoch": 43.21494559638678, | |
| "grad_norm": 0.14309245347976685, | |
| "learning_rate": 6.79901254885826e-05, | |
| "loss": 0.0273, | |
| "step": 421000 | |
| }, | |
| { | |
| "epoch": 43.21494559638678, | |
| "eval_cer": 0.00784952606635071, | |
| "eval_loss": 0.012923350557684898, | |
| "eval_runtime": 22.6017, | |
| "eval_samples_per_second": 3.982, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.032171581769437, | |
| "step": 421000 | |
| }, | |
| { | |
| "epoch": 43.31759392321905, | |
| "grad_norm": 0.3173372447490692, | |
| "learning_rate": 6.69615305492697e-05, | |
| "loss": 0.0271, | |
| "step": 422000 | |
| }, | |
| { | |
| "epoch": 43.31759392321905, | |
| "eval_cer": 0.007701421800947867, | |
| "eval_loss": 0.013023743405938148, | |
| "eval_runtime": 22.4573, | |
| "eval_samples_per_second": 4.008, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.032171581769437, | |
| "step": 422000 | |
| }, | |
| { | |
| "epoch": 43.420242250051324, | |
| "grad_norm": 0.2641269266605377, | |
| "learning_rate": 6.593293560995681e-05, | |
| "loss": 0.0274, | |
| "step": 423000 | |
| }, | |
| { | |
| "epoch": 43.420242250051324, | |
| "eval_cer": 0.007701421800947867, | |
| "eval_loss": 0.013114223256707191, | |
| "eval_runtime": 22.3329, | |
| "eval_samples_per_second": 4.03, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.032171581769437, | |
| "step": 423000 | |
| }, | |
| { | |
| "epoch": 43.5228905768836, | |
| "grad_norm": 0.20256099104881287, | |
| "learning_rate": 6.49043406706439e-05, | |
| "loss": 0.0273, | |
| "step": 424000 | |
| }, | |
| { | |
| "epoch": 43.5228905768836, | |
| "eval_cer": 0.008441943127962086, | |
| "eval_loss": 0.013389894738793373, | |
| "eval_runtime": 22.4353, | |
| "eval_samples_per_second": 4.012, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 424000 | |
| }, | |
| { | |
| "epoch": 43.62553890371587, | |
| "grad_norm": 0.2633448541164398, | |
| "learning_rate": 6.387574573133101e-05, | |
| "loss": 0.0273, | |
| "step": 425000 | |
| }, | |
| { | |
| "epoch": 43.62553890371587, | |
| "eval_cer": 0.008293838862559242, | |
| "eval_loss": 0.013175377622246742, | |
| "eval_runtime": 22.0476, | |
| "eval_samples_per_second": 4.082, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 425000 | |
| }, | |
| { | |
| "epoch": 43.728187230548144, | |
| "grad_norm": 0.14357689023017883, | |
| "learning_rate": 6.28471507920181e-05, | |
| "loss": 0.0271, | |
| "step": 426000 | |
| }, | |
| { | |
| "epoch": 43.728187230548144, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.013218458741903305, | |
| "eval_runtime": 22.1794, | |
| "eval_samples_per_second": 4.058, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 426000 | |
| }, | |
| { | |
| "epoch": 43.83083555738042, | |
| "grad_norm": 0.20153464376926422, | |
| "learning_rate": 6.181855585270521e-05, | |
| "loss": 0.0275, | |
| "step": 427000 | |
| }, | |
| { | |
| "epoch": 43.83083555738042, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.013227393850684166, | |
| "eval_runtime": 22.262, | |
| "eval_samples_per_second": 4.043, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 427000 | |
| }, | |
| { | |
| "epoch": 43.933483884212684, | |
| "grad_norm": 0.1699199080467224, | |
| "learning_rate": 6.07899609133923e-05, | |
| "loss": 0.0272, | |
| "step": 428000 | |
| }, | |
| { | |
| "epoch": 43.933483884212684, | |
| "eval_cer": 0.0071090047393364926, | |
| "eval_loss": 0.013098928146064281, | |
| "eval_runtime": 22.2399, | |
| "eval_samples_per_second": 4.047, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 428000 | |
| }, | |
| { | |
| "epoch": 44.03613221104496, | |
| "grad_norm": 0.21964909136295319, | |
| "learning_rate": 5.97613659740794e-05, | |
| "loss": 0.0268, | |
| "step": 429000 | |
| }, | |
| { | |
| "epoch": 44.03613221104496, | |
| "eval_cer": 0.0071090047393364926, | |
| "eval_loss": 0.013152836821973324, | |
| "eval_runtime": 22.2431, | |
| "eval_samples_per_second": 4.046, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 429000 | |
| }, | |
| { | |
| "epoch": 44.13878053787723, | |
| "grad_norm": 0.2735491693019867, | |
| "learning_rate": 5.873277103476652e-05, | |
| "loss": 0.0265, | |
| "step": 430000 | |
| }, | |
| { | |
| "epoch": 44.13878053787723, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.013426227495074272, | |
| "eval_runtime": 22.6558, | |
| "eval_samples_per_second": 3.973, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.032171581769437, | |
| "step": 430000 | |
| }, | |
| { | |
| "epoch": 44.241428864709505, | |
| "grad_norm": 0.18371999263763428, | |
| "learning_rate": 5.7704176095453617e-05, | |
| "loss": 0.0269, | |
| "step": 431000 | |
| }, | |
| { | |
| "epoch": 44.241428864709505, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.013290656730532646, | |
| "eval_runtime": 22.5067, | |
| "eval_samples_per_second": 3.999, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.032171581769437, | |
| "step": 431000 | |
| }, | |
| { | |
| "epoch": 44.34407719154178, | |
| "grad_norm": 0.3648182451725006, | |
| "learning_rate": 5.6675581156140717e-05, | |
| "loss": 0.0269, | |
| "step": 432000 | |
| }, | |
| { | |
| "epoch": 44.34407719154178, | |
| "eval_cer": 0.0071090047393364926, | |
| "eval_loss": 0.013338846154510975, | |
| "eval_runtime": 22.5224, | |
| "eval_samples_per_second": 3.996, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 432000 | |
| }, | |
| { | |
| "epoch": 44.44672551837405, | |
| "grad_norm": 0.14803066849708557, | |
| "learning_rate": 5.5646986216827816e-05, | |
| "loss": 0.0272, | |
| "step": 433000 | |
| }, | |
| { | |
| "epoch": 44.44672551837405, | |
| "eval_cer": 0.007701421800947867, | |
| "eval_loss": 0.013501363806426525, | |
| "eval_runtime": 22.4789, | |
| "eval_samples_per_second": 4.004, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.032171581769437, | |
| "step": 433000 | |
| }, | |
| { | |
| "epoch": 44.549373845206325, | |
| "grad_norm": 0.13751809298992157, | |
| "learning_rate": 5.4618391277514916e-05, | |
| "loss": 0.0275, | |
| "step": 434000 | |
| }, | |
| { | |
| "epoch": 44.549373845206325, | |
| "eval_cer": 0.007701421800947867, | |
| "eval_loss": 0.013389664702117443, | |
| "eval_runtime": 22.5113, | |
| "eval_samples_per_second": 3.998, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.032171581769437, | |
| "step": 434000 | |
| }, | |
| { | |
| "epoch": 44.6520221720386, | |
| "grad_norm": 0.18634863197803497, | |
| "learning_rate": 5.3589796338202016e-05, | |
| "loss": 0.0271, | |
| "step": 435000 | |
| }, | |
| { | |
| "epoch": 44.6520221720386, | |
| "eval_cer": 0.007701421800947867, | |
| "eval_loss": 0.013616056181490421, | |
| "eval_runtime": 22.6218, | |
| "eval_samples_per_second": 3.978, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.032171581769437, | |
| "step": 435000 | |
| }, | |
| { | |
| "epoch": 44.754670498870865, | |
| "grad_norm": 0.20694060623645782, | |
| "learning_rate": 5.2561201398889116e-05, | |
| "loss": 0.0272, | |
| "step": 436000 | |
| }, | |
| { | |
| "epoch": 44.754670498870865, | |
| "eval_cer": 0.007997630331753554, | |
| "eval_loss": 0.013776379637420177, | |
| "eval_runtime": 22.7323, | |
| "eval_samples_per_second": 3.959, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 436000 | |
| }, | |
| { | |
| "epoch": 44.85731882570314, | |
| "grad_norm": 0.22977876663208008, | |
| "learning_rate": 5.153260645957622e-05, | |
| "loss": 0.0269, | |
| "step": 437000 | |
| }, | |
| { | |
| "epoch": 44.85731882570314, | |
| "eval_cer": 0.006664691943127962, | |
| "eval_loss": 0.01339475717395544, | |
| "eval_runtime": 22.3891, | |
| "eval_samples_per_second": 4.02, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.028596961572832886, | |
| "step": 437000 | |
| }, | |
| { | |
| "epoch": 44.95996715253541, | |
| "grad_norm": 0.16584943234920502, | |
| "learning_rate": 5.050401152026332e-05, | |
| "loss": 0.0269, | |
| "step": 438000 | |
| }, | |
| { | |
| "epoch": 44.95996715253541, | |
| "eval_cer": 0.00740521327014218, | |
| "eval_loss": 0.013209763914346695, | |
| "eval_runtime": 22.5389, | |
| "eval_samples_per_second": 3.993, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 438000 | |
| }, | |
| { | |
| "epoch": 45.062615479367686, | |
| "grad_norm": 0.2402484118938446, | |
| "learning_rate": 4.947541658095042e-05, | |
| "loss": 0.0268, | |
| "step": 439000 | |
| }, | |
| { | |
| "epoch": 45.062615479367686, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.013325601816177368, | |
| "eval_runtime": 22.5418, | |
| "eval_samples_per_second": 3.993, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 439000 | |
| }, | |
| { | |
| "epoch": 45.16526380619996, | |
| "grad_norm": 0.18944856524467468, | |
| "learning_rate": 4.844682164163752e-05, | |
| "loss": 0.0265, | |
| "step": 440000 | |
| }, | |
| { | |
| "epoch": 45.16526380619996, | |
| "eval_cer": 0.00740521327014218, | |
| "eval_loss": 0.012947525829076767, | |
| "eval_runtime": 22.3674, | |
| "eval_samples_per_second": 4.024, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 440000 | |
| }, | |
| { | |
| "epoch": 45.26791213303223, | |
| "grad_norm": 0.18352921307086945, | |
| "learning_rate": 4.741822670232462e-05, | |
| "loss": 0.0265, | |
| "step": 441000 | |
| }, | |
| { | |
| "epoch": 45.26791213303223, | |
| "eval_cer": 0.006960900473933649, | |
| "eval_loss": 0.012937680818140507, | |
| "eval_runtime": 22.3946, | |
| "eval_samples_per_second": 4.019, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 441000 | |
| }, | |
| { | |
| "epoch": 45.37056045986451, | |
| "grad_norm": 0.2022509127855301, | |
| "learning_rate": 4.638963176301172e-05, | |
| "loss": 0.0266, | |
| "step": 442000 | |
| }, | |
| { | |
| "epoch": 45.37056045986451, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.012964108027517796, | |
| "eval_runtime": 22.5768, | |
| "eval_samples_per_second": 3.986, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 442000 | |
| }, | |
| { | |
| "epoch": 45.47320878669678, | |
| "grad_norm": 0.3211962878704071, | |
| "learning_rate": 4.536103682369882e-05, | |
| "loss": 0.0269, | |
| "step": 443000 | |
| }, | |
| { | |
| "epoch": 45.47320878669678, | |
| "eval_cer": 0.006960900473933649, | |
| "eval_loss": 0.01303518284112215, | |
| "eval_runtime": 22.4387, | |
| "eval_samples_per_second": 4.011, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.029490616621983913, | |
| "step": 443000 | |
| }, | |
| { | |
| "epoch": 45.57585711352905, | |
| "grad_norm": 0.1719512939453125, | |
| "learning_rate": 4.433244188438593e-05, | |
| "loss": 0.0268, | |
| "step": 444000 | |
| }, | |
| { | |
| "epoch": 45.57585711352905, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.012754334136843681, | |
| "eval_runtime": 22.842, | |
| "eval_samples_per_second": 3.94, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 444000 | |
| }, | |
| { | |
| "epoch": 45.67850544036132, | |
| "grad_norm": 0.1725946068763733, | |
| "learning_rate": 4.330384694507303e-05, | |
| "loss": 0.0265, | |
| "step": 445000 | |
| }, | |
| { | |
| "epoch": 45.67850544036132, | |
| "eval_cer": 0.007553317535545024, | |
| "eval_loss": 0.012887077406048775, | |
| "eval_runtime": 22.7726, | |
| "eval_samples_per_second": 3.952, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.032171581769437, | |
| "step": 445000 | |
| }, | |
| { | |
| "epoch": 45.781153767193594, | |
| "grad_norm": 0.17726309597492218, | |
| "learning_rate": 4.2275252005760136e-05, | |
| "loss": 0.0269, | |
| "step": 446000 | |
| }, | |
| { | |
| "epoch": 45.781153767193594, | |
| "eval_cer": 0.006812796208530806, | |
| "eval_loss": 0.012967323884367943, | |
| "eval_runtime": 22.8533, | |
| "eval_samples_per_second": 3.938, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.029490616621983913, | |
| "step": 446000 | |
| }, | |
| { | |
| "epoch": 45.88380209402587, | |
| "grad_norm": 0.2569744288921356, | |
| "learning_rate": 4.1246657066447236e-05, | |
| "loss": 0.0268, | |
| "step": 447000 | |
| }, | |
| { | |
| "epoch": 45.88380209402587, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.013006458058953285, | |
| "eval_runtime": 22.479, | |
| "eval_samples_per_second": 4.004, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 447000 | |
| }, | |
| { | |
| "epoch": 45.98645042085814, | |
| "grad_norm": 0.18339526653289795, | |
| "learning_rate": 4.0218062127134336e-05, | |
| "loss": 0.0272, | |
| "step": 448000 | |
| }, | |
| { | |
| "epoch": 45.98645042085814, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.012899891473352909, | |
| "eval_runtime": 22.2876, | |
| "eval_samples_per_second": 4.038, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 448000 | |
| }, | |
| { | |
| "epoch": 46.089098747690414, | |
| "grad_norm": 0.21979086101055145, | |
| "learning_rate": 3.9189467187821436e-05, | |
| "loss": 0.0266, | |
| "step": 449000 | |
| }, | |
| { | |
| "epoch": 46.089098747690414, | |
| "eval_cer": 0.007553317535545024, | |
| "eval_loss": 0.012851923704147339, | |
| "eval_runtime": 22.1005, | |
| "eval_samples_per_second": 4.072, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.032171581769437, | |
| "step": 449000 | |
| }, | |
| { | |
| "epoch": 46.19174707452269, | |
| "grad_norm": 0.18747855722904205, | |
| "learning_rate": 3.816087224850854e-05, | |
| "loss": 0.0263, | |
| "step": 450000 | |
| }, | |
| { | |
| "epoch": 46.19174707452269, | |
| "eval_cer": 0.007553317535545024, | |
| "eval_loss": 0.012800745666027069, | |
| "eval_runtime": 22.2631, | |
| "eval_samples_per_second": 4.043, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.032171581769437, | |
| "step": 450000 | |
| }, | |
| { | |
| "epoch": 46.29439540135496, | |
| "grad_norm": 0.20750053226947784, | |
| "learning_rate": 3.713227730919564e-05, | |
| "loss": 0.0263, | |
| "step": 451000 | |
| }, | |
| { | |
| "epoch": 46.29439540135496, | |
| "eval_cer": 0.00740521327014218, | |
| "eval_loss": 0.012725234031677246, | |
| "eval_runtime": 22.496, | |
| "eval_samples_per_second": 4.001, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.032171581769437, | |
| "step": 451000 | |
| }, | |
| { | |
| "epoch": 46.39704372818723, | |
| "grad_norm": 0.19758926331996918, | |
| "learning_rate": 3.610368236988274e-05, | |
| "loss": 0.027, | |
| "step": 452000 | |
| }, | |
| { | |
| "epoch": 46.39704372818723, | |
| "eval_cer": 0.00740521327014218, | |
| "eval_loss": 0.012631443329155445, | |
| "eval_runtime": 22.4978, | |
| "eval_samples_per_second": 4.0, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.032171581769437, | |
| "step": 452000 | |
| }, | |
| { | |
| "epoch": 46.4996920550195, | |
| "grad_norm": 0.23830477893352509, | |
| "learning_rate": 3.507508743056984e-05, | |
| "loss": 0.0267, | |
| "step": 453000 | |
| }, | |
| { | |
| "epoch": 46.4996920550195, | |
| "eval_cer": 0.007701421800947867, | |
| "eval_loss": 0.012887844815850258, | |
| "eval_runtime": 22.6407, | |
| "eval_samples_per_second": 3.975, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 453000 | |
| }, | |
| { | |
| "epoch": 46.602340381851775, | |
| "grad_norm": 0.22783514857292175, | |
| "learning_rate": 3.404649249125694e-05, | |
| "loss": 0.026, | |
| "step": 454000 | |
| }, | |
| { | |
| "epoch": 46.602340381851775, | |
| "eval_cer": 0.007553317535545024, | |
| "eval_loss": 0.012926483526825905, | |
| "eval_runtime": 22.7995, | |
| "eval_samples_per_second": 3.947, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 454000 | |
| }, | |
| { | |
| "epoch": 46.70498870868405, | |
| "grad_norm": 0.17195259034633636, | |
| "learning_rate": 3.301789755194404e-05, | |
| "loss": 0.0269, | |
| "step": 455000 | |
| }, | |
| { | |
| "epoch": 46.70498870868405, | |
| "eval_cer": 0.007553317535545024, | |
| "eval_loss": 0.012897716835141182, | |
| "eval_runtime": 22.735, | |
| "eval_samples_per_second": 3.959, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.032171581769437, | |
| "step": 455000 | |
| }, | |
| { | |
| "epoch": 46.80763703551632, | |
| "grad_norm": 0.21375824511051178, | |
| "learning_rate": 3.198930261263114e-05, | |
| "loss": 0.0267, | |
| "step": 456000 | |
| }, | |
| { | |
| "epoch": 46.80763703551632, | |
| "eval_cer": 0.00784952606635071, | |
| "eval_loss": 0.01305030845105648, | |
| "eval_runtime": 22.6049, | |
| "eval_samples_per_second": 3.981, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 456000 | |
| }, | |
| { | |
| "epoch": 46.910285362348596, | |
| "grad_norm": 0.2291615605354309, | |
| "learning_rate": 3.096070767331825e-05, | |
| "loss": 0.0269, | |
| "step": 457000 | |
| }, | |
| { | |
| "epoch": 46.910285362348596, | |
| "eval_cer": 0.007553317535545024, | |
| "eval_loss": 0.013074580579996109, | |
| "eval_runtime": 22.4586, | |
| "eval_samples_per_second": 4.007, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03395889186773905, | |
| "step": 457000 | |
| }, | |
| { | |
| "epoch": 47.01293368918087, | |
| "grad_norm": 0.18737538158893585, | |
| "learning_rate": 2.993211273400535e-05, | |
| "loss": 0.0266, | |
| "step": 458000 | |
| }, | |
| { | |
| "epoch": 47.01293368918087, | |
| "eval_cer": 0.0071090047393364926, | |
| "eval_loss": 0.01272270642220974, | |
| "eval_runtime": 22.7233, | |
| "eval_samples_per_second": 3.961, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 458000 | |
| }, | |
| { | |
| "epoch": 47.115582016013136, | |
| "grad_norm": 0.20762291550636292, | |
| "learning_rate": 2.8903517794692452e-05, | |
| "loss": 0.0263, | |
| "step": 459000 | |
| }, | |
| { | |
| "epoch": 47.115582016013136, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.01276144664734602, | |
| "eval_runtime": 22.6501, | |
| "eval_samples_per_second": 3.973, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 459000 | |
| }, | |
| { | |
| "epoch": 47.21823034284541, | |
| "grad_norm": 0.22940364480018616, | |
| "learning_rate": 2.7874922855379552e-05, | |
| "loss": 0.0263, | |
| "step": 460000 | |
| }, | |
| { | |
| "epoch": 47.21823034284541, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.012876234017312527, | |
| "eval_runtime": 22.497, | |
| "eval_samples_per_second": 4.001, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 460000 | |
| }, | |
| { | |
| "epoch": 47.32087866967768, | |
| "grad_norm": 0.1963769495487213, | |
| "learning_rate": 2.6846327916066652e-05, | |
| "loss": 0.0262, | |
| "step": 461000 | |
| }, | |
| { | |
| "epoch": 47.32087866967768, | |
| "eval_cer": 0.00740521327014218, | |
| "eval_loss": 0.012759842909872532, | |
| "eval_runtime": 22.4933, | |
| "eval_samples_per_second": 4.001, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.032171581769437, | |
| "step": 461000 | |
| }, | |
| { | |
| "epoch": 47.423526996509956, | |
| "grad_norm": 0.13898225128650665, | |
| "learning_rate": 2.5817732976753755e-05, | |
| "loss": 0.0262, | |
| "step": 462000 | |
| }, | |
| { | |
| "epoch": 47.423526996509956, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.012645237147808075, | |
| "eval_runtime": 22.7141, | |
| "eval_samples_per_second": 3.962, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 462000 | |
| }, | |
| { | |
| "epoch": 47.52617532334223, | |
| "grad_norm": 0.295411616563797, | |
| "learning_rate": 2.4789138037440855e-05, | |
| "loss": 0.0264, | |
| "step": 463000 | |
| }, | |
| { | |
| "epoch": 47.52617532334223, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.012794570997357368, | |
| "eval_runtime": 22.5479, | |
| "eval_samples_per_second": 3.992, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 463000 | |
| }, | |
| { | |
| "epoch": 47.6288236501745, | |
| "grad_norm": 0.2703556418418884, | |
| "learning_rate": 2.3760543098127955e-05, | |
| "loss": 0.0264, | |
| "step": 464000 | |
| }, | |
| { | |
| "epoch": 47.6288236501745, | |
| "eval_cer": 0.00784952606635071, | |
| "eval_loss": 0.012676162645220757, | |
| "eval_runtime": 22.746, | |
| "eval_samples_per_second": 3.957, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 464000 | |
| }, | |
| { | |
| "epoch": 47.73147197700678, | |
| "grad_norm": 0.20613588392734528, | |
| "learning_rate": 2.2731948158815062e-05, | |
| "loss": 0.0267, | |
| "step": 465000 | |
| }, | |
| { | |
| "epoch": 47.73147197700678, | |
| "eval_cer": 0.00740521327014218, | |
| "eval_loss": 0.012650624848902225, | |
| "eval_runtime": 22.6051, | |
| "eval_samples_per_second": 3.981, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.032171581769437, | |
| "step": 465000 | |
| }, | |
| { | |
| "epoch": 47.83412030383905, | |
| "grad_norm": 0.19190002977848053, | |
| "learning_rate": 2.1703353219502162e-05, | |
| "loss": 0.0261, | |
| "step": 466000 | |
| }, | |
| { | |
| "epoch": 47.83412030383905, | |
| "eval_cer": 0.00740521327014218, | |
| "eval_loss": 0.012826332822442055, | |
| "eval_runtime": 22.6709, | |
| "eval_samples_per_second": 3.97, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03306523681858802, | |
| "step": 466000 | |
| }, | |
| { | |
| "epoch": 47.93676863067132, | |
| "grad_norm": 0.1679309457540512, | |
| "learning_rate": 2.0674758280189262e-05, | |
| "loss": 0.0262, | |
| "step": 467000 | |
| }, | |
| { | |
| "epoch": 47.93676863067132, | |
| "eval_cer": 0.006960900473933649, | |
| "eval_loss": 0.012559423223137856, | |
| "eval_runtime": 22.4179, | |
| "eval_samples_per_second": 4.015, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.029490616621983913, | |
| "step": 467000 | |
| }, | |
| { | |
| "epoch": 48.03941695750359, | |
| "grad_norm": 0.21128496527671814, | |
| "learning_rate": 1.9646163340876362e-05, | |
| "loss": 0.0264, | |
| "step": 468000 | |
| }, | |
| { | |
| "epoch": 48.03941695750359, | |
| "eval_cer": 0.0071090047393364926, | |
| "eval_loss": 0.01262712012976408, | |
| "eval_runtime": 22.5251, | |
| "eval_samples_per_second": 3.996, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 468000 | |
| }, | |
| { | |
| "epoch": 48.142065284335864, | |
| "grad_norm": 0.23150426149368286, | |
| "learning_rate": 1.8617568401563465e-05, | |
| "loss": 0.0263, | |
| "step": 469000 | |
| }, | |
| { | |
| "epoch": 48.142065284335864, | |
| "eval_cer": 0.0071090047393364926, | |
| "eval_loss": 0.012799305841326714, | |
| "eval_runtime": 22.402, | |
| "eval_samples_per_second": 4.018, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 469000 | |
| }, | |
| { | |
| "epoch": 48.24471361116814, | |
| "grad_norm": 0.23995615541934967, | |
| "learning_rate": 1.7588973462250565e-05, | |
| "loss": 0.0259, | |
| "step": 470000 | |
| }, | |
| { | |
| "epoch": 48.24471361116814, | |
| "eval_cer": 0.006960900473933649, | |
| "eval_loss": 0.012804466299712658, | |
| "eval_runtime": 22.5738, | |
| "eval_samples_per_second": 3.987, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.029490616621983913, | |
| "step": 470000 | |
| }, | |
| { | |
| "epoch": 48.34736193800041, | |
| "grad_norm": 0.2322542518377304, | |
| "learning_rate": 1.6560378522937665e-05, | |
| "loss": 0.0263, | |
| "step": 471000 | |
| }, | |
| { | |
| "epoch": 48.34736193800041, | |
| "eval_cer": 0.006960900473933649, | |
| "eval_loss": 0.012536253780126572, | |
| "eval_runtime": 22.4337, | |
| "eval_samples_per_second": 4.012, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.029490616621983913, | |
| "step": 471000 | |
| }, | |
| { | |
| "epoch": 48.450010264832684, | |
| "grad_norm": 0.19019187986850739, | |
| "learning_rate": 1.553178358362477e-05, | |
| "loss": 0.0258, | |
| "step": 472000 | |
| }, | |
| { | |
| "epoch": 48.450010264832684, | |
| "eval_cer": 0.0071090047393364926, | |
| "eval_loss": 0.012600379064679146, | |
| "eval_runtime": 22.4073, | |
| "eval_samples_per_second": 4.017, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 472000 | |
| }, | |
| { | |
| "epoch": 48.55265859166496, | |
| "grad_norm": 0.16290856897830963, | |
| "learning_rate": 1.450318864431187e-05, | |
| "loss": 0.0262, | |
| "step": 473000 | |
| }, | |
| { | |
| "epoch": 48.55265859166496, | |
| "eval_cer": 0.0071090047393364926, | |
| "eval_loss": 0.012459825724363327, | |
| "eval_runtime": 22.3028, | |
| "eval_samples_per_second": 4.035, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 473000 | |
| }, | |
| { | |
| "epoch": 48.65530691849723, | |
| "grad_norm": 0.20070821046829224, | |
| "learning_rate": 1.3474593704998972e-05, | |
| "loss": 0.0258, | |
| "step": 474000 | |
| }, | |
| { | |
| "epoch": 48.65530691849723, | |
| "eval_cer": 0.0071090047393364926, | |
| "eval_loss": 0.01252325065433979, | |
| "eval_runtime": 22.5396, | |
| "eval_samples_per_second": 3.993, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 474000 | |
| }, | |
| { | |
| "epoch": 48.7579552453295, | |
| "grad_norm": 0.21549555659294128, | |
| "learning_rate": 1.2445998765686073e-05, | |
| "loss": 0.026, | |
| "step": 475000 | |
| }, | |
| { | |
| "epoch": 48.7579552453295, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.0124691566452384, | |
| "eval_runtime": 22.5036, | |
| "eval_samples_per_second": 3.999, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 475000 | |
| }, | |
| { | |
| "epoch": 48.86060357216177, | |
| "grad_norm": 0.27654412388801575, | |
| "learning_rate": 1.1417403826373175e-05, | |
| "loss": 0.0265, | |
| "step": 476000 | |
| }, | |
| { | |
| "epoch": 48.86060357216177, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.012546148151159286, | |
| "eval_runtime": 22.6064, | |
| "eval_samples_per_second": 3.981, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 476000 | |
| }, | |
| { | |
| "epoch": 48.963251898994045, | |
| "grad_norm": 0.1903195083141327, | |
| "learning_rate": 1.0388808887060275e-05, | |
| "loss": 0.0262, | |
| "step": 477000 | |
| }, | |
| { | |
| "epoch": 48.963251898994045, | |
| "eval_cer": 0.00740521327014218, | |
| "eval_loss": 0.012493513524532318, | |
| "eval_runtime": 22.4828, | |
| "eval_samples_per_second": 4.003, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.032171581769437, | |
| "step": 477000 | |
| }, | |
| { | |
| "epoch": 49.06590022582632, | |
| "grad_norm": 0.2395348995923996, | |
| "learning_rate": 9.360213947747377e-06, | |
| "loss": 0.0261, | |
| "step": 478000 | |
| }, | |
| { | |
| "epoch": 49.06590022582632, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.012535948306322098, | |
| "eval_runtime": 22.8324, | |
| "eval_samples_per_second": 3.942, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 478000 | |
| }, | |
| { | |
| "epoch": 49.16854855265859, | |
| "grad_norm": 0.20346806943416595, | |
| "learning_rate": 8.33161900843448e-06, | |
| "loss": 0.0259, | |
| "step": 479000 | |
| }, | |
| { | |
| "epoch": 49.16854855265859, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.012490322813391685, | |
| "eval_runtime": 22.7561, | |
| "eval_samples_per_second": 3.955, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 479000 | |
| }, | |
| { | |
| "epoch": 49.271196879490866, | |
| "grad_norm": 0.18711692094802856, | |
| "learning_rate": 7.30302406912158e-06, | |
| "loss": 0.0261, | |
| "step": 480000 | |
| }, | |
| { | |
| "epoch": 49.271196879490866, | |
| "eval_cer": 0.0071090047393364926, | |
| "eval_loss": 0.01254476048052311, | |
| "eval_runtime": 22.575, | |
| "eval_samples_per_second": 3.987, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 480000 | |
| }, | |
| { | |
| "epoch": 49.37384520632314, | |
| "grad_norm": 0.19651705026626587, | |
| "learning_rate": 6.274429129808681e-06, | |
| "loss": 0.0257, | |
| "step": 481000 | |
| }, | |
| { | |
| "epoch": 49.37384520632314, | |
| "eval_cer": 0.0071090047393364926, | |
| "eval_loss": 0.012482382357120514, | |
| "eval_runtime": 22.6241, | |
| "eval_samples_per_second": 3.978, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 481000 | |
| }, | |
| { | |
| "epoch": 49.47649353315541, | |
| "grad_norm": 0.1909824162721634, | |
| "learning_rate": 5.245834190495783e-06, | |
| "loss": 0.0261, | |
| "step": 482000 | |
| }, | |
| { | |
| "epoch": 49.47649353315541, | |
| "eval_cer": 0.0071090047393364926, | |
| "eval_loss": 0.012513377703726292, | |
| "eval_runtime": 22.706, | |
| "eval_samples_per_second": 3.964, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 482000 | |
| }, | |
| { | |
| "epoch": 49.57914185998768, | |
| "grad_norm": 0.18398302793502808, | |
| "learning_rate": 4.217239251182884e-06, | |
| "loss": 0.0262, | |
| "step": 483000 | |
| }, | |
| { | |
| "epoch": 49.57914185998768, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.012494869530200958, | |
| "eval_runtime": 22.545, | |
| "eval_samples_per_second": 3.992, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 483000 | |
| }, | |
| { | |
| "epoch": 49.68179018681995, | |
| "grad_norm": 0.2315966933965683, | |
| "learning_rate": 3.1886443118699856e-06, | |
| "loss": 0.0255, | |
| "step": 484000 | |
| }, | |
| { | |
| "epoch": 49.68179018681995, | |
| "eval_cer": 0.007257109004739337, | |
| "eval_loss": 0.012507443316280842, | |
| "eval_runtime": 22.6822, | |
| "eval_samples_per_second": 3.968, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03127792672028597, | |
| "step": 484000 | |
| }, | |
| { | |
| "epoch": 49.784438513652226, | |
| "grad_norm": 0.16276974976062775, | |
| "learning_rate": 2.1600493725570872e-06, | |
| "loss": 0.0258, | |
| "step": 485000 | |
| }, | |
| { | |
| "epoch": 49.784438513652226, | |
| "eval_cer": 0.0071090047393364926, | |
| "eval_loss": 0.012460754252970219, | |
| "eval_runtime": 22.3925, | |
| "eval_samples_per_second": 4.019, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 485000 | |
| }, | |
| { | |
| "epoch": 49.8870868404845, | |
| "grad_norm": 0.16273947060108185, | |
| "learning_rate": 1.1314544332441885e-06, | |
| "loss": 0.0261, | |
| "step": 486000 | |
| }, | |
| { | |
| "epoch": 49.8870868404845, | |
| "eval_cer": 0.0071090047393364926, | |
| "eval_loss": 0.012488246895372868, | |
| "eval_runtime": 22.5881, | |
| "eval_samples_per_second": 3.984, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 486000 | |
| }, | |
| { | |
| "epoch": 49.98973516731677, | |
| "grad_norm": 0.2636996805667877, | |
| "learning_rate": 1.0285949393128985e-07, | |
| "loss": 0.0263, | |
| "step": 487000 | |
| }, | |
| { | |
| "epoch": 49.98973516731677, | |
| "eval_cer": 0.0071090047393364926, | |
| "eval_loss": 0.01250074990093708, | |
| "eval_runtime": 22.7073, | |
| "eval_samples_per_second": 3.963, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.03038427167113494, | |
| "step": 487000 | |
| } | |
| ], | |
| "logging_steps": 1000, | |
| "max_steps": 487100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 4000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.703420432200581e+17, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |