| { | |
| "best_metric": 0.02429259568452835, | |
| "best_model_checkpoint": "./phase3-t5/checkpoint-484000", | |
| "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": 0.2675953805446625, | |
| "learning_rate": 0.0005, | |
| "loss": 0.1517, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.10264832683227264, | |
| "eval_cer": 0.3379739336492891, | |
| "eval_loss": 0.09135649353265762, | |
| "eval_runtime": 61.2416, | |
| "eval_samples_per_second": 1.47, | |
| "eval_steps_per_second": 0.016, | |
| "eval_wer": 0.42716711349419123, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.20529665366454528, | |
| "grad_norm": 0.21362937986850739, | |
| "learning_rate": 0.0004989714050606871, | |
| "loss": 0.1197, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.20529665366454528, | |
| "eval_cer": 0.15506516587677724, | |
| "eval_loss": 0.08493143320083618, | |
| "eval_runtime": 60.3144, | |
| "eval_samples_per_second": 1.492, | |
| "eval_steps_per_second": 0.017, | |
| "eval_wer": 0.2645218945487042, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.3079449804968179, | |
| "grad_norm": 0.381354957818985, | |
| "learning_rate": 0.0004979428101213742, | |
| "loss": 0.1119, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.3079449804968179, | |
| "eval_cer": 0.13122037914691942, | |
| "eval_loss": 0.07921701669692993, | |
| "eval_runtime": 41.6322, | |
| "eval_samples_per_second": 2.162, | |
| "eval_steps_per_second": 0.024, | |
| "eval_wer": 0.23145665773011617, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.41059330732909055, | |
| "grad_norm": 0.21338549256324768, | |
| "learning_rate": 0.0004969142151820613, | |
| "loss": 0.1083, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.41059330732909055, | |
| "eval_cer": 0.05139218009478673, | |
| "eval_loss": 0.07416867464780807, | |
| "eval_runtime": 24.7028, | |
| "eval_samples_per_second": 3.643, | |
| "eval_steps_per_second": 0.04, | |
| "eval_wer": 0.14655942806076855, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5132416341613631, | |
| "grad_norm": 0.16844697296619415, | |
| "learning_rate": 0.0004958856202427484, | |
| "loss": 0.1054, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.5132416341613631, | |
| "eval_cer": 0.06916469194312796, | |
| "eval_loss": 0.0707884430885315, | |
| "eval_runtime": 32.9432, | |
| "eval_samples_per_second": 2.732, | |
| "eval_steps_per_second": 0.03, | |
| "eval_wer": 0.1572832886505809, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.6158899609936358, | |
| "grad_norm": 0.2411990612745285, | |
| "learning_rate": 0.0004948570253034355, | |
| "loss": 0.1033, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.6158899609936358, | |
| "eval_cer": 0.1552132701421801, | |
| "eval_loss": 0.07135774940252304, | |
| "eval_runtime": 46.0141, | |
| "eval_samples_per_second": 1.956, | |
| "eval_steps_per_second": 0.022, | |
| "eval_wer": 0.24128686327077747, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.7185382878259085, | |
| "grad_norm": 0.24947325885295868, | |
| "learning_rate": 0.0004938284303641226, | |
| "loss": 0.1017, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.7185382878259085, | |
| "eval_cer": 0.059537914691943125, | |
| "eval_loss": 0.06659836322069168, | |
| "eval_runtime": 45.8436, | |
| "eval_samples_per_second": 1.963, | |
| "eval_steps_per_second": 0.022, | |
| "eval_wer": 0.14298480786416443, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.8211866146581811, | |
| "grad_norm": 0.14766907691955566, | |
| "learning_rate": 0.0004927998354248098, | |
| "loss": 0.1006, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.8211866146581811, | |
| "eval_cer": 0.0727191943127962, | |
| "eval_loss": 0.06510724872350693, | |
| "eval_runtime": 51.2808, | |
| "eval_samples_per_second": 1.755, | |
| "eval_steps_per_second": 0.02, | |
| "eval_wer": 0.15638963360142985, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.9238349414904538, | |
| "grad_norm": 0.25097745656967163, | |
| "learning_rate": 0.0004917712404854969, | |
| "loss": 0.0982, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.9238349414904538, | |
| "eval_cer": 0.04576421800947867, | |
| "eval_loss": 0.06489837914705276, | |
| "eval_runtime": 45.8464, | |
| "eval_samples_per_second": 1.963, | |
| "eval_steps_per_second": 0.022, | |
| "eval_wer": 0.13047363717605004, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.0264832683227263, | |
| "grad_norm": 0.31185394525527954, | |
| "learning_rate": 0.000490742645546184, | |
| "loss": 0.0976, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.0264832683227263, | |
| "eval_cer": 0.050355450236966824, | |
| "eval_loss": 0.06680955737829208, | |
| "eval_runtime": 46.015, | |
| "eval_samples_per_second": 1.956, | |
| "eval_steps_per_second": 0.022, | |
| "eval_wer": 0.13494191242180517, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.129131595154999, | |
| "grad_norm": 0.2225043922662735, | |
| "learning_rate": 0.000489714050606871, | |
| "loss": 0.0944, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.129131595154999, | |
| "eval_cer": 0.03806279620853081, | |
| "eval_loss": 0.06565282493829727, | |
| "eval_runtime": 39.9826, | |
| "eval_samples_per_second": 2.251, | |
| "eval_steps_per_second": 0.025, | |
| "eval_wer": 0.1224307417336908, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.2317799219872716, | |
| "grad_norm": 0.2643093168735504, | |
| "learning_rate": 0.0004886854556675581, | |
| "loss": 0.095, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.2317799219872716, | |
| "eval_cer": 0.042950236966824644, | |
| "eval_loss": 0.06407604366540909, | |
| "eval_runtime": 38.3229, | |
| "eval_samples_per_second": 2.348, | |
| "eval_steps_per_second": 0.026, | |
| "eval_wer": 0.12779267202859695, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.3344282488195442, | |
| "grad_norm": 0.1704595983028412, | |
| "learning_rate": 0.00048765686072824524, | |
| "loss": 0.0934, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.3344282488195442, | |
| "eval_cer": 0.030953791469194313, | |
| "eval_loss": 0.06052744388580322, | |
| "eval_runtime": 43.0244, | |
| "eval_samples_per_second": 2.092, | |
| "eval_steps_per_second": 0.023, | |
| "eval_wer": 0.11081322609472744, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.437076575651817, | |
| "grad_norm": 0.28452351689338684, | |
| "learning_rate": 0.00048662826578893233, | |
| "loss": 0.0922, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.437076575651817, | |
| "eval_cer": 0.030065165876777253, | |
| "eval_loss": 0.06290669739246368, | |
| "eval_runtime": 38.566, | |
| "eval_samples_per_second": 2.334, | |
| "eval_steps_per_second": 0.026, | |
| "eval_wer": 0.1063449508489723, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.5397249024840896, | |
| "grad_norm": 0.18739064037799835, | |
| "learning_rate": 0.0004855996708496194, | |
| "loss": 0.0914, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.5397249024840896, | |
| "eval_cer": 0.09389810426540285, | |
| "eval_loss": 0.060126081109046936, | |
| "eval_runtime": 60.9362, | |
| "eval_samples_per_second": 1.477, | |
| "eval_steps_per_second": 0.016, | |
| "eval_wer": 0.1742627345844504, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.642373229316362, | |
| "grad_norm": 0.14926160871982574, | |
| "learning_rate": 0.0004845710759103065, | |
| "loss": 0.0923, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.642373229316362, | |
| "eval_cer": 0.04428317535545024, | |
| "eval_loss": 0.05910157784819603, | |
| "eval_runtime": 35.0608, | |
| "eval_samples_per_second": 2.567, | |
| "eval_steps_per_second": 0.029, | |
| "eval_wer": 0.11974977658623771, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.7450215561486346, | |
| "grad_norm": 0.15932171046733856, | |
| "learning_rate": 0.0004835424809709936, | |
| "loss": 0.0923, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.7450215561486346, | |
| "eval_cer": 0.08975118483412323, | |
| "eval_loss": 0.05897140130400658, | |
| "eval_runtime": 37.0041, | |
| "eval_samples_per_second": 2.432, | |
| "eval_steps_per_second": 0.027, | |
| "eval_wer": 0.16979445933869527, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.8476698829809073, | |
| "grad_norm": 0.23555859923362732, | |
| "learning_rate": 0.0004825138860316807, | |
| "loss": 0.0901, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.8476698829809073, | |
| "eval_cer": 0.033619668246445494, | |
| "eval_loss": 0.057106491178274155, | |
| "eval_runtime": 20.6425, | |
| "eval_samples_per_second": 4.36, | |
| "eval_steps_per_second": 0.048, | |
| "eval_wer": 0.10723860589812333, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.95031820981318, | |
| "grad_norm": 0.16271623969078064, | |
| "learning_rate": 0.00048148529109236785, | |
| "loss": 0.0896, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.95031820981318, | |
| "eval_cer": 0.039543838862559244, | |
| "eval_loss": 0.05746171995997429, | |
| "eval_runtime": 21.3164, | |
| "eval_samples_per_second": 4.222, | |
| "eval_steps_per_second": 0.047, | |
| "eval_wer": 0.11349419124218052, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.0529665366454526, | |
| "grad_norm": 0.225450336933136, | |
| "learning_rate": 0.00048045669615305494, | |
| "loss": 0.0879, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.0529665366454526, | |
| "eval_cer": 0.041321090047393365, | |
| "eval_loss": 0.05702373385429382, | |
| "eval_runtime": 22.0309, | |
| "eval_samples_per_second": 4.085, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.11170688114387846, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.1556148634777252, | |
| "grad_norm": 0.2128904014825821, | |
| "learning_rate": 0.00047942810121374204, | |
| "loss": 0.0875, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.1556148634777252, | |
| "eval_cer": 0.03154620853080569, | |
| "eval_loss": 0.05841705948114395, | |
| "eval_runtime": 22.0788, | |
| "eval_samples_per_second": 4.076, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.10187667560321716, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.258263190309998, | |
| "grad_norm": 0.24170714616775513, | |
| "learning_rate": 0.00047839950627442913, | |
| "loss": 0.0869, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 2.258263190309998, | |
| "eval_cer": 0.034212085308056875, | |
| "eval_loss": 0.057785358279943466, | |
| "eval_runtime": 22.1563, | |
| "eval_samples_per_second": 4.062, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.10902591599642537, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 2.3609115171422705, | |
| "grad_norm": 0.261454313993454, | |
| "learning_rate": 0.0004773709113351162, | |
| "loss": 0.0868, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 2.3609115171422705, | |
| "eval_cer": 0.03450829383886256, | |
| "eval_loss": 0.05649031326174736, | |
| "eval_runtime": 22.1442, | |
| "eval_samples_per_second": 4.064, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.10723860589812333, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 2.463559843974543, | |
| "grad_norm": 0.24116463959217072, | |
| "learning_rate": 0.0004763423163958033, | |
| "loss": 0.0861, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.463559843974543, | |
| "eval_cer": 0.03169431279620853, | |
| "eval_loss": 0.05647359788417816, | |
| "eval_runtime": 23.1195, | |
| "eval_samples_per_second": 3.893, | |
| "eval_steps_per_second": 0.043, | |
| "eval_wer": 0.10187667560321716, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.566208170806816, | |
| "grad_norm": 0.19247783720493317, | |
| "learning_rate": 0.0004753137214564904, | |
| "loss": 0.0861, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.566208170806816, | |
| "eval_cer": 0.027103080568720378, | |
| "eval_loss": 0.05473396182060242, | |
| "eval_runtime": 21.8207, | |
| "eval_samples_per_second": 4.125, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.10008936550491511, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.6688564976390885, | |
| "grad_norm": 0.20277945697307587, | |
| "learning_rate": 0.00047428512651717756, | |
| "loss": 0.0859, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.6688564976390885, | |
| "eval_cer": 0.028880331753554502, | |
| "eval_loss": 0.05520312860608101, | |
| "eval_runtime": 22.4109, | |
| "eval_samples_per_second": 4.016, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.10187667560321716, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.771504824471361, | |
| "grad_norm": 0.17546099424362183, | |
| "learning_rate": 0.00047325653157786465, | |
| "loss": 0.0853, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.771504824471361, | |
| "eval_cer": 0.027251184834123223, | |
| "eval_loss": 0.05452750623226166, | |
| "eval_runtime": 22.3468, | |
| "eval_samples_per_second": 4.027, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.09562109025915996, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.874153151303634, | |
| "grad_norm": 0.23071998357772827, | |
| "learning_rate": 0.00047222793663855174, | |
| "loss": 0.0857, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 2.874153151303634, | |
| "eval_cer": 0.027399289099526065, | |
| "eval_loss": 0.05141612887382507, | |
| "eval_runtime": 22.353, | |
| "eval_samples_per_second": 4.026, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.09651474530831099, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 2.9768014781359065, | |
| "grad_norm": 0.2768128216266632, | |
| "learning_rate": 0.00047119934169923884, | |
| "loss": 0.0844, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 2.9768014781359065, | |
| "eval_cer": 0.030805687203791468, | |
| "eval_loss": 0.053941383957862854, | |
| "eval_runtime": 22.2529, | |
| "eval_samples_per_second": 4.044, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.1063449508489723, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 3.079449804968179, | |
| "grad_norm": 0.21525971591472626, | |
| "learning_rate": 0.00047017074675992593, | |
| "loss": 0.0828, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 3.079449804968179, | |
| "eval_cer": 0.027843601895734597, | |
| "eval_loss": 0.05354895442724228, | |
| "eval_runtime": 22.123, | |
| "eval_samples_per_second": 4.068, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0938337801608579, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 3.1820981318004518, | |
| "grad_norm": 0.36673811078071594, | |
| "learning_rate": 0.000469142151820613, | |
| "loss": 0.0825, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 3.1820981318004518, | |
| "eval_cer": 0.02843601895734597, | |
| "eval_loss": 0.05313113331794739, | |
| "eval_runtime": 22.1652, | |
| "eval_samples_per_second": 4.06, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.10008936550491511, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 3.2847464586327244, | |
| "grad_norm": 0.20910222828388214, | |
| "learning_rate": 0.0004681135568813001, | |
| "loss": 0.0815, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 3.2847464586327244, | |
| "eval_cer": 0.04191350710900474, | |
| "eval_loss": 0.05280559882521629, | |
| "eval_runtime": 25.447, | |
| "eval_samples_per_second": 3.537, | |
| "eval_steps_per_second": 0.039, | |
| "eval_wer": 0.11349419124218052, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 3.387394785464997, | |
| "grad_norm": 0.18754634261131287, | |
| "learning_rate": 0.00046708496194198726, | |
| "loss": 0.0815, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 3.387394785464997, | |
| "eval_cer": 0.029472748815165876, | |
| "eval_loss": 0.05334876477718353, | |
| "eval_runtime": 21.9504, | |
| "eval_samples_per_second": 4.1, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.10545129579982127, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 3.4900431122972697, | |
| "grad_norm": 0.2228628695011139, | |
| "learning_rate": 0.00046605636700267436, | |
| "loss": 0.0821, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 3.4900431122972697, | |
| "eval_cer": 0.029917061611374408, | |
| "eval_loss": 0.05586351081728935, | |
| "eval_runtime": 21.6771, | |
| "eval_samples_per_second": 4.152, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.10545129579982127, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 3.592691439129542, | |
| "grad_norm": 0.13316944241523743, | |
| "learning_rate": 0.00046502777206336145, | |
| "loss": 0.0813, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 3.592691439129542, | |
| "eval_cer": 0.029324644549763034, | |
| "eval_loss": 0.05374256148934364, | |
| "eval_runtime": 21.9747, | |
| "eval_samples_per_second": 4.096, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.10277033065236818, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 3.6953397659618146, | |
| "grad_norm": 0.16168580949306488, | |
| "learning_rate": 0.00046399917712404854, | |
| "loss": 0.0812, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 3.6953397659618146, | |
| "eval_cer": 0.03450829383886256, | |
| "eval_loss": 0.05287105217576027, | |
| "eval_runtime": 24.1708, | |
| "eval_samples_per_second": 3.724, | |
| "eval_steps_per_second": 0.041, | |
| "eval_wer": 0.10008936550491511, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 3.7979880927940872, | |
| "grad_norm": 0.20095530152320862, | |
| "learning_rate": 0.00046297058218473564, | |
| "loss": 0.0814, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 3.7979880927940872, | |
| "eval_cer": 0.035545023696682464, | |
| "eval_loss": 0.05389825999736786, | |
| "eval_runtime": 21.7718, | |
| "eval_samples_per_second": 4.134, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.10723860589812333, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 3.90063641962636, | |
| "grad_norm": 0.19599634408950806, | |
| "learning_rate": 0.00046194198724542273, | |
| "loss": 0.0815, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 3.90063641962636, | |
| "eval_cer": 0.026954976303317536, | |
| "eval_loss": 0.05257488787174225, | |
| "eval_runtime": 22.0572, | |
| "eval_samples_per_second": 4.08, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.09651474530831099, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 4.0032847464586325, | |
| "grad_norm": 0.15417757630348206, | |
| "learning_rate": 0.0004609133923061098, | |
| "loss": 0.0806, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 4.0032847464586325, | |
| "eval_cer": 0.027695497630331755, | |
| "eval_loss": 0.05188766494393349, | |
| "eval_runtime": 22.1269, | |
| "eval_samples_per_second": 4.067, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.09830205540661305, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 4.105933073290905, | |
| "grad_norm": 0.26163730025291443, | |
| "learning_rate": 0.00045988479736679697, | |
| "loss": 0.0794, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 4.105933073290905, | |
| "eval_cer": 0.025622037914691944, | |
| "eval_loss": 0.048877034336328506, | |
| "eval_runtime": 22.2755, | |
| "eval_samples_per_second": 4.04, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08847184986595175, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 4.208581400123178, | |
| "grad_norm": 0.20826220512390137, | |
| "learning_rate": 0.00045885620242748406, | |
| "loss": 0.0785, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 4.208581400123178, | |
| "eval_cer": 0.025622037914691944, | |
| "eval_loss": 0.05260869115591049, | |
| "eval_runtime": 22.0992, | |
| "eval_samples_per_second": 4.073, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.09204647006255585, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 4.3112297269554505, | |
| "grad_norm": 0.3389241695404053, | |
| "learning_rate": 0.00045782760748817116, | |
| "loss": 0.0782, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 4.3112297269554505, | |
| "eval_cer": 0.026954976303317536, | |
| "eval_loss": 0.04903939738869667, | |
| "eval_runtime": 22.5525, | |
| "eval_samples_per_second": 3.991, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.09204647006255585, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 4.413878053787723, | |
| "grad_norm": 0.29210948944091797, | |
| "learning_rate": 0.00045679901254885825, | |
| "loss": 0.0779, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 4.413878053787723, | |
| "eval_cer": 0.02754739336492891, | |
| "eval_loss": 0.04861212149262428, | |
| "eval_runtime": 22.0013, | |
| "eval_samples_per_second": 4.091, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.09204647006255585, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 4.516526380619996, | |
| "grad_norm": 0.19757212698459625, | |
| "learning_rate": 0.00045577041760954534, | |
| "loss": 0.0786, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 4.516526380619996, | |
| "eval_cer": 0.023992890995260665, | |
| "eval_loss": 0.04914968088269234, | |
| "eval_runtime": 22.6286, | |
| "eval_samples_per_second": 3.977, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.08936550491510277, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 4.619174707452268, | |
| "grad_norm": 0.23518621921539307, | |
| "learning_rate": 0.00045474182267023244, | |
| "loss": 0.0778, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 4.619174707452268, | |
| "eval_cer": 0.025770142180094786, | |
| "eval_loss": 0.04797298088669777, | |
| "eval_runtime": 22.0387, | |
| "eval_samples_per_second": 4.084, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08847184986595175, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 4.721823034284541, | |
| "grad_norm": 0.18045727908611298, | |
| "learning_rate": 0.00045371322773091953, | |
| "loss": 0.0782, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 4.721823034284541, | |
| "eval_cer": 0.023252369668246446, | |
| "eval_loss": 0.049161382019519806, | |
| "eval_runtime": 22.0981, | |
| "eval_samples_per_second": 4.073, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08579088471849866, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 4.824471361116814, | |
| "grad_norm": 0.2216973602771759, | |
| "learning_rate": 0.0004526846327916067, | |
| "loss": 0.0778, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 4.824471361116814, | |
| "eval_cer": 0.027103080568720378, | |
| "eval_loss": 0.04854327812790871, | |
| "eval_runtime": 22.1254, | |
| "eval_samples_per_second": 4.068, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.09204647006255585, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 4.927119687949086, | |
| "grad_norm": 0.21196693181991577, | |
| "learning_rate": 0.00045165603785229377, | |
| "loss": 0.078, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 4.927119687949086, | |
| "eval_cer": 0.0254739336492891, | |
| "eval_loss": 0.04801648482680321, | |
| "eval_runtime": 22.3465, | |
| "eval_samples_per_second": 4.027, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08847184986595175, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 5.029768014781359, | |
| "grad_norm": 0.22312557697296143, | |
| "learning_rate": 0.00045062744291298086, | |
| "loss": 0.0771, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 5.029768014781359, | |
| "eval_cer": 0.027843601895734597, | |
| "eval_loss": 0.04988682270050049, | |
| "eval_runtime": 22.2617, | |
| "eval_samples_per_second": 4.043, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0938337801608579, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 5.132416341613632, | |
| "grad_norm": 0.22252117097377777, | |
| "learning_rate": 0.00044959884797366796, | |
| "loss": 0.0753, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 5.132416341613632, | |
| "eval_cer": 0.024140995260663507, | |
| "eval_loss": 0.04682554677128792, | |
| "eval_runtime": 22.1777, | |
| "eval_samples_per_second": 4.058, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08936550491510277, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 5.235064668445904, | |
| "grad_norm": 0.15847323834896088, | |
| "learning_rate": 0.00044857025303435505, | |
| "loss": 0.0756, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 5.235064668445904, | |
| "eval_cer": 0.02502962085308057, | |
| "eval_loss": 0.047456566244363785, | |
| "eval_runtime": 22.3306, | |
| "eval_samples_per_second": 4.03, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08847184986595175, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 5.337712995278177, | |
| "grad_norm": 0.2008858174085617, | |
| "learning_rate": 0.00044754165809504214, | |
| "loss": 0.0763, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 5.337712995278177, | |
| "eval_cer": 0.021178909952606635, | |
| "eval_loss": 0.0479045994579792, | |
| "eval_runtime": 22.2995, | |
| "eval_samples_per_second": 4.036, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0777479892761394, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 5.44036132211045, | |
| "grad_norm": 0.31161248683929443, | |
| "learning_rate": 0.00044651306315572923, | |
| "loss": 0.0761, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 5.44036132211045, | |
| "eval_cer": 0.026066350710900472, | |
| "eval_loss": 0.047916192561388016, | |
| "eval_runtime": 22.0152, | |
| "eval_samples_per_second": 4.088, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08847184986595175, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 5.543009648942722, | |
| "grad_norm": 0.19102248549461365, | |
| "learning_rate": 0.0004454844682164164, | |
| "loss": 0.0756, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 5.543009648942722, | |
| "eval_cer": 0.023252369668246446, | |
| "eval_loss": 0.04646703228354454, | |
| "eval_runtime": 22.0815, | |
| "eval_samples_per_second": 4.076, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08668453976764968, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 5.645657975774995, | |
| "grad_norm": 0.3801836669445038, | |
| "learning_rate": 0.0004444558732771035, | |
| "loss": 0.0758, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 5.645657975774995, | |
| "eval_cer": 0.025770142180094786, | |
| "eval_loss": 0.04690609872341156, | |
| "eval_runtime": 22.1797, | |
| "eval_samples_per_second": 4.058, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.09294012511170688, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 5.748306302607268, | |
| "grad_norm": 0.24007199704647064, | |
| "learning_rate": 0.00044342727833779057, | |
| "loss": 0.0751, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 5.748306302607268, | |
| "eval_cer": 0.026066350710900472, | |
| "eval_loss": 0.04649204760789871, | |
| "eval_runtime": 22.2606, | |
| "eval_samples_per_second": 4.043, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0902591599642538, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 5.85095462943954, | |
| "grad_norm": 0.177778959274292, | |
| "learning_rate": 0.00044239868339847766, | |
| "loss": 0.0751, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 5.85095462943954, | |
| "eval_cer": 0.02665876777251185, | |
| "eval_loss": 0.04591357707977295, | |
| "eval_runtime": 22.0753, | |
| "eval_samples_per_second": 4.077, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.09294012511170688, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 5.953602956271813, | |
| "grad_norm": 0.14689143002033234, | |
| "learning_rate": 0.00044137008845916475, | |
| "loss": 0.0754, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 5.953602956271813, | |
| "eval_cer": 0.025622037914691944, | |
| "eval_loss": 0.04754678159952164, | |
| "eval_runtime": 22.1351, | |
| "eval_samples_per_second": 4.066, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.09294012511170688, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 6.056251283104086, | |
| "grad_norm": 0.1848069429397583, | |
| "learning_rate": 0.00044034149351985185, | |
| "loss": 0.0738, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 6.056251283104086, | |
| "eval_cer": 0.027103080568720378, | |
| "eval_loss": 0.04577142372727394, | |
| "eval_runtime": 22.1342, | |
| "eval_samples_per_second": 4.066, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08936550491510277, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 6.158899609936358, | |
| "grad_norm": 0.22594769299030304, | |
| "learning_rate": 0.00043931289858053894, | |
| "loss": 0.0735, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 6.158899609936358, | |
| "eval_cer": 0.025177725118483412, | |
| "eval_loss": 0.04629523307085037, | |
| "eval_runtime": 22.2033, | |
| "eval_samples_per_second": 4.053, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08847184986595175, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 6.261547936768631, | |
| "grad_norm": 0.20947369933128357, | |
| "learning_rate": 0.0004382843036412261, | |
| "loss": 0.0734, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 6.261547936768631, | |
| "eval_cer": 0.026510663507109004, | |
| "eval_loss": 0.04853444546461105, | |
| "eval_runtime": 21.9491, | |
| "eval_samples_per_second": 4.1, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.09204647006255585, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 6.3641962636009035, | |
| "grad_norm": 0.2018922120332718, | |
| "learning_rate": 0.0004372557087019132, | |
| "loss": 0.0734, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 6.3641962636009035, | |
| "eval_cer": 0.022808056872037914, | |
| "eval_loss": 0.04681561887264252, | |
| "eval_runtime": 22.0488, | |
| "eval_samples_per_second": 4.082, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07864164432529044, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 6.466844590433176, | |
| "grad_norm": 0.2090204358100891, | |
| "learning_rate": 0.0004362271137626003, | |
| "loss": 0.0737, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 6.466844590433176, | |
| "eval_cer": 0.03865521327014218, | |
| "eval_loss": 0.046171579509973526, | |
| "eval_runtime": 24.3572, | |
| "eval_samples_per_second": 3.695, | |
| "eval_steps_per_second": 0.041, | |
| "eval_wer": 0.09651474530831099, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 6.569492917265449, | |
| "grad_norm": 0.21769754588603973, | |
| "learning_rate": 0.0004351985188232874, | |
| "loss": 0.0726, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 6.569492917265449, | |
| "eval_cer": 0.0254739336492891, | |
| "eval_loss": 0.047106679528951645, | |
| "eval_runtime": 21.6323, | |
| "eval_samples_per_second": 4.16, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.08936550491510277, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 6.6721412440977215, | |
| "grad_norm": 0.17357899248600006, | |
| "learning_rate": 0.0004341699238839745, | |
| "loss": 0.0734, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 6.6721412440977215, | |
| "eval_cer": 0.026214454976303318, | |
| "eval_loss": 0.04763852432370186, | |
| "eval_runtime": 21.7494, | |
| "eval_samples_per_second": 4.138, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.08757819481680071, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 6.774789570929994, | |
| "grad_norm": 0.23694564402103424, | |
| "learning_rate": 0.0004331413289446616, | |
| "loss": 0.0737, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 6.774789570929994, | |
| "eval_cer": 0.02384478672985782, | |
| "eval_loss": 0.04740356281399727, | |
| "eval_runtime": 21.8747, | |
| "eval_samples_per_second": 4.114, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.0840035746201966, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 6.877437897762267, | |
| "grad_norm": 0.24957220256328583, | |
| "learning_rate": 0.00043211273400534876, | |
| "loss": 0.0725, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 6.877437897762267, | |
| "eval_cer": 0.023548578199052133, | |
| "eval_loss": 0.04718530550599098, | |
| "eval_runtime": 22.1234, | |
| "eval_samples_per_second": 4.068, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08132260947274352, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 6.980086224594539, | |
| "grad_norm": 0.20797890424728394, | |
| "learning_rate": 0.00043108413906603585, | |
| "loss": 0.0728, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 6.980086224594539, | |
| "eval_cer": 0.02384478672985782, | |
| "eval_loss": 0.0460374690592289, | |
| "eval_runtime": 22.126, | |
| "eval_samples_per_second": 4.068, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08042895442359249, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 7.082734551426812, | |
| "grad_norm": 0.21543003618717194, | |
| "learning_rate": 0.00043005554412672294, | |
| "loss": 0.0714, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 7.082734551426812, | |
| "eval_cer": 0.024881516587677725, | |
| "eval_loss": 0.04654213413596153, | |
| "eval_runtime": 22.1075, | |
| "eval_samples_per_second": 4.071, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08579088471849866, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 7.185382878259085, | |
| "grad_norm": 0.2328251153230667, | |
| "learning_rate": 0.00042902694918741004, | |
| "loss": 0.0711, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 7.185382878259085, | |
| "eval_cer": 0.02458530805687204, | |
| "eval_loss": 0.0458548367023468, | |
| "eval_runtime": 22.1559, | |
| "eval_samples_per_second": 4.062, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08489722966934764, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 7.288031205091357, | |
| "grad_norm": 0.2861992418766022, | |
| "learning_rate": 0.00042799835424809713, | |
| "loss": 0.0711, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 7.288031205091357, | |
| "eval_cer": 0.0254739336492891, | |
| "eval_loss": 0.04648789018392563, | |
| "eval_runtime": 22.2028, | |
| "eval_samples_per_second": 4.054, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08489722966934764, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 7.39067953192363, | |
| "grad_norm": 0.21449844539165497, | |
| "learning_rate": 0.0004269697593087842, | |
| "loss": 0.0709, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 7.39067953192363, | |
| "eval_cer": 0.023548578199052133, | |
| "eval_loss": 0.04555143415927887, | |
| "eval_runtime": 22.2316, | |
| "eval_samples_per_second": 4.048, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0777479892761394, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 7.493327858755903, | |
| "grad_norm": 0.21393579244613647, | |
| "learning_rate": 0.0004259411643694713, | |
| "loss": 0.0719, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 7.493327858755903, | |
| "eval_cer": 0.023548578199052133, | |
| "eval_loss": 0.04599784314632416, | |
| "eval_runtime": 22.1857, | |
| "eval_samples_per_second": 4.057, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0777479892761394, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 7.5959761855881744, | |
| "grad_norm": 0.26174065470695496, | |
| "learning_rate": 0.00042491256943015846, | |
| "loss": 0.0707, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 7.5959761855881744, | |
| "eval_cer": 0.026362559241706163, | |
| "eval_loss": 0.046083446592092514, | |
| "eval_runtime": 22.1298, | |
| "eval_samples_per_second": 4.067, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08668453976764968, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 7.698624512420448, | |
| "grad_norm": 0.23440520465373993, | |
| "learning_rate": 0.00042388397449084556, | |
| "loss": 0.0707, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 7.698624512420448, | |
| "eval_cer": 0.02502962085308057, | |
| "eval_loss": 0.046040162444114685, | |
| "eval_runtime": 22.0318, | |
| "eval_samples_per_second": 4.085, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08579088471849866, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 7.80127283925272, | |
| "grad_norm": 0.210636705160141, | |
| "learning_rate": 0.00042285537955153265, | |
| "loss": 0.0717, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 7.80127283925272, | |
| "eval_cer": 0.023252369668246446, | |
| "eval_loss": 0.044275738298892975, | |
| "eval_runtime": 22.1574, | |
| "eval_samples_per_second": 4.062, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08132260947274352, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 7.903921166084993, | |
| "grad_norm": 0.19636699557304382, | |
| "learning_rate": 0.00042182678461221974, | |
| "loss": 0.0713, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 7.903921166084993, | |
| "eval_cer": 0.025177725118483412, | |
| "eval_loss": 0.04342404752969742, | |
| "eval_runtime": 22.2249, | |
| "eval_samples_per_second": 4.05, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08221626452189455, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 8.006569492917265, | |
| "grad_norm": 0.23213474452495575, | |
| "learning_rate": 0.00042079818967290683, | |
| "loss": 0.0703, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 8.006569492917265, | |
| "eval_cer": 0.02428909952606635, | |
| "eval_loss": 0.04461972787976265, | |
| "eval_runtime": 22.2476, | |
| "eval_samples_per_second": 4.045, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08489722966934764, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 8.109217819749539, | |
| "grad_norm": 0.5032120943069458, | |
| "learning_rate": 0.00041976959473359393, | |
| "loss": 0.0692, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 8.109217819749539, | |
| "eval_cer": 0.025177725118483412, | |
| "eval_loss": 0.0442008450627327, | |
| "eval_runtime": 22.2279, | |
| "eval_samples_per_second": 4.049, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08310991957104558, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 8.21186614658181, | |
| "grad_norm": 0.2267637550830841, | |
| "learning_rate": 0.000418740999794281, | |
| "loss": 0.0694, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 8.21186614658181, | |
| "eval_cer": 0.02473341232227488, | |
| "eval_loss": 0.041503287851810455, | |
| "eval_runtime": 22.2205, | |
| "eval_samples_per_second": 4.05, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08310991957104558, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 8.314514473414084, | |
| "grad_norm": 0.22922959923744202, | |
| "learning_rate": 0.00041771240485496817, | |
| "loss": 0.0698, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 8.314514473414084, | |
| "eval_cer": 0.02132701421800948, | |
| "eval_loss": 0.0416925847530365, | |
| "eval_runtime": 22.7674, | |
| "eval_samples_per_second": 3.953, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.07506702412868632, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 8.417162800246356, | |
| "grad_norm": 0.2807318866252899, | |
| "learning_rate": 0.00041668380991565526, | |
| "loss": 0.0687, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 8.417162800246356, | |
| "eval_cer": 0.021475118483412322, | |
| "eval_loss": 0.042673755437135696, | |
| "eval_runtime": 21.9781, | |
| "eval_samples_per_second": 4.095, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07327971403038427, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 8.51981112707863, | |
| "grad_norm": 0.19773901998996735, | |
| "learning_rate": 0.00041565521497634235, | |
| "loss": 0.0697, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 8.51981112707863, | |
| "eval_cer": 0.025770142180094786, | |
| "eval_loss": 0.04353732243180275, | |
| "eval_runtime": 21.9415, | |
| "eval_samples_per_second": 4.102, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.08668453976764968, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 8.622459453910901, | |
| "grad_norm": 0.2049770951271057, | |
| "learning_rate": 0.00041462662003702945, | |
| "loss": 0.0689, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 8.622459453910901, | |
| "eval_cer": 0.02295616113744076, | |
| "eval_loss": 0.044080935418605804, | |
| "eval_runtime": 22.0767, | |
| "eval_samples_per_second": 4.077, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08042895442359249, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 8.725107780743175, | |
| "grad_norm": 0.19069017469882965, | |
| "learning_rate": 0.00041359802509771654, | |
| "loss": 0.0692, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 8.725107780743175, | |
| "eval_cer": 0.022363744075829382, | |
| "eval_loss": 0.042362380772829056, | |
| "eval_runtime": 22.0792, | |
| "eval_samples_per_second": 4.076, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07864164432529044, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 8.827756107575446, | |
| "grad_norm": 0.26284459233283997, | |
| "learning_rate": 0.00041256943015840363, | |
| "loss": 0.0694, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 8.827756107575446, | |
| "eval_cer": 0.08486374407582939, | |
| "eval_loss": 0.04263956472277641, | |
| "eval_runtime": 39.7879, | |
| "eval_samples_per_second": 2.262, | |
| "eval_steps_per_second": 0.025, | |
| "eval_wer": 0.13762287756925826, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 8.93040443440772, | |
| "grad_norm": 0.12760530412197113, | |
| "learning_rate": 0.00041154083521909073, | |
| "loss": 0.0691, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 8.93040443440772, | |
| "eval_cer": 0.02502962085308057, | |
| "eval_loss": 0.043798867613077164, | |
| "eval_runtime": 23.4939, | |
| "eval_samples_per_second": 3.831, | |
| "eval_steps_per_second": 0.043, | |
| "eval_wer": 0.08936550491510277, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 9.033052761239992, | |
| "grad_norm": 0.3002149760723114, | |
| "learning_rate": 0.0004105122402797779, | |
| "loss": 0.0689, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 9.033052761239992, | |
| "eval_cer": 0.03036137440758294, | |
| "eval_loss": 0.04081055149435997, | |
| "eval_runtime": 28.0519, | |
| "eval_samples_per_second": 3.208, | |
| "eval_steps_per_second": 0.036, | |
| "eval_wer": 0.08489722966934764, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 9.135701088072265, | |
| "grad_norm": 0.3071158230304718, | |
| "learning_rate": 0.00040948364534046497, | |
| "loss": 0.0672, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 9.135701088072265, | |
| "eval_cer": 0.028880331753554502, | |
| "eval_loss": 0.0426529198884964, | |
| "eval_runtime": 21.5055, | |
| "eval_samples_per_second": 4.185, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.08042895442359249, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 9.238349414904537, | |
| "grad_norm": 0.2852329909801483, | |
| "learning_rate": 0.00040845505040115206, | |
| "loss": 0.0672, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 9.238349414904537, | |
| "eval_cer": 0.023252369668246446, | |
| "eval_loss": 0.04157470539212227, | |
| "eval_runtime": 20.7416, | |
| "eval_samples_per_second": 4.339, | |
| "eval_steps_per_second": 0.048, | |
| "eval_wer": 0.0777479892761394, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 9.34099774173681, | |
| "grad_norm": 0.2886292338371277, | |
| "learning_rate": 0.00040742645546183915, | |
| "loss": 0.0677, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 9.34099774173681, | |
| "eval_cer": 0.029768957345971563, | |
| "eval_loss": 0.04205571115016937, | |
| "eval_runtime": 21.8561, | |
| "eval_samples_per_second": 4.118, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.08310991957104558, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 9.443646068569082, | |
| "grad_norm": 0.20502915978431702, | |
| "learning_rate": 0.00040639786052252625, | |
| "loss": 0.068, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 9.443646068569082, | |
| "eval_cer": 0.050651658767772514, | |
| "eval_loss": 0.04323223605751991, | |
| "eval_runtime": 34.4159, | |
| "eval_samples_per_second": 2.615, | |
| "eval_steps_per_second": 0.029, | |
| "eval_wer": 0.10723860589812333, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 9.546294395401356, | |
| "grad_norm": 0.31369203329086304, | |
| "learning_rate": 0.00040536926558321334, | |
| "loss": 0.0675, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 9.546294395401356, | |
| "eval_cer": 0.023548578199052133, | |
| "eval_loss": 0.040974486619234085, | |
| "eval_runtime": 21.5483, | |
| "eval_samples_per_second": 4.177, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.08042895442359249, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 9.648942722233627, | |
| "grad_norm": 0.2164764255285263, | |
| "learning_rate": 0.00040434067064390043, | |
| "loss": 0.0674, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 9.648942722233627, | |
| "eval_cer": 0.021623222748815167, | |
| "eval_loss": 0.04193877801299095, | |
| "eval_runtime": 20.5784, | |
| "eval_samples_per_second": 4.374, | |
| "eval_steps_per_second": 0.049, | |
| "eval_wer": 0.0741733690795353, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 9.751591049065901, | |
| "grad_norm": 0.20598456263542175, | |
| "learning_rate": 0.0004033120757045876, | |
| "loss": 0.0683, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 9.751591049065901, | |
| "eval_cer": 0.021623222748815167, | |
| "eval_loss": 0.04229186475276947, | |
| "eval_runtime": 21.7515, | |
| "eval_samples_per_second": 4.138, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.07685433422698838, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 9.854239375898173, | |
| "grad_norm": 0.17217130959033966, | |
| "learning_rate": 0.0004022834807652747, | |
| "loss": 0.0679, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 9.854239375898173, | |
| "eval_cer": 0.02221563981042654, | |
| "eval_loss": 0.04353512451052666, | |
| "eval_runtime": 21.856, | |
| "eval_samples_per_second": 4.118, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.08042895442359249, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 9.956887702730446, | |
| "grad_norm": 0.30278000235557556, | |
| "learning_rate": 0.00040125488582596177, | |
| "loss": 0.0676, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 9.956887702730446, | |
| "eval_cer": 0.021623222748815167, | |
| "eval_loss": 0.04357661306858063, | |
| "eval_runtime": 21.958, | |
| "eval_samples_per_second": 4.099, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.07685433422698838, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 10.059536029562718, | |
| "grad_norm": 0.17858092486858368, | |
| "learning_rate": 0.00040022629088664886, | |
| "loss": 0.0659, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 10.059536029562718, | |
| "eval_cer": 0.02428909952606635, | |
| "eval_loss": 0.04361404851078987, | |
| "eval_runtime": 22.0796, | |
| "eval_samples_per_second": 4.076, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.08221626452189455, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 10.162184356394992, | |
| "grad_norm": 0.16541603207588196, | |
| "learning_rate": 0.00039919769594733595, | |
| "loss": 0.066, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 10.162184356394992, | |
| "eval_cer": 0.019845971563981043, | |
| "eval_loss": 0.04267050698399544, | |
| "eval_runtime": 21.9748, | |
| "eval_samples_per_second": 4.096, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.07149240393208221, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 10.264832683227263, | |
| "grad_norm": 0.16135546565055847, | |
| "learning_rate": 0.00039816910100802305, | |
| "loss": 0.066, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 10.264832683227263, | |
| "eval_cer": 0.02458530805687204, | |
| "eval_loss": 0.043835073709487915, | |
| "eval_runtime": 22.2259, | |
| "eval_samples_per_second": 4.049, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0840035746201966, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 10.367481010059535, | |
| "grad_norm": 0.3332918882369995, | |
| "learning_rate": 0.00039714050606871014, | |
| "loss": 0.0667, | |
| "step": 101000 | |
| }, | |
| { | |
| "epoch": 10.367481010059535, | |
| "eval_cer": 0.021919431279620854, | |
| "eval_loss": 0.04083102196455002, | |
| "eval_runtime": 22.2761, | |
| "eval_samples_per_second": 4.04, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0777479892761394, | |
| "step": 101000 | |
| }, | |
| { | |
| "epoch": 10.470129336891809, | |
| "grad_norm": 0.2997848093509674, | |
| "learning_rate": 0.0003961119111293973, | |
| "loss": 0.0659, | |
| "step": 102000 | |
| }, | |
| { | |
| "epoch": 10.470129336891809, | |
| "eval_cer": 0.021475118483412322, | |
| "eval_loss": 0.04166368022561073, | |
| "eval_runtime": 22.1637, | |
| "eval_samples_per_second": 4.061, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07327971403038427, | |
| "step": 102000 | |
| }, | |
| { | |
| "epoch": 10.572777663724082, | |
| "grad_norm": 0.2358590066432953, | |
| "learning_rate": 0.0003950833161900844, | |
| "loss": 0.0661, | |
| "step": 103000 | |
| }, | |
| { | |
| "epoch": 10.572777663724082, | |
| "eval_cer": 0.018957345971563982, | |
| "eval_loss": 0.04044094681739807, | |
| "eval_runtime": 22.1865, | |
| "eval_samples_per_second": 4.057, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06523681858802502, | |
| "step": 103000 | |
| }, | |
| { | |
| "epoch": 10.675425990556354, | |
| "grad_norm": 0.2147412747144699, | |
| "learning_rate": 0.0003940547212507715, | |
| "loss": 0.0662, | |
| "step": 104000 | |
| }, | |
| { | |
| "epoch": 10.675425990556354, | |
| "eval_cer": 0.022067535545023696, | |
| "eval_loss": 0.042136672884225845, | |
| "eval_runtime": 22.3442, | |
| "eval_samples_per_second": 4.028, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0741733690795353, | |
| "step": 104000 | |
| }, | |
| { | |
| "epoch": 10.778074317388626, | |
| "grad_norm": 0.21466469764709473, | |
| "learning_rate": 0.00039302612631145857, | |
| "loss": 0.066, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 10.778074317388626, | |
| "eval_cer": 0.020290284360189575, | |
| "eval_loss": 0.04227345809340477, | |
| "eval_runtime": 22.2243, | |
| "eval_samples_per_second": 4.05, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06881143878462913, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 10.8807226442209, | |
| "grad_norm": 0.20175763964653015, | |
| "learning_rate": 0.00039199753137214566, | |
| "loss": 0.0665, | |
| "step": 106000 | |
| }, | |
| { | |
| "epoch": 10.8807226442209, | |
| "eval_cer": 0.021919431279620854, | |
| "eval_loss": 0.04261546581983566, | |
| "eval_runtime": 22.0605, | |
| "eval_samples_per_second": 4.08, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07327971403038427, | |
| "step": 106000 | |
| }, | |
| { | |
| "epoch": 10.983370971053171, | |
| "grad_norm": 0.3140750229358673, | |
| "learning_rate": 0.00039096893643283275, | |
| "loss": 0.066, | |
| "step": 107000 | |
| }, | |
| { | |
| "epoch": 10.983370971053171, | |
| "eval_cer": 0.018957345971563982, | |
| "eval_loss": 0.040347784757614136, | |
| "eval_runtime": 23.0302, | |
| "eval_samples_per_second": 3.908, | |
| "eval_steps_per_second": 0.043, | |
| "eval_wer": 0.06523681858802502, | |
| "step": 107000 | |
| }, | |
| { | |
| "epoch": 11.086019297885445, | |
| "grad_norm": 0.2863214313983917, | |
| "learning_rate": 0.00038994034149351985, | |
| "loss": 0.0643, | |
| "step": 108000 | |
| }, | |
| { | |
| "epoch": 11.086019297885445, | |
| "eval_cer": 0.019105450236966824, | |
| "eval_loss": 0.040664974600076675, | |
| "eval_runtime": 22.169, | |
| "eval_samples_per_second": 4.06, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07149240393208221, | |
| "step": 108000 | |
| }, | |
| { | |
| "epoch": 11.188667624717716, | |
| "grad_norm": 0.1757292002439499, | |
| "learning_rate": 0.000388911746554207, | |
| "loss": 0.0644, | |
| "step": 109000 | |
| }, | |
| { | |
| "epoch": 11.188667624717716, | |
| "eval_cer": 0.01762440758293839, | |
| "eval_loss": 0.03967958316206932, | |
| "eval_runtime": 22.6062, | |
| "eval_samples_per_second": 3.981, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.06344950848972297, | |
| "step": 109000 | |
| }, | |
| { | |
| "epoch": 11.29131595154999, | |
| "grad_norm": 0.25020548701286316, | |
| "learning_rate": 0.0003878831516148941, | |
| "loss": 0.0646, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 11.29131595154999, | |
| "eval_cer": 0.019697867298578198, | |
| "eval_loss": 0.0391419492661953, | |
| "eval_runtime": 21.9075, | |
| "eval_samples_per_second": 4.108, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.06881143878462913, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 11.393964278382262, | |
| "grad_norm": 0.2499699741601944, | |
| "learning_rate": 0.0003868545566755812, | |
| "loss": 0.0643, | |
| "step": 111000 | |
| }, | |
| { | |
| "epoch": 11.393964278382262, | |
| "eval_cer": 0.02058649289099526, | |
| "eval_loss": 0.04037711024284363, | |
| "eval_runtime": 22.1604, | |
| "eval_samples_per_second": 4.061, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0741733690795353, | |
| "step": 111000 | |
| }, | |
| { | |
| "epoch": 11.496612605214535, | |
| "grad_norm": 0.16554132103919983, | |
| "learning_rate": 0.0003858259617362683, | |
| "loss": 0.0649, | |
| "step": 112000 | |
| }, | |
| { | |
| "epoch": 11.496612605214535, | |
| "eval_cer": 0.019845971563981043, | |
| "eval_loss": 0.04021435230970383, | |
| "eval_runtime": 22.1316, | |
| "eval_samples_per_second": 4.067, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07327971403038427, | |
| "step": 112000 | |
| }, | |
| { | |
| "epoch": 11.599260932046807, | |
| "grad_norm": 0.18748946487903595, | |
| "learning_rate": 0.00038479736679695537, | |
| "loss": 0.0655, | |
| "step": 113000 | |
| }, | |
| { | |
| "epoch": 11.599260932046807, | |
| "eval_cer": 0.022511848341232227, | |
| "eval_loss": 0.040546808391809464, | |
| "eval_runtime": 22.1622, | |
| "eval_samples_per_second": 4.061, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07596067917783736, | |
| "step": 113000 | |
| }, | |
| { | |
| "epoch": 11.70190925887908, | |
| "grad_norm": 0.4686923623085022, | |
| "learning_rate": 0.00038376877185764246, | |
| "loss": 0.0647, | |
| "step": 114000 | |
| }, | |
| { | |
| "epoch": 11.70190925887908, | |
| "eval_cer": 0.018809241706161137, | |
| "eval_loss": 0.04154360666871071, | |
| "eval_runtime": 22.0458, | |
| "eval_samples_per_second": 4.082, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0679177837354781, | |
| "step": 114000 | |
| }, | |
| { | |
| "epoch": 11.804557585711352, | |
| "grad_norm": 0.20417150855064392, | |
| "learning_rate": 0.00038274017691832955, | |
| "loss": 0.0649, | |
| "step": 115000 | |
| }, | |
| { | |
| "epoch": 11.804557585711352, | |
| "eval_cer": 0.018661137440758292, | |
| "eval_loss": 0.04089859500527382, | |
| "eval_runtime": 22.0112, | |
| "eval_samples_per_second": 4.089, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06702412868632708, | |
| "step": 115000 | |
| }, | |
| { | |
| "epoch": 11.907205912543626, | |
| "grad_norm": 0.27801695466041565, | |
| "learning_rate": 0.0003817115819790167, | |
| "loss": 0.0643, | |
| "step": 116000 | |
| }, | |
| { | |
| "epoch": 11.907205912543626, | |
| "eval_cer": 0.020438388625592416, | |
| "eval_loss": 0.04090258479118347, | |
| "eval_runtime": 22.1223, | |
| "eval_samples_per_second": 4.068, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07059874888293119, | |
| "step": 116000 | |
| }, | |
| { | |
| "epoch": 12.009854239375898, | |
| "grad_norm": 0.24761049449443817, | |
| "learning_rate": 0.0003806829870397038, | |
| "loss": 0.0641, | |
| "step": 117000 | |
| }, | |
| { | |
| "epoch": 12.009854239375898, | |
| "eval_cer": 0.02103080568720379, | |
| "eval_loss": 0.041068486869335175, | |
| "eval_runtime": 22.0016, | |
| "eval_samples_per_second": 4.091, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0741733690795353, | |
| "step": 117000 | |
| }, | |
| { | |
| "epoch": 12.112502566208171, | |
| "grad_norm": 0.18434394896030426, | |
| "learning_rate": 0.0003796543921003909, | |
| "loss": 0.0625, | |
| "step": 118000 | |
| }, | |
| { | |
| "epoch": 12.112502566208171, | |
| "eval_cer": 0.020438388625592416, | |
| "eval_loss": 0.04070517420768738, | |
| "eval_runtime": 22.1605, | |
| "eval_samples_per_second": 4.061, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07149240393208221, | |
| "step": 118000 | |
| }, | |
| { | |
| "epoch": 12.215150893040443, | |
| "grad_norm": 0.24979180097579956, | |
| "learning_rate": 0.000378625797161078, | |
| "loss": 0.0629, | |
| "step": 119000 | |
| }, | |
| { | |
| "epoch": 12.215150893040443, | |
| "eval_cer": 0.019845971563981043, | |
| "eval_loss": 0.038933165371418, | |
| "eval_runtime": 22.1583, | |
| "eval_samples_per_second": 4.062, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06613047363717604, | |
| "step": 119000 | |
| }, | |
| { | |
| "epoch": 12.317799219872716, | |
| "grad_norm": 0.29532870650291443, | |
| "learning_rate": 0.0003775972022217651, | |
| "loss": 0.0634, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 12.317799219872716, | |
| "eval_cer": 0.020438388625592416, | |
| "eval_loss": 0.04176652058959007, | |
| "eval_runtime": 22.5206, | |
| "eval_samples_per_second": 3.996, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.07506702412868632, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 12.420447546704988, | |
| "grad_norm": 0.2764800190925598, | |
| "learning_rate": 0.00037656860728245217, | |
| "loss": 0.0628, | |
| "step": 121000 | |
| }, | |
| { | |
| "epoch": 12.420447546704988, | |
| "eval_cer": 0.019845971563981043, | |
| "eval_loss": 0.038935501128435135, | |
| "eval_runtime": 22.4914, | |
| "eval_samples_per_second": 4.002, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.06970509383378017, | |
| "step": 121000 | |
| }, | |
| { | |
| "epoch": 12.523095873537262, | |
| "grad_norm": 0.20742572844028473, | |
| "learning_rate": 0.00037554001234313926, | |
| "loss": 0.0638, | |
| "step": 122000 | |
| }, | |
| { | |
| "epoch": 12.523095873537262, | |
| "eval_cer": 0.020734597156398103, | |
| "eval_loss": 0.040505990386009216, | |
| "eval_runtime": 22.4563, | |
| "eval_samples_per_second": 4.008, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07059874888293119, | |
| "step": 122000 | |
| }, | |
| { | |
| "epoch": 12.625744200369533, | |
| "grad_norm": 0.22677256166934967, | |
| "learning_rate": 0.0003745114174038264, | |
| "loss": 0.063, | |
| "step": 123000 | |
| }, | |
| { | |
| "epoch": 12.625744200369533, | |
| "eval_cer": 0.02058649289099526, | |
| "eval_loss": 0.0383678562939167, | |
| "eval_runtime": 22.3288, | |
| "eval_samples_per_second": 4.031, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06881143878462913, | |
| "step": 123000 | |
| }, | |
| { | |
| "epoch": 12.728392527201807, | |
| "grad_norm": 0.2105027288198471, | |
| "learning_rate": 0.0003734828224645135, | |
| "loss": 0.0633, | |
| "step": 124000 | |
| }, | |
| { | |
| "epoch": 12.728392527201807, | |
| "eval_cer": 0.01925355450236967, | |
| "eval_loss": 0.03879451006650925, | |
| "eval_runtime": 22.3197, | |
| "eval_samples_per_second": 4.032, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07059874888293119, | |
| "step": 124000 | |
| }, | |
| { | |
| "epoch": 12.831040854034079, | |
| "grad_norm": 0.21093320846557617, | |
| "learning_rate": 0.0003724542275252006, | |
| "loss": 0.0636, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 12.831040854034079, | |
| "eval_cer": 0.020734597156398103, | |
| "eval_loss": 0.038905300199985504, | |
| "eval_runtime": 22.1791, | |
| "eval_samples_per_second": 4.058, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06702412868632708, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 12.933689180866352, | |
| "grad_norm": 0.21501018106937408, | |
| "learning_rate": 0.0003714256325858877, | |
| "loss": 0.0634, | |
| "step": 126000 | |
| }, | |
| { | |
| "epoch": 12.933689180866352, | |
| "eval_cer": 0.018809241706161137, | |
| "eval_loss": 0.03828004002571106, | |
| "eval_runtime": 22.175, | |
| "eval_samples_per_second": 4.059, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06613047363717604, | |
| "step": 126000 | |
| }, | |
| { | |
| "epoch": 13.036337507698624, | |
| "grad_norm": 0.2869448661804199, | |
| "learning_rate": 0.0003703970376465748, | |
| "loss": 0.0623, | |
| "step": 127000 | |
| }, | |
| { | |
| "epoch": 13.036337507698624, | |
| "eval_cer": 0.02221563981042654, | |
| "eval_loss": 0.03908955305814743, | |
| "eval_runtime": 22.0973, | |
| "eval_samples_per_second": 4.073, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0741733690795353, | |
| "step": 127000 | |
| }, | |
| { | |
| "epoch": 13.138985834530898, | |
| "grad_norm": 0.20509861409664154, | |
| "learning_rate": 0.00036936844270726187, | |
| "loss": 0.0616, | |
| "step": 128000 | |
| }, | |
| { | |
| "epoch": 13.138985834530898, | |
| "eval_cer": 0.021919431279620854, | |
| "eval_loss": 0.0386139452457428, | |
| "eval_runtime": 22.2024, | |
| "eval_samples_per_second": 4.054, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07149240393208221, | |
| "step": 128000 | |
| }, | |
| { | |
| "epoch": 13.24163416136317, | |
| "grad_norm": 0.17765522003173828, | |
| "learning_rate": 0.00036833984776794897, | |
| "loss": 0.0618, | |
| "step": 129000 | |
| }, | |
| { | |
| "epoch": 13.24163416136317, | |
| "eval_cer": 0.01925355450236967, | |
| "eval_loss": 0.0378284677863121, | |
| "eval_runtime": 22.0041, | |
| "eval_samples_per_second": 4.09, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06613047363717604, | |
| "step": 129000 | |
| }, | |
| { | |
| "epoch": 13.344282488195443, | |
| "grad_norm": 0.24877817928791046, | |
| "learning_rate": 0.0003673112528286361, | |
| "loss": 0.0619, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 13.344282488195443, | |
| "eval_cer": 0.021475118483412322, | |
| "eval_loss": 0.03674837946891785, | |
| "eval_runtime": 22.0962, | |
| "eval_samples_per_second": 4.073, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0741733690795353, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 13.446930815027715, | |
| "grad_norm": 0.25357529520988464, | |
| "learning_rate": 0.0003662826578893232, | |
| "loss": 0.0623, | |
| "step": 131000 | |
| }, | |
| { | |
| "epoch": 13.446930815027715, | |
| "eval_cer": 0.019994075829383885, | |
| "eval_loss": 0.03821048513054848, | |
| "eval_runtime": 22.1382, | |
| "eval_samples_per_second": 4.065, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06970509383378017, | |
| "step": 131000 | |
| }, | |
| { | |
| "epoch": 13.549579141859988, | |
| "grad_norm": 0.32752060890197754, | |
| "learning_rate": 0.0003652540629500103, | |
| "loss": 0.0622, | |
| "step": 132000 | |
| }, | |
| { | |
| "epoch": 13.549579141859988, | |
| "eval_cer": 0.02014218009478673, | |
| "eval_loss": 0.038477059453725815, | |
| "eval_runtime": 22.2355, | |
| "eval_samples_per_second": 4.048, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06702412868632708, | |
| "step": 132000 | |
| }, | |
| { | |
| "epoch": 13.65222746869226, | |
| "grad_norm": 0.22504030168056488, | |
| "learning_rate": 0.0003642254680106974, | |
| "loss": 0.0622, | |
| "step": 133000 | |
| }, | |
| { | |
| "epoch": 13.65222746869226, | |
| "eval_cer": 0.018809241706161137, | |
| "eval_loss": 0.038305170834064484, | |
| "eval_runtime": 22.2134, | |
| "eval_samples_per_second": 4.052, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06970509383378017, | |
| "step": 133000 | |
| }, | |
| { | |
| "epoch": 13.754875795524534, | |
| "grad_norm": 0.23793594539165497, | |
| "learning_rate": 0.0003631968730713845, | |
| "loss": 0.062, | |
| "step": 134000 | |
| }, | |
| { | |
| "epoch": 13.754875795524534, | |
| "eval_cer": 0.02058649289099526, | |
| "eval_loss": 0.03838730975985527, | |
| "eval_runtime": 22.3142, | |
| "eval_samples_per_second": 4.033, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0679177837354781, | |
| "step": 134000 | |
| }, | |
| { | |
| "epoch": 13.857524122356805, | |
| "grad_norm": 0.18214410543441772, | |
| "learning_rate": 0.0003621682781320716, | |
| "loss": 0.0625, | |
| "step": 135000 | |
| }, | |
| { | |
| "epoch": 13.857524122356805, | |
| "eval_cer": 0.02014218009478673, | |
| "eval_loss": 0.03858475759625435, | |
| "eval_runtime": 22.3822, | |
| "eval_samples_per_second": 4.021, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06881143878462913, | |
| "step": 135000 | |
| }, | |
| { | |
| "epoch": 13.960172449189079, | |
| "grad_norm": 0.24008499085903168, | |
| "learning_rate": 0.00036113968319275867, | |
| "loss": 0.0621, | |
| "step": 136000 | |
| }, | |
| { | |
| "epoch": 13.960172449189079, | |
| "eval_cer": 0.019697867298578198, | |
| "eval_loss": 0.03749372810125351, | |
| "eval_runtime": 22.3573, | |
| "eval_samples_per_second": 4.026, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06702412868632708, | |
| "step": 136000 | |
| }, | |
| { | |
| "epoch": 14.06282077602135, | |
| "grad_norm": 0.20032504200935364, | |
| "learning_rate": 0.0003601110882534458, | |
| "loss": 0.0616, | |
| "step": 137000 | |
| }, | |
| { | |
| "epoch": 14.06282077602135, | |
| "eval_cer": 0.02058649289099526, | |
| "eval_loss": 0.03601989150047302, | |
| "eval_runtime": 22.1524, | |
| "eval_samples_per_second": 4.063, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06970509383378017, | |
| "step": 137000 | |
| }, | |
| { | |
| "epoch": 14.165469102853624, | |
| "grad_norm": 0.29345226287841797, | |
| "learning_rate": 0.0003590824933141329, | |
| "loss": 0.0611, | |
| "step": 138000 | |
| }, | |
| { | |
| "epoch": 14.165469102853624, | |
| "eval_cer": 0.018809241706161137, | |
| "eval_loss": 0.036261823028326035, | |
| "eval_runtime": 22.1903, | |
| "eval_samples_per_second": 4.056, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06344950848972297, | |
| "step": 138000 | |
| }, | |
| { | |
| "epoch": 14.268117429685896, | |
| "grad_norm": 0.19695940613746643, | |
| "learning_rate": 0.00035805389837482, | |
| "loss": 0.0606, | |
| "step": 139000 | |
| }, | |
| { | |
| "epoch": 14.268117429685896, | |
| "eval_cer": 0.021623222748815167, | |
| "eval_loss": 0.0368872731924057, | |
| "eval_runtime": 22.1734, | |
| "eval_samples_per_second": 4.059, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07238605898123325, | |
| "step": 139000 | |
| }, | |
| { | |
| "epoch": 14.37076575651817, | |
| "grad_norm": 0.1869696080684662, | |
| "learning_rate": 0.0003570253034355071, | |
| "loss": 0.0609, | |
| "step": 140000 | |
| }, | |
| { | |
| "epoch": 14.37076575651817, | |
| "eval_cer": 0.019845971563981043, | |
| "eval_loss": 0.036232832819223404, | |
| "eval_runtime": 22.2094, | |
| "eval_samples_per_second": 4.052, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06613047363717604, | |
| "step": 140000 | |
| }, | |
| { | |
| "epoch": 14.473414083350441, | |
| "grad_norm": 0.3025355041027069, | |
| "learning_rate": 0.0003559967084961942, | |
| "loss": 0.0611, | |
| "step": 141000 | |
| }, | |
| { | |
| "epoch": 14.473414083350441, | |
| "eval_cer": 0.02088270142180095, | |
| "eval_loss": 0.036524925380945206, | |
| "eval_runtime": 22.0892, | |
| "eval_samples_per_second": 4.074, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07149240393208221, | |
| "step": 141000 | |
| }, | |
| { | |
| "epoch": 14.576062410182715, | |
| "grad_norm": 0.2218203842639923, | |
| "learning_rate": 0.0003549681135568813, | |
| "loss": 0.0612, | |
| "step": 142000 | |
| }, | |
| { | |
| "epoch": 14.576062410182715, | |
| "eval_cer": 0.02058649289099526, | |
| "eval_loss": 0.03738318383693695, | |
| "eval_runtime": 22.4443, | |
| "eval_samples_per_second": 4.01, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07149240393208221, | |
| "step": 142000 | |
| }, | |
| { | |
| "epoch": 14.678710737014987, | |
| "grad_norm": 0.18811815977096558, | |
| "learning_rate": 0.0003539395186175684, | |
| "loss": 0.0606, | |
| "step": 143000 | |
| }, | |
| { | |
| "epoch": 14.678710737014987, | |
| "eval_cer": 0.019697867298578198, | |
| "eval_loss": 0.03684472665190697, | |
| "eval_runtime": 22.3128, | |
| "eval_samples_per_second": 4.034, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07238605898123325, | |
| "step": 143000 | |
| }, | |
| { | |
| "epoch": 14.78135906384726, | |
| "grad_norm": 0.20071063935756683, | |
| "learning_rate": 0.0003529109236782555, | |
| "loss": 0.0609, | |
| "step": 144000 | |
| }, | |
| { | |
| "epoch": 14.78135906384726, | |
| "eval_cer": 0.018364928909952605, | |
| "eval_loss": 0.037738535553216934, | |
| "eval_runtime": 22.5071, | |
| "eval_samples_per_second": 3.999, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.064343163538874, | |
| "step": 144000 | |
| }, | |
| { | |
| "epoch": 14.884007390679532, | |
| "grad_norm": 0.2046099305152893, | |
| "learning_rate": 0.0003518823287389426, | |
| "loss": 0.0614, | |
| "step": 145000 | |
| }, | |
| { | |
| "epoch": 14.884007390679532, | |
| "eval_cer": 0.019697867298578198, | |
| "eval_loss": 0.03707651048898697, | |
| "eval_runtime": 22.4459, | |
| "eval_samples_per_second": 4.01, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06523681858802502, | |
| "step": 145000 | |
| }, | |
| { | |
| "epoch": 14.986655717511805, | |
| "grad_norm": 0.27886244654655457, | |
| "learning_rate": 0.0003508537337996297, | |
| "loss": 0.0616, | |
| "step": 146000 | |
| }, | |
| { | |
| "epoch": 14.986655717511805, | |
| "eval_cer": 0.022511848341232227, | |
| "eval_loss": 0.038485873490571976, | |
| "eval_runtime": 22.2607, | |
| "eval_samples_per_second": 4.043, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07685433422698838, | |
| "step": 146000 | |
| }, | |
| { | |
| "epoch": 15.089304044344077, | |
| "grad_norm": 0.3120444118976593, | |
| "learning_rate": 0.0003498251388603168, | |
| "loss": 0.0598, | |
| "step": 147000 | |
| }, | |
| { | |
| "epoch": 15.089304044344077, | |
| "eval_cer": 0.021475118483412322, | |
| "eval_loss": 0.03692319989204407, | |
| "eval_runtime": 22.2577, | |
| "eval_samples_per_second": 4.044, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07685433422698838, | |
| "step": 147000 | |
| }, | |
| { | |
| "epoch": 15.19195237117635, | |
| "grad_norm": 0.2510223388671875, | |
| "learning_rate": 0.0003487965439210039, | |
| "loss": 0.06, | |
| "step": 148000 | |
| }, | |
| { | |
| "epoch": 15.19195237117635, | |
| "eval_cer": 0.019549763033175356, | |
| "eval_loss": 0.036081377416849136, | |
| "eval_runtime": 22.3708, | |
| "eval_samples_per_second": 4.023, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06881143878462913, | |
| "step": 148000 | |
| }, | |
| { | |
| "epoch": 15.294600698008622, | |
| "grad_norm": 0.2539554536342621, | |
| "learning_rate": 0.000347767948981691, | |
| "loss": 0.0603, | |
| "step": 149000 | |
| }, | |
| { | |
| "epoch": 15.294600698008622, | |
| "eval_cer": 0.019105450236966824, | |
| "eval_loss": 0.036284659057855606, | |
| "eval_runtime": 22.2203, | |
| "eval_samples_per_second": 4.05, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07059874888293119, | |
| "step": 149000 | |
| }, | |
| { | |
| "epoch": 15.397249024840896, | |
| "grad_norm": 0.3537413775920868, | |
| "learning_rate": 0.0003467393540423781, | |
| "loss": 0.0602, | |
| "step": 150000 | |
| }, | |
| { | |
| "epoch": 15.397249024840896, | |
| "eval_cer": 0.01925355450236967, | |
| "eval_loss": 0.037329014390707016, | |
| "eval_runtime": 22.3303, | |
| "eval_samples_per_second": 4.03, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0679177837354781, | |
| "step": 150000 | |
| }, | |
| { | |
| "epoch": 15.499897351673168, | |
| "grad_norm": 0.31098708510398865, | |
| "learning_rate": 0.00034571075910306523, | |
| "loss": 0.0599, | |
| "step": 151000 | |
| }, | |
| { | |
| "epoch": 15.499897351673168, | |
| "eval_cer": 0.019697867298578198, | |
| "eval_loss": 0.0374312698841095, | |
| "eval_runtime": 22.4269, | |
| "eval_samples_per_second": 4.013, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06613047363717604, | |
| "step": 151000 | |
| }, | |
| { | |
| "epoch": 15.60254567850544, | |
| "grad_norm": 0.18067589402198792, | |
| "learning_rate": 0.0003446821641637523, | |
| "loss": 0.0603, | |
| "step": 152000 | |
| }, | |
| { | |
| "epoch": 15.60254567850544, | |
| "eval_cer": 0.019105450236966824, | |
| "eval_loss": 0.03638828173279762, | |
| "eval_runtime": 22.5221, | |
| "eval_samples_per_second": 3.996, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.06344950848972297, | |
| "step": 152000 | |
| }, | |
| { | |
| "epoch": 15.705194005337713, | |
| "grad_norm": 0.26867741346359253, | |
| "learning_rate": 0.0003436535692244394, | |
| "loss": 0.0605, | |
| "step": 153000 | |
| }, | |
| { | |
| "epoch": 15.705194005337713, | |
| "eval_cer": 0.019845971563981043, | |
| "eval_loss": 0.03720884397625923, | |
| "eval_runtime": 22.3098, | |
| "eval_samples_per_second": 4.034, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06523681858802502, | |
| "step": 153000 | |
| }, | |
| { | |
| "epoch": 15.807842332169985, | |
| "grad_norm": 0.18743161857128143, | |
| "learning_rate": 0.0003426249742851265, | |
| "loss": 0.06, | |
| "step": 154000 | |
| }, | |
| { | |
| "epoch": 15.807842332169985, | |
| "eval_cer": 0.021919431279620854, | |
| "eval_loss": 0.03767675533890724, | |
| "eval_runtime": 22.2871, | |
| "eval_samples_per_second": 4.038, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0741733690795353, | |
| "step": 154000 | |
| }, | |
| { | |
| "epoch": 15.910490659002258, | |
| "grad_norm": 0.2153768688440323, | |
| "learning_rate": 0.0003415963793458136, | |
| "loss": 0.0599, | |
| "step": 155000 | |
| }, | |
| { | |
| "epoch": 15.910490659002258, | |
| "eval_cer": 0.020290284360189575, | |
| "eval_loss": 0.03886905685067177, | |
| "eval_runtime": 22.1149, | |
| "eval_samples_per_second": 4.07, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07149240393208221, | |
| "step": 155000 | |
| }, | |
| { | |
| "epoch": 16.01313898583453, | |
| "grad_norm": 0.24325600266456604, | |
| "learning_rate": 0.0003405677844065007, | |
| "loss": 0.0598, | |
| "step": 156000 | |
| }, | |
| { | |
| "epoch": 16.01313898583453, | |
| "eval_cer": 0.02177132701421801, | |
| "eval_loss": 0.03822890669107437, | |
| "eval_runtime": 22.2699, | |
| "eval_samples_per_second": 4.041, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07059874888293119, | |
| "step": 156000 | |
| }, | |
| { | |
| "epoch": 16.115787312666804, | |
| "grad_norm": 0.2438814342021942, | |
| "learning_rate": 0.0003395391894671878, | |
| "loss": 0.0586, | |
| "step": 157000 | |
| }, | |
| { | |
| "epoch": 16.115787312666804, | |
| "eval_cer": 0.019697867298578198, | |
| "eval_loss": 0.038069114089012146, | |
| "eval_runtime": 22.2427, | |
| "eval_samples_per_second": 4.046, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06970509383378017, | |
| "step": 157000 | |
| }, | |
| { | |
| "epoch": 16.218435639499077, | |
| "grad_norm": 0.22072641551494598, | |
| "learning_rate": 0.00033851059452787494, | |
| "loss": 0.0584, | |
| "step": 158000 | |
| }, | |
| { | |
| "epoch": 16.218435639499077, | |
| "eval_cer": 0.0173281990521327, | |
| "eval_loss": 0.03663622587919235, | |
| "eval_runtime": 22.2026, | |
| "eval_samples_per_second": 4.054, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06166219839142091, | |
| "step": 158000 | |
| }, | |
| { | |
| "epoch": 16.321083966331347, | |
| "grad_norm": 0.15790699422359467, | |
| "learning_rate": 0.00033748199958856203, | |
| "loss": 0.0587, | |
| "step": 159000 | |
| }, | |
| { | |
| "epoch": 16.321083966331347, | |
| "eval_cer": 0.019549763033175356, | |
| "eval_loss": 0.03735322132706642, | |
| "eval_runtime": 22.315, | |
| "eval_samples_per_second": 4.033, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06702412868632708, | |
| "step": 159000 | |
| }, | |
| { | |
| "epoch": 16.42373229316362, | |
| "grad_norm": 0.1823723465204239, | |
| "learning_rate": 0.0003364534046492491, | |
| "loss": 0.0591, | |
| "step": 160000 | |
| }, | |
| { | |
| "epoch": 16.42373229316362, | |
| "eval_cer": 0.021178909952606635, | |
| "eval_loss": 0.03700649365782738, | |
| "eval_runtime": 22.6895, | |
| "eval_samples_per_second": 3.967, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.07059874888293119, | |
| "step": 160000 | |
| }, | |
| { | |
| "epoch": 16.526380619995894, | |
| "grad_norm": 0.24684032797813416, | |
| "learning_rate": 0.0003354248097099362, | |
| "loss": 0.0595, | |
| "step": 161000 | |
| }, | |
| { | |
| "epoch": 16.526380619995894, | |
| "eval_cer": 0.019697867298578198, | |
| "eval_loss": 0.03638649359345436, | |
| "eval_runtime": 22.3004, | |
| "eval_samples_per_second": 4.036, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.064343163538874, | |
| "step": 161000 | |
| }, | |
| { | |
| "epoch": 16.629028946828168, | |
| "grad_norm": 0.2088789939880371, | |
| "learning_rate": 0.0003343962147706233, | |
| "loss": 0.059, | |
| "step": 162000 | |
| }, | |
| { | |
| "epoch": 16.629028946828168, | |
| "eval_cer": 0.017772511848341232, | |
| "eval_loss": 0.03632904216647148, | |
| "eval_runtime": 22.2253, | |
| "eval_samples_per_second": 4.049, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06255585344057193, | |
| "step": 162000 | |
| }, | |
| { | |
| "epoch": 16.731677273660438, | |
| "grad_norm": 0.20426543056964874, | |
| "learning_rate": 0.0003333676198313104, | |
| "loss": 0.0598, | |
| "step": 163000 | |
| }, | |
| { | |
| "epoch": 16.731677273660438, | |
| "eval_cer": 0.019697867298578198, | |
| "eval_loss": 0.03629469498991966, | |
| "eval_runtime": 22.3453, | |
| "eval_samples_per_second": 4.028, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06523681858802502, | |
| "step": 163000 | |
| }, | |
| { | |
| "epoch": 16.83432560049271, | |
| "grad_norm": 0.2737009525299072, | |
| "learning_rate": 0.0003323390248919975, | |
| "loss": 0.0591, | |
| "step": 164000 | |
| }, | |
| { | |
| "epoch": 16.83432560049271, | |
| "eval_cer": 0.017772511848341232, | |
| "eval_loss": 0.03445509076118469, | |
| "eval_runtime": 22.3887, | |
| "eval_samples_per_second": 4.02, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06344950848972297, | |
| "step": 164000 | |
| }, | |
| { | |
| "epoch": 16.936973927324985, | |
| "grad_norm": 0.33581215143203735, | |
| "learning_rate": 0.00033131042995268465, | |
| "loss": 0.0586, | |
| "step": 165000 | |
| }, | |
| { | |
| "epoch": 16.936973927324985, | |
| "eval_cer": 0.020290284360189575, | |
| "eval_loss": 0.036238256841897964, | |
| "eval_runtime": 22.1998, | |
| "eval_samples_per_second": 4.054, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06970509383378017, | |
| "step": 165000 | |
| }, | |
| { | |
| "epoch": 17.03962225415726, | |
| "grad_norm": 0.277771532535553, | |
| "learning_rate": 0.00033028183501337174, | |
| "loss": 0.0591, | |
| "step": 166000 | |
| }, | |
| { | |
| "epoch": 17.03962225415726, | |
| "eval_cer": 0.01806872037914692, | |
| "eval_loss": 0.035338886082172394, | |
| "eval_runtime": 22.132, | |
| "eval_samples_per_second": 4.067, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.064343163538874, | |
| "step": 166000 | |
| }, | |
| { | |
| "epoch": 17.14227058098953, | |
| "grad_norm": 0.17568770051002502, | |
| "learning_rate": 0.00032925324007405883, | |
| "loss": 0.0578, | |
| "step": 167000 | |
| }, | |
| { | |
| "epoch": 17.14227058098953, | |
| "eval_cer": 0.017772511848341232, | |
| "eval_loss": 0.03632746636867523, | |
| "eval_runtime": 22.2554, | |
| "eval_samples_per_second": 4.044, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.064343163538874, | |
| "step": 167000 | |
| }, | |
| { | |
| "epoch": 17.244918907821802, | |
| "grad_norm": 0.22311237454414368, | |
| "learning_rate": 0.0003282246451347459, | |
| "loss": 0.0576, | |
| "step": 168000 | |
| }, | |
| { | |
| "epoch": 17.244918907821802, | |
| "eval_cer": 0.01925355450236967, | |
| "eval_loss": 0.036676324903964996, | |
| "eval_runtime": 22.484, | |
| "eval_samples_per_second": 4.003, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.06702412868632708, | |
| "step": 168000 | |
| }, | |
| { | |
| "epoch": 17.347567234654075, | |
| "grad_norm": 0.3649640381336212, | |
| "learning_rate": 0.000327196050195433, | |
| "loss": 0.0576, | |
| "step": 169000 | |
| }, | |
| { | |
| "epoch": 17.347567234654075, | |
| "eval_cer": 0.019105450236966824, | |
| "eval_loss": 0.035018790513277054, | |
| "eval_runtime": 22.2563, | |
| "eval_samples_per_second": 4.044, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06613047363717604, | |
| "step": 169000 | |
| }, | |
| { | |
| "epoch": 17.45021556148635, | |
| "grad_norm": 0.2809694707393646, | |
| "learning_rate": 0.0003261674552561201, | |
| "loss": 0.0582, | |
| "step": 170000 | |
| }, | |
| { | |
| "epoch": 17.45021556148635, | |
| "eval_cer": 0.018661137440758292, | |
| "eval_loss": 0.035496581345796585, | |
| "eval_runtime": 22.1719, | |
| "eval_samples_per_second": 4.059, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06613047363717604, | |
| "step": 170000 | |
| }, | |
| { | |
| "epoch": 17.55286388831862, | |
| "grad_norm": 0.23279882967472076, | |
| "learning_rate": 0.0003251388603168072, | |
| "loss": 0.0587, | |
| "step": 171000 | |
| }, | |
| { | |
| "epoch": 17.55286388831862, | |
| "eval_cer": 0.018364928909952605, | |
| "eval_loss": 0.03641456365585327, | |
| "eval_runtime": 22.1562, | |
| "eval_samples_per_second": 4.062, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06613047363717604, | |
| "step": 171000 | |
| }, | |
| { | |
| "epoch": 17.655512215150893, | |
| "grad_norm": 0.23531590402126312, | |
| "learning_rate": 0.00032411026537749435, | |
| "loss": 0.0587, | |
| "step": 172000 | |
| }, | |
| { | |
| "epoch": 17.655512215150893, | |
| "eval_cer": 0.018216824644549764, | |
| "eval_loss": 0.034297019243240356, | |
| "eval_runtime": 22.3222, | |
| "eval_samples_per_second": 4.032, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.058981233243967826, | |
| "step": 172000 | |
| }, | |
| { | |
| "epoch": 17.758160541983166, | |
| "grad_norm": 0.38004380464553833, | |
| "learning_rate": 0.00032308167043818144, | |
| "loss": 0.0589, | |
| "step": 173000 | |
| }, | |
| { | |
| "epoch": 17.758160541983166, | |
| "eval_cer": 0.017772511848341232, | |
| "eval_loss": 0.03497479110956192, | |
| "eval_runtime": 22.1444, | |
| "eval_samples_per_second": 4.064, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06255585344057193, | |
| "step": 173000 | |
| }, | |
| { | |
| "epoch": 17.86080886881544, | |
| "grad_norm": 0.22771824896335602, | |
| "learning_rate": 0.00032205307549886854, | |
| "loss": 0.0578, | |
| "step": 174000 | |
| }, | |
| { | |
| "epoch": 17.86080886881544, | |
| "eval_cer": 0.019697867298578198, | |
| "eval_loss": 0.034568045288324356, | |
| "eval_runtime": 22.2545, | |
| "eval_samples_per_second": 4.044, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06613047363717604, | |
| "step": 174000 | |
| }, | |
| { | |
| "epoch": 17.96345719564771, | |
| "grad_norm": 0.3113113343715668, | |
| "learning_rate": 0.00032102448055955563, | |
| "loss": 0.0586, | |
| "step": 175000 | |
| }, | |
| { | |
| "epoch": 17.96345719564771, | |
| "eval_cer": 0.01851303317535545, | |
| "eval_loss": 0.035004787147045135, | |
| "eval_runtime": 22.3269, | |
| "eval_samples_per_second": 4.031, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.064343163538874, | |
| "step": 175000 | |
| }, | |
| { | |
| "epoch": 18.066105522479983, | |
| "grad_norm": 0.24952055513858795, | |
| "learning_rate": 0.0003199958856202427, | |
| "loss": 0.0571, | |
| "step": 176000 | |
| }, | |
| { | |
| "epoch": 18.066105522479983, | |
| "eval_cer": 0.019105450236966824, | |
| "eval_loss": 0.03602117672562599, | |
| "eval_runtime": 22.4032, | |
| "eval_samples_per_second": 4.017, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.064343163538874, | |
| "step": 176000 | |
| }, | |
| { | |
| "epoch": 18.168753849312257, | |
| "grad_norm": 0.33917441964149475, | |
| "learning_rate": 0.0003189672906809298, | |
| "loss": 0.0569, | |
| "step": 177000 | |
| }, | |
| { | |
| "epoch": 18.168753849312257, | |
| "eval_cer": 0.019105450236966824, | |
| "eval_loss": 0.036349765956401825, | |
| "eval_runtime": 22.273, | |
| "eval_samples_per_second": 4.041, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06613047363717604, | |
| "step": 177000 | |
| }, | |
| { | |
| "epoch": 18.27140217614453, | |
| "grad_norm": 0.18371905386447906, | |
| "learning_rate": 0.0003179386957416169, | |
| "loss": 0.0566, | |
| "step": 178000 | |
| }, | |
| { | |
| "epoch": 18.27140217614453, | |
| "eval_cer": 0.018661137440758292, | |
| "eval_loss": 0.03550698608160019, | |
| "eval_runtime": 22.3312, | |
| "eval_samples_per_second": 4.03, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06613047363717604, | |
| "step": 178000 | |
| }, | |
| { | |
| "epoch": 18.3740505029768, | |
| "grad_norm": 0.2601664662361145, | |
| "learning_rate": 0.00031691010080230406, | |
| "loss": 0.057, | |
| "step": 179000 | |
| }, | |
| { | |
| "epoch": 18.3740505029768, | |
| "eval_cer": 0.018957345971563982, | |
| "eval_loss": 0.035991400480270386, | |
| "eval_runtime": 22.2432, | |
| "eval_samples_per_second": 4.046, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06613047363717604, | |
| "step": 179000 | |
| }, | |
| { | |
| "epoch": 18.476698829809074, | |
| "grad_norm": 0.17889824509620667, | |
| "learning_rate": 0.00031588150586299115, | |
| "loss": 0.057, | |
| "step": 180000 | |
| }, | |
| { | |
| "epoch": 18.476698829809074, | |
| "eval_cer": 0.018661137440758292, | |
| "eval_loss": 0.03573347255587578, | |
| "eval_runtime": 22.2321, | |
| "eval_samples_per_second": 4.048, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06523681858802502, | |
| "step": 180000 | |
| }, | |
| { | |
| "epoch": 18.579347156641347, | |
| "grad_norm": 0.5736141800880432, | |
| "learning_rate": 0.00031485291092367824, | |
| "loss": 0.0578, | |
| "step": 181000 | |
| }, | |
| { | |
| "epoch": 18.579347156641347, | |
| "eval_cer": 0.01851303317535545, | |
| "eval_loss": 0.035990502685308456, | |
| "eval_runtime": 22.2434, | |
| "eval_samples_per_second": 4.046, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06523681858802502, | |
| "step": 181000 | |
| }, | |
| { | |
| "epoch": 18.68199548347362, | |
| "grad_norm": 0.2320089340209961, | |
| "learning_rate": 0.00031382431598436534, | |
| "loss": 0.0581, | |
| "step": 182000 | |
| }, | |
| { | |
| "epoch": 18.68199548347362, | |
| "eval_cer": 0.0173281990521327, | |
| "eval_loss": 0.035308029502630234, | |
| "eval_runtime": 22.161, | |
| "eval_samples_per_second": 4.061, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.064343163538874, | |
| "step": 182000 | |
| }, | |
| { | |
| "epoch": 18.78464381030589, | |
| "grad_norm": 0.31181567907333374, | |
| "learning_rate": 0.00031279572104505243, | |
| "loss": 0.0575, | |
| "step": 183000 | |
| }, | |
| { | |
| "epoch": 18.78464381030589, | |
| "eval_cer": 0.017920616113744077, | |
| "eval_loss": 0.03400159254670143, | |
| "eval_runtime": 22.1991, | |
| "eval_samples_per_second": 4.054, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06076854334226988, | |
| "step": 183000 | |
| }, | |
| { | |
| "epoch": 18.887292137138164, | |
| "grad_norm": 0.26530441641807556, | |
| "learning_rate": 0.0003117671261057395, | |
| "loss": 0.0574, | |
| "step": 184000 | |
| }, | |
| { | |
| "epoch": 18.887292137138164, | |
| "eval_cer": 0.020290284360189575, | |
| "eval_loss": 0.03510544076561928, | |
| "eval_runtime": 22.2589, | |
| "eval_samples_per_second": 4.043, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07059874888293119, | |
| "step": 184000 | |
| }, | |
| { | |
| "epoch": 18.989940463970438, | |
| "grad_norm": 0.2680375576019287, | |
| "learning_rate": 0.0003107385311664266, | |
| "loss": 0.0576, | |
| "step": 185000 | |
| }, | |
| { | |
| "epoch": 18.989940463970438, | |
| "eval_cer": 0.021178909952606635, | |
| "eval_loss": 0.037547577172517776, | |
| "eval_runtime": 22.2321, | |
| "eval_samples_per_second": 4.048, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07238605898123325, | |
| "step": 185000 | |
| }, | |
| { | |
| "epoch": 19.09258879080271, | |
| "grad_norm": 0.2254854440689087, | |
| "learning_rate": 0.00030970993622711376, | |
| "loss": 0.0561, | |
| "step": 186000 | |
| }, | |
| { | |
| "epoch": 19.09258879080271, | |
| "eval_cer": 0.01940165876777251, | |
| "eval_loss": 0.03498771786689758, | |
| "eval_runtime": 22.2675, | |
| "eval_samples_per_second": 4.042, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06344950848972297, | |
| "step": 186000 | |
| }, | |
| { | |
| "epoch": 19.19523711763498, | |
| "grad_norm": 0.2187499850988388, | |
| "learning_rate": 0.00030868134128780086, | |
| "loss": 0.0563, | |
| "step": 187000 | |
| }, | |
| { | |
| "epoch": 19.19523711763498, | |
| "eval_cer": 0.020734597156398103, | |
| "eval_loss": 0.035063955932855606, | |
| "eval_runtime": 22.2231, | |
| "eval_samples_per_second": 4.05, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07059874888293119, | |
| "step": 187000 | |
| }, | |
| { | |
| "epoch": 19.297885444467255, | |
| "grad_norm": 0.3217693567276001, | |
| "learning_rate": 0.000307652746348488, | |
| "loss": 0.0566, | |
| "step": 188000 | |
| }, | |
| { | |
| "epoch": 19.297885444467255, | |
| "eval_cer": 0.020734597156398103, | |
| "eval_loss": 0.034946467727422714, | |
| "eval_runtime": 22.2783, | |
| "eval_samples_per_second": 4.04, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0679177837354781, | |
| "step": 188000 | |
| }, | |
| { | |
| "epoch": 19.40053377129953, | |
| "grad_norm": 0.2570216655731201, | |
| "learning_rate": 0.0003066241514091751, | |
| "loss": 0.056, | |
| "step": 189000 | |
| }, | |
| { | |
| "epoch": 19.40053377129953, | |
| "eval_cer": 0.019105450236966824, | |
| "eval_loss": 0.03498660773038864, | |
| "eval_runtime": 22.3189, | |
| "eval_samples_per_second": 4.032, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0679177837354781, | |
| "step": 189000 | |
| }, | |
| { | |
| "epoch": 19.503182098131802, | |
| "grad_norm": 0.24969050288200378, | |
| "learning_rate": 0.0003055955564698622, | |
| "loss": 0.0565, | |
| "step": 190000 | |
| }, | |
| { | |
| "epoch": 19.503182098131802, | |
| "eval_cer": 0.018809241706161137, | |
| "eval_loss": 0.034704625606536865, | |
| "eval_runtime": 22.2267, | |
| "eval_samples_per_second": 4.049, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06076854334226988, | |
| "step": 190000 | |
| }, | |
| { | |
| "epoch": 19.605830424964072, | |
| "grad_norm": 0.19874796271324158, | |
| "learning_rate": 0.0003045669615305493, | |
| "loss": 0.0565, | |
| "step": 191000 | |
| }, | |
| { | |
| "epoch": 19.605830424964072, | |
| "eval_cer": 0.017772511848341232, | |
| "eval_loss": 0.03414672613143921, | |
| "eval_runtime": 22.4133, | |
| "eval_samples_per_second": 4.015, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05987488829311886, | |
| "step": 191000 | |
| }, | |
| { | |
| "epoch": 19.708478751796346, | |
| "grad_norm": 0.21598631143569946, | |
| "learning_rate": 0.0003035383665912364, | |
| "loss": 0.0569, | |
| "step": 192000 | |
| }, | |
| { | |
| "epoch": 19.708478751796346, | |
| "eval_cer": 0.02088270142180095, | |
| "eval_loss": 0.03441624715924263, | |
| "eval_runtime": 22.4978, | |
| "eval_samples_per_second": 4.0, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.0679177837354781, | |
| "step": 192000 | |
| }, | |
| { | |
| "epoch": 19.81112707862862, | |
| "grad_norm": 0.740821361541748, | |
| "learning_rate": 0.0003025097716519235, | |
| "loss": 0.0566, | |
| "step": 193000 | |
| }, | |
| { | |
| "epoch": 19.81112707862862, | |
| "eval_cer": 0.018809241706161137, | |
| "eval_loss": 0.03505128249526024, | |
| "eval_runtime": 22.2297, | |
| "eval_samples_per_second": 4.049, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.064343163538874, | |
| "step": 193000 | |
| }, | |
| { | |
| "epoch": 19.913775405460893, | |
| "grad_norm": 0.25967320799827576, | |
| "learning_rate": 0.0003014811767126106, | |
| "loss": 0.0566, | |
| "step": 194000 | |
| }, | |
| { | |
| "epoch": 19.913775405460893, | |
| "eval_cer": 0.02088270142180095, | |
| "eval_loss": 0.03540065139532089, | |
| "eval_runtime": 22.328, | |
| "eval_samples_per_second": 4.031, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07327971403038427, | |
| "step": 194000 | |
| }, | |
| { | |
| "epoch": 20.016423732293163, | |
| "grad_norm": 0.2352762520313263, | |
| "learning_rate": 0.0003004525817732977, | |
| "loss": 0.0563, | |
| "step": 195000 | |
| }, | |
| { | |
| "epoch": 20.016423732293163, | |
| "eval_cer": 0.021178909952606635, | |
| "eval_loss": 0.0357496440410614, | |
| "eval_runtime": 22.3495, | |
| "eval_samples_per_second": 4.027, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.07059874888293119, | |
| "step": 195000 | |
| }, | |
| { | |
| "epoch": 20.119072059125436, | |
| "grad_norm": 0.31662923097610474, | |
| "learning_rate": 0.0002994239868339848, | |
| "loss": 0.0556, | |
| "step": 196000 | |
| }, | |
| { | |
| "epoch": 20.119072059125436, | |
| "eval_cer": 0.019697867298578198, | |
| "eval_loss": 0.03499302640557289, | |
| "eval_runtime": 22.3332, | |
| "eval_samples_per_second": 4.03, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06523681858802502, | |
| "step": 196000 | |
| }, | |
| { | |
| "epoch": 20.22172038595771, | |
| "grad_norm": 0.16036640107631683, | |
| "learning_rate": 0.0002983953918946719, | |
| "loss": 0.0556, | |
| "step": 197000 | |
| }, | |
| { | |
| "epoch": 20.22172038595771, | |
| "eval_cer": 0.019845971563981043, | |
| "eval_loss": 0.03571586683392525, | |
| "eval_runtime": 22.3554, | |
| "eval_samples_per_second": 4.026, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06970509383378017, | |
| "step": 197000 | |
| }, | |
| { | |
| "epoch": 20.324368712789983, | |
| "grad_norm": 0.18146023154258728, | |
| "learning_rate": 0.000297366796955359, | |
| "loss": 0.0556, | |
| "step": 198000 | |
| }, | |
| { | |
| "epoch": 20.324368712789983, | |
| "eval_cer": 0.01851303317535545, | |
| "eval_loss": 0.03470243141055107, | |
| "eval_runtime": 22.2669, | |
| "eval_samples_per_second": 4.042, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06523681858802502, | |
| "step": 198000 | |
| }, | |
| { | |
| "epoch": 20.427017039622253, | |
| "grad_norm": 0.18593771755695343, | |
| "learning_rate": 0.0002963382020160461, | |
| "loss": 0.0553, | |
| "step": 199000 | |
| }, | |
| { | |
| "epoch": 20.427017039622253, | |
| "eval_cer": 0.01762440758293839, | |
| "eval_loss": 0.034099601209163666, | |
| "eval_runtime": 22.1758, | |
| "eval_samples_per_second": 4.058, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06523681858802502, | |
| "step": 199000 | |
| }, | |
| { | |
| "epoch": 20.529665366454527, | |
| "grad_norm": 0.16802427172660828, | |
| "learning_rate": 0.00029530960707673323, | |
| "loss": 0.0558, | |
| "step": 200000 | |
| }, | |
| { | |
| "epoch": 20.529665366454527, | |
| "eval_cer": 0.016587677725118485, | |
| "eval_loss": 0.034865912050008774, | |
| "eval_runtime": 22.2605, | |
| "eval_samples_per_second": 4.043, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05987488829311886, | |
| "step": 200000 | |
| }, | |
| { | |
| "epoch": 20.6323136932868, | |
| "grad_norm": 0.20501871407032013, | |
| "learning_rate": 0.0002942810121374203, | |
| "loss": 0.0559, | |
| "step": 201000 | |
| }, | |
| { | |
| "epoch": 20.6323136932868, | |
| "eval_cer": 0.018957345971563982, | |
| "eval_loss": 0.03544703871011734, | |
| "eval_runtime": 22.0876, | |
| "eval_samples_per_second": 4.075, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.064343163538874, | |
| "step": 201000 | |
| }, | |
| { | |
| "epoch": 20.73496202011907, | |
| "grad_norm": 0.19251494109630585, | |
| "learning_rate": 0.0002932524171981074, | |
| "loss": 0.0555, | |
| "step": 202000 | |
| }, | |
| { | |
| "epoch": 20.73496202011907, | |
| "eval_cer": 0.01851303317535545, | |
| "eval_loss": 0.03462394326925278, | |
| "eval_runtime": 22.1969, | |
| "eval_samples_per_second": 4.055, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06166219839142091, | |
| "step": 202000 | |
| }, | |
| { | |
| "epoch": 20.837610346951344, | |
| "grad_norm": 0.12611526250839233, | |
| "learning_rate": 0.0002922238222587945, | |
| "loss": 0.0558, | |
| "step": 203000 | |
| }, | |
| { | |
| "epoch": 20.837610346951344, | |
| "eval_cer": 0.017920616113744077, | |
| "eval_loss": 0.03471866995096207, | |
| "eval_runtime": 22.9419, | |
| "eval_samples_per_second": 3.923, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.06344950848972297, | |
| "step": 203000 | |
| }, | |
| { | |
| "epoch": 20.940258673783617, | |
| "grad_norm": 0.23571810126304626, | |
| "learning_rate": 0.0002911952273194816, | |
| "loss": 0.0553, | |
| "step": 204000 | |
| }, | |
| { | |
| "epoch": 20.940258673783617, | |
| "eval_cer": 0.016735781990521326, | |
| "eval_loss": 0.033052537590265274, | |
| "eval_runtime": 22.2403, | |
| "eval_samples_per_second": 4.047, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05630026809651475, | |
| "step": 204000 | |
| }, | |
| { | |
| "epoch": 21.04290700061589, | |
| "grad_norm": 0.16654832661151886, | |
| "learning_rate": 0.0002901666323801687, | |
| "loss": 0.0552, | |
| "step": 205000 | |
| }, | |
| { | |
| "epoch": 21.04290700061589, | |
| "eval_cer": 0.017920616113744077, | |
| "eval_loss": 0.03388543054461479, | |
| "eval_runtime": 22.3538, | |
| "eval_samples_per_second": 4.026, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06613047363717604, | |
| "step": 205000 | |
| }, | |
| { | |
| "epoch": 21.14555532744816, | |
| "grad_norm": 0.30480700731277466, | |
| "learning_rate": 0.0002891380374408558, | |
| "loss": 0.054, | |
| "step": 206000 | |
| }, | |
| { | |
| "epoch": 21.14555532744816, | |
| "eval_cer": 0.017180094786729858, | |
| "eval_loss": 0.03281432017683983, | |
| "eval_runtime": 22.4896, | |
| "eval_samples_per_second": 4.002, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05540661304736372, | |
| "step": 206000 | |
| }, | |
| { | |
| "epoch": 21.248203654280434, | |
| "grad_norm": 0.22593766450881958, | |
| "learning_rate": 0.00028810944250154294, | |
| "loss": 0.0544, | |
| "step": 207000 | |
| }, | |
| { | |
| "epoch": 21.248203654280434, | |
| "eval_cer": 0.017920616113744077, | |
| "eval_loss": 0.03340643644332886, | |
| "eval_runtime": 22.4028, | |
| "eval_samples_per_second": 4.017, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06076854334226988, | |
| "step": 207000 | |
| }, | |
| { | |
| "epoch": 21.350851981112708, | |
| "grad_norm": 0.1935175508260727, | |
| "learning_rate": 0.00028708084756223003, | |
| "loss": 0.0549, | |
| "step": 208000 | |
| }, | |
| { | |
| "epoch": 21.350851981112708, | |
| "eval_cer": 0.018216824644549764, | |
| "eval_loss": 0.03223665431141853, | |
| "eval_runtime": 22.4082, | |
| "eval_samples_per_second": 4.016, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06255585344057193, | |
| "step": 208000 | |
| }, | |
| { | |
| "epoch": 21.45350030794498, | |
| "grad_norm": 0.18650276958942413, | |
| "learning_rate": 0.0002860522526229171, | |
| "loss": 0.0549, | |
| "step": 209000 | |
| }, | |
| { | |
| "epoch": 21.45350030794498, | |
| "eval_cer": 0.01762440758293839, | |
| "eval_loss": 0.0338866226375103, | |
| "eval_runtime": 22.2564, | |
| "eval_samples_per_second": 4.044, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05987488829311886, | |
| "step": 209000 | |
| }, | |
| { | |
| "epoch": 21.55614863477725, | |
| "grad_norm": 0.2135065495967865, | |
| "learning_rate": 0.0002850236576836042, | |
| "loss": 0.055, | |
| "step": 210000 | |
| }, | |
| { | |
| "epoch": 21.55614863477725, | |
| "eval_cer": 0.019845971563981043, | |
| "eval_loss": 0.033525411039590836, | |
| "eval_runtime": 22.1326, | |
| "eval_samples_per_second": 4.066, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06523681858802502, | |
| "step": 210000 | |
| }, | |
| { | |
| "epoch": 21.658796961609525, | |
| "grad_norm": 0.21206118166446686, | |
| "learning_rate": 0.0002839950627442913, | |
| "loss": 0.0547, | |
| "step": 211000 | |
| }, | |
| { | |
| "epoch": 21.658796961609525, | |
| "eval_cer": 0.01940165876777251, | |
| "eval_loss": 0.033743493258953094, | |
| "eval_runtime": 22.1212, | |
| "eval_samples_per_second": 4.068, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06613047363717604, | |
| "step": 211000 | |
| }, | |
| { | |
| "epoch": 21.7614452884418, | |
| "grad_norm": 0.41899242997169495, | |
| "learning_rate": 0.0002829664678049784, | |
| "loss": 0.0551, | |
| "step": 212000 | |
| }, | |
| { | |
| "epoch": 21.7614452884418, | |
| "eval_cer": 0.018661137440758292, | |
| "eval_loss": 0.03284008055925369, | |
| "eval_runtime": 22.3424, | |
| "eval_samples_per_second": 4.028, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06255585344057193, | |
| "step": 212000 | |
| }, | |
| { | |
| "epoch": 21.864093615274072, | |
| "grad_norm": 0.16063542664051056, | |
| "learning_rate": 0.0002819378728656655, | |
| "loss": 0.0547, | |
| "step": 213000 | |
| }, | |
| { | |
| "epoch": 21.864093615274072, | |
| "eval_cer": 0.017031990521327013, | |
| "eval_loss": 0.03267466276884079, | |
| "eval_runtime": 22.247, | |
| "eval_samples_per_second": 4.045, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05719392314566577, | |
| "step": 213000 | |
| }, | |
| { | |
| "epoch": 21.966741942106342, | |
| "grad_norm": 0.3092317283153534, | |
| "learning_rate": 0.00028090927792635264, | |
| "loss": 0.055, | |
| "step": 214000 | |
| }, | |
| { | |
| "epoch": 21.966741942106342, | |
| "eval_cer": 0.016291469194312798, | |
| "eval_loss": 0.032835327088832855, | |
| "eval_runtime": 22.1659, | |
| "eval_samples_per_second": 4.06, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0580875781948168, | |
| "step": 214000 | |
| }, | |
| { | |
| "epoch": 22.069390268938616, | |
| "grad_norm": 0.21150179207324982, | |
| "learning_rate": 0.00027988068298703974, | |
| "loss": 0.0544, | |
| "step": 215000 | |
| }, | |
| { | |
| "epoch": 22.069390268938616, | |
| "eval_cer": 0.018661137440758292, | |
| "eval_loss": 0.03344175964593887, | |
| "eval_runtime": 22.2595, | |
| "eval_samples_per_second": 4.043, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05987488829311886, | |
| "step": 215000 | |
| }, | |
| { | |
| "epoch": 22.17203859577089, | |
| "grad_norm": 0.7867759466171265, | |
| "learning_rate": 0.00027885208804772683, | |
| "loss": 0.0535, | |
| "step": 216000 | |
| }, | |
| { | |
| "epoch": 22.17203859577089, | |
| "eval_cer": 0.016587677725118485, | |
| "eval_loss": 0.03438210114836693, | |
| "eval_runtime": 22.4442, | |
| "eval_samples_per_second": 4.01, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06076854334226988, | |
| "step": 216000 | |
| }, | |
| { | |
| "epoch": 22.274686922603163, | |
| "grad_norm": 0.37553030252456665, | |
| "learning_rate": 0.0002778234931084139, | |
| "loss": 0.0546, | |
| "step": 217000 | |
| }, | |
| { | |
| "epoch": 22.274686922603163, | |
| "eval_cer": 0.018216824644549764, | |
| "eval_loss": 0.033362455666065216, | |
| "eval_runtime": 22.2715, | |
| "eval_samples_per_second": 4.041, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06523681858802502, | |
| "step": 217000 | |
| }, | |
| { | |
| "epoch": 22.377335249435433, | |
| "grad_norm": 0.29263338446617126, | |
| "learning_rate": 0.000276794898169101, | |
| "loss": 0.0537, | |
| "step": 218000 | |
| }, | |
| { | |
| "epoch": 22.377335249435433, | |
| "eval_cer": 0.017772511848341232, | |
| "eval_loss": 0.03228195384144783, | |
| "eval_runtime": 22.3277, | |
| "eval_samples_per_second": 4.031, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06166219839142091, | |
| "step": 218000 | |
| }, | |
| { | |
| "epoch": 22.479983576267706, | |
| "grad_norm": 0.25038591027259827, | |
| "learning_rate": 0.0002757663032297881, | |
| "loss": 0.0533, | |
| "step": 219000 | |
| }, | |
| { | |
| "epoch": 22.479983576267706, | |
| "eval_cer": 0.018957345971563982, | |
| "eval_loss": 0.033720944076776505, | |
| "eval_runtime": 22.5305, | |
| "eval_samples_per_second": 3.995, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.06702412868632708, | |
| "step": 219000 | |
| }, | |
| { | |
| "epoch": 22.58263190309998, | |
| "grad_norm": 0.1750280261039734, | |
| "learning_rate": 0.0002747377082904752, | |
| "loss": 0.055, | |
| "step": 220000 | |
| }, | |
| { | |
| "epoch": 22.58263190309998, | |
| "eval_cer": 0.016735781990521326, | |
| "eval_loss": 0.03310966119170189, | |
| "eval_runtime": 22.4023, | |
| "eval_samples_per_second": 4.017, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05987488829311886, | |
| "step": 220000 | |
| }, | |
| { | |
| "epoch": 22.685280229932253, | |
| "grad_norm": 0.24312840402126312, | |
| "learning_rate": 0.00027370911335116235, | |
| "loss": 0.0541, | |
| "step": 221000 | |
| }, | |
| { | |
| "epoch": 22.685280229932253, | |
| "eval_cer": 0.014958530805687204, | |
| "eval_loss": 0.03354882076382637, | |
| "eval_runtime": 22.1565, | |
| "eval_samples_per_second": 4.062, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05630026809651475, | |
| "step": 221000 | |
| }, | |
| { | |
| "epoch": 22.787928556764523, | |
| "grad_norm": 0.31420522928237915, | |
| "learning_rate": 0.00027268051841184944, | |
| "loss": 0.0541, | |
| "step": 222000 | |
| }, | |
| { | |
| "epoch": 22.787928556764523, | |
| "eval_cer": 0.015106635071090047, | |
| "eval_loss": 0.03195945546030998, | |
| "eval_runtime": 22.1178, | |
| "eval_samples_per_second": 4.069, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05540661304736372, | |
| "step": 222000 | |
| }, | |
| { | |
| "epoch": 22.890576883596797, | |
| "grad_norm": 0.5160701870918274, | |
| "learning_rate": 0.00027165192347253654, | |
| "loss": 0.0544, | |
| "step": 223000 | |
| }, | |
| { | |
| "epoch": 22.890576883596797, | |
| "eval_cer": 0.018661137440758292, | |
| "eval_loss": 0.0328957661986351, | |
| "eval_runtime": 22.2935, | |
| "eval_samples_per_second": 4.037, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.064343163538874, | |
| "step": 223000 | |
| }, | |
| { | |
| "epoch": 22.99322521042907, | |
| "grad_norm": 0.18564113974571228, | |
| "learning_rate": 0.00027062332853322363, | |
| "loss": 0.0538, | |
| "step": 224000 | |
| }, | |
| { | |
| "epoch": 22.99322521042907, | |
| "eval_cer": 0.019105450236966824, | |
| "eval_loss": 0.033330611884593964, | |
| "eval_runtime": 22.2627, | |
| "eval_samples_per_second": 4.043, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06344950848972297, | |
| "step": 224000 | |
| }, | |
| { | |
| "epoch": 23.095873537261344, | |
| "grad_norm": 0.19497232139110565, | |
| "learning_rate": 0.0002695947335939107, | |
| "loss": 0.0527, | |
| "step": 225000 | |
| }, | |
| { | |
| "epoch": 23.095873537261344, | |
| "eval_cer": 0.01643957345971564, | |
| "eval_loss": 0.03324893116950989, | |
| "eval_runtime": 22.2973, | |
| "eval_samples_per_second": 4.036, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06344950848972297, | |
| "step": 225000 | |
| }, | |
| { | |
| "epoch": 23.198521864093614, | |
| "grad_norm": 0.24956081807613373, | |
| "learning_rate": 0.0002685661386545978, | |
| "loss": 0.0529, | |
| "step": 226000 | |
| }, | |
| { | |
| "epoch": 23.198521864093614, | |
| "eval_cer": 0.017772511848341232, | |
| "eval_loss": 0.03351821005344391, | |
| "eval_runtime": 22.2956, | |
| "eval_samples_per_second": 4.037, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.058981233243967826, | |
| "step": 226000 | |
| }, | |
| { | |
| "epoch": 23.301170190925887, | |
| "grad_norm": 0.1977517157793045, | |
| "learning_rate": 0.00026753754371528496, | |
| "loss": 0.0532, | |
| "step": 227000 | |
| }, | |
| { | |
| "epoch": 23.301170190925887, | |
| "eval_cer": 0.017920616113744077, | |
| "eval_loss": 0.03354490175843239, | |
| "eval_runtime": 22.2726, | |
| "eval_samples_per_second": 4.041, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06523681858802502, | |
| "step": 227000 | |
| }, | |
| { | |
| "epoch": 23.40381851775816, | |
| "grad_norm": 0.23078420758247375, | |
| "learning_rate": 0.00026650894877597206, | |
| "loss": 0.0537, | |
| "step": 228000 | |
| }, | |
| { | |
| "epoch": 23.40381851775816, | |
| "eval_cer": 0.01688388625592417, | |
| "eval_loss": 0.03308222442865372, | |
| "eval_runtime": 22.3431, | |
| "eval_samples_per_second": 4.028, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.058981233243967826, | |
| "step": 228000 | |
| }, | |
| { | |
| "epoch": 23.506466844590435, | |
| "grad_norm": 0.3596203029155731, | |
| "learning_rate": 0.00026548035383665915, | |
| "loss": 0.0534, | |
| "step": 229000 | |
| }, | |
| { | |
| "epoch": 23.506466844590435, | |
| "eval_cer": 0.017031990521327013, | |
| "eval_loss": 0.03161655366420746, | |
| "eval_runtime": 22.3127, | |
| "eval_samples_per_second": 4.034, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0580875781948168, | |
| "step": 229000 | |
| }, | |
| { | |
| "epoch": 23.609115171422705, | |
| "grad_norm": 0.26955559849739075, | |
| "learning_rate": 0.00026445175889734624, | |
| "loss": 0.0538, | |
| "step": 230000 | |
| }, | |
| { | |
| "epoch": 23.609115171422705, | |
| "eval_cer": 0.01940165876777251, | |
| "eval_loss": 0.03247096389532089, | |
| "eval_runtime": 22.3142, | |
| "eval_samples_per_second": 4.033, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06523681858802502, | |
| "step": 230000 | |
| }, | |
| { | |
| "epoch": 23.711763498254978, | |
| "grad_norm": 0.305859237909317, | |
| "learning_rate": 0.00026342316395803334, | |
| "loss": 0.0534, | |
| "step": 231000 | |
| }, | |
| { | |
| "epoch": 23.711763498254978, | |
| "eval_cer": 0.017180094786729858, | |
| "eval_loss": 0.03188026696443558, | |
| "eval_runtime": 22.28, | |
| "eval_samples_per_second": 4.039, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05630026809651475, | |
| "step": 231000 | |
| }, | |
| { | |
| "epoch": 23.81441182508725, | |
| "grad_norm": 0.414869099855423, | |
| "learning_rate": 0.00026239456901872043, | |
| "loss": 0.0534, | |
| "step": 232000 | |
| }, | |
| { | |
| "epoch": 23.81441182508725, | |
| "eval_cer": 0.01940165876777251, | |
| "eval_loss": 0.033250004053115845, | |
| "eval_runtime": 22.3878, | |
| "eval_samples_per_second": 4.02, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06613047363717604, | |
| "step": 232000 | |
| }, | |
| { | |
| "epoch": 23.917060151919525, | |
| "grad_norm": 0.24189621210098267, | |
| "learning_rate": 0.0002613659740794075, | |
| "loss": 0.0537, | |
| "step": 233000 | |
| }, | |
| { | |
| "epoch": 23.917060151919525, | |
| "eval_cer": 0.018216824644549764, | |
| "eval_loss": 0.033227771520614624, | |
| "eval_runtime": 22.1167, | |
| "eval_samples_per_second": 4.069, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06255585344057193, | |
| "step": 233000 | |
| }, | |
| { | |
| "epoch": 24.019708478751795, | |
| "grad_norm": 0.19833236932754517, | |
| "learning_rate": 0.00026033737914009467, | |
| "loss": 0.0533, | |
| "step": 234000 | |
| }, | |
| { | |
| "epoch": 24.019708478751795, | |
| "eval_cer": 0.017920616113744077, | |
| "eval_loss": 0.03290673345327377, | |
| "eval_runtime": 22.278, | |
| "eval_samples_per_second": 4.04, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06076854334226988, | |
| "step": 234000 | |
| }, | |
| { | |
| "epoch": 24.12235680558407, | |
| "grad_norm": 0.15947362780570984, | |
| "learning_rate": 0.00025930878420078176, | |
| "loss": 0.0523, | |
| "step": 235000 | |
| }, | |
| { | |
| "epoch": 24.12235680558407, | |
| "eval_cer": 0.017772511848341232, | |
| "eval_loss": 0.03324710577726364, | |
| "eval_runtime": 22.2183, | |
| "eval_samples_per_second": 4.051, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06076854334226988, | |
| "step": 235000 | |
| }, | |
| { | |
| "epoch": 24.225005132416342, | |
| "grad_norm": 0.19958557188510895, | |
| "learning_rate": 0.00025828018926146886, | |
| "loss": 0.0522, | |
| "step": 236000 | |
| }, | |
| { | |
| "epoch": 24.225005132416342, | |
| "eval_cer": 0.015847156398104266, | |
| "eval_loss": 0.031826525926589966, | |
| "eval_runtime": 23.5228, | |
| "eval_samples_per_second": 3.826, | |
| "eval_steps_per_second": 0.043, | |
| "eval_wer": 0.05451295799821269, | |
| "step": 236000 | |
| }, | |
| { | |
| "epoch": 24.327653459248616, | |
| "grad_norm": 0.20118238031864166, | |
| "learning_rate": 0.00025725159432215595, | |
| "loss": 0.0518, | |
| "step": 237000 | |
| }, | |
| { | |
| "epoch": 24.327653459248616, | |
| "eval_cer": 0.014218009478672985, | |
| "eval_loss": 0.03189582750201225, | |
| "eval_runtime": 22.1576, | |
| "eval_samples_per_second": 4.062, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05451295799821269, | |
| "step": 237000 | |
| }, | |
| { | |
| "epoch": 24.430301786080886, | |
| "grad_norm": 0.2729335129261017, | |
| "learning_rate": 0.00025622299938284304, | |
| "loss": 0.0529, | |
| "step": 238000 | |
| }, | |
| { | |
| "epoch": 24.430301786080886, | |
| "eval_cer": 0.017180094786729858, | |
| "eval_loss": 0.03211754932999611, | |
| "eval_runtime": 22.2144, | |
| "eval_samples_per_second": 4.051, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06076854334226988, | |
| "step": 238000 | |
| }, | |
| { | |
| "epoch": 24.53295011291316, | |
| "grad_norm": 0.18871107697486877, | |
| "learning_rate": 0.00025519440444353014, | |
| "loss": 0.0531, | |
| "step": 239000 | |
| }, | |
| { | |
| "epoch": 24.53295011291316, | |
| "eval_cer": 0.01806872037914692, | |
| "eval_loss": 0.031803932040929794, | |
| "eval_runtime": 22.1636, | |
| "eval_samples_per_second": 4.061, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06166219839142091, | |
| "step": 239000 | |
| }, | |
| { | |
| "epoch": 24.635598439745433, | |
| "grad_norm": 0.22552721202373505, | |
| "learning_rate": 0.00025416580950421723, | |
| "loss": 0.053, | |
| "step": 240000 | |
| }, | |
| { | |
| "epoch": 24.635598439745433, | |
| "eval_cer": 0.018661137440758292, | |
| "eval_loss": 0.03302132338285446, | |
| "eval_runtime": 22.305, | |
| "eval_samples_per_second": 4.035, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06255585344057193, | |
| "step": 240000 | |
| }, | |
| { | |
| "epoch": 24.738246766577706, | |
| "grad_norm": 0.3143594264984131, | |
| "learning_rate": 0.0002531372145649044, | |
| "loss": 0.0535, | |
| "step": 241000 | |
| }, | |
| { | |
| "epoch": 24.738246766577706, | |
| "eval_cer": 0.017031990521327013, | |
| "eval_loss": 0.032325536012649536, | |
| "eval_runtime": 22.2087, | |
| "eval_samples_per_second": 4.052, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05451295799821269, | |
| "step": 241000 | |
| }, | |
| { | |
| "epoch": 24.840895093409976, | |
| "grad_norm": 0.2053222805261612, | |
| "learning_rate": 0.00025210861962559147, | |
| "loss": 0.053, | |
| "step": 242000 | |
| }, | |
| { | |
| "epoch": 24.840895093409976, | |
| "eval_cer": 0.01851303317535545, | |
| "eval_loss": 0.03247794508934021, | |
| "eval_runtime": 22.8241, | |
| "eval_samples_per_second": 3.943, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.06255585344057193, | |
| "step": 242000 | |
| }, | |
| { | |
| "epoch": 24.94354342024225, | |
| "grad_norm": 0.2229388952255249, | |
| "learning_rate": 0.00025108002468627856, | |
| "loss": 0.0525, | |
| "step": 243000 | |
| }, | |
| { | |
| "epoch": 24.94354342024225, | |
| "eval_cer": 0.020290284360189575, | |
| "eval_loss": 0.032272905111312866, | |
| "eval_runtime": 22.7477, | |
| "eval_samples_per_second": 3.956, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.06702412868632708, | |
| "step": 243000 | |
| }, | |
| { | |
| "epoch": 25.046191747074523, | |
| "grad_norm": 0.21171027421951294, | |
| "learning_rate": 0.00025005142974696566, | |
| "loss": 0.052, | |
| "step": 244000 | |
| }, | |
| { | |
| "epoch": 25.046191747074523, | |
| "eval_cer": 0.018809241706161137, | |
| "eval_loss": 0.03205866739153862, | |
| "eval_runtime": 22.1495, | |
| "eval_samples_per_second": 4.063, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06255585344057193, | |
| "step": 244000 | |
| }, | |
| { | |
| "epoch": 25.148840073906797, | |
| "grad_norm": 0.28519150614738464, | |
| "learning_rate": 0.00024902283480765275, | |
| "loss": 0.0518, | |
| "step": 245000 | |
| }, | |
| { | |
| "epoch": 25.148840073906797, | |
| "eval_cer": 0.017772511848341232, | |
| "eval_loss": 0.0315067283809185, | |
| "eval_runtime": 22.1659, | |
| "eval_samples_per_second": 4.06, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06255585344057193, | |
| "step": 245000 | |
| }, | |
| { | |
| "epoch": 25.251488400739067, | |
| "grad_norm": 0.19159762561321259, | |
| "learning_rate": 0.00024799423986833984, | |
| "loss": 0.0519, | |
| "step": 246000 | |
| }, | |
| { | |
| "epoch": 25.251488400739067, | |
| "eval_cer": 0.019105450236966824, | |
| "eval_loss": 0.03155896067619324, | |
| "eval_runtime": 22.1685, | |
| "eval_samples_per_second": 4.06, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06523681858802502, | |
| "step": 246000 | |
| }, | |
| { | |
| "epoch": 25.35413672757134, | |
| "grad_norm": 0.2302646040916443, | |
| "learning_rate": 0.00024696564492902694, | |
| "loss": 0.0521, | |
| "step": 247000 | |
| }, | |
| { | |
| "epoch": 25.35413672757134, | |
| "eval_cer": 0.018364928909952605, | |
| "eval_loss": 0.032100409269332886, | |
| "eval_runtime": 22.1736, | |
| "eval_samples_per_second": 4.059, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06344950848972297, | |
| "step": 247000 | |
| }, | |
| { | |
| "epoch": 25.456785054403614, | |
| "grad_norm": 0.2624431848526001, | |
| "learning_rate": 0.0002459370499897141, | |
| "loss": 0.0525, | |
| "step": 248000 | |
| }, | |
| { | |
| "epoch": 25.456785054403614, | |
| "eval_cer": 0.016587677725118485, | |
| "eval_loss": 0.03178829327225685, | |
| "eval_runtime": 22.2239, | |
| "eval_samples_per_second": 4.05, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.058981233243967826, | |
| "step": 248000 | |
| }, | |
| { | |
| "epoch": 25.559433381235884, | |
| "grad_norm": 0.24456505477428436, | |
| "learning_rate": 0.0002449084550504012, | |
| "loss": 0.052, | |
| "step": 249000 | |
| }, | |
| { | |
| "epoch": 25.559433381235884, | |
| "eval_cer": 0.01851303317535545, | |
| "eval_loss": 0.030579831451177597, | |
| "eval_runtime": 22.261, | |
| "eval_samples_per_second": 4.043, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06076854334226988, | |
| "step": 249000 | |
| }, | |
| { | |
| "epoch": 25.662081708068158, | |
| "grad_norm": 0.1688804030418396, | |
| "learning_rate": 0.00024387986011108827, | |
| "loss": 0.0518, | |
| "step": 250000 | |
| }, | |
| { | |
| "epoch": 25.662081708068158, | |
| "eval_cer": 0.01688388625592417, | |
| "eval_loss": 0.033347100019454956, | |
| "eval_runtime": 22.153, | |
| "eval_samples_per_second": 4.063, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05987488829311886, | |
| "step": 250000 | |
| }, | |
| { | |
| "epoch": 25.76473003490043, | |
| "grad_norm": 0.1677083969116211, | |
| "learning_rate": 0.00024285126517177536, | |
| "loss": 0.0524, | |
| "step": 251000 | |
| }, | |
| { | |
| "epoch": 25.76473003490043, | |
| "eval_cer": 0.017031990521327013, | |
| "eval_loss": 0.031150901690125465, | |
| "eval_runtime": 22.137, | |
| "eval_samples_per_second": 4.066, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05719392314566577, | |
| "step": 251000 | |
| }, | |
| { | |
| "epoch": 25.867378361732705, | |
| "grad_norm": 0.28190451860427856, | |
| "learning_rate": 0.00024182267023246246, | |
| "loss": 0.0525, | |
| "step": 252000 | |
| }, | |
| { | |
| "epoch": 25.867378361732705, | |
| "eval_cer": 0.014514218009478674, | |
| "eval_loss": 0.030852019786834717, | |
| "eval_runtime": 22.2042, | |
| "eval_samples_per_second": 4.053, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05272564789991063, | |
| "step": 252000 | |
| }, | |
| { | |
| "epoch": 25.970026688564978, | |
| "grad_norm": 0.30484601855278015, | |
| "learning_rate": 0.00024079407529314955, | |
| "loss": 0.0525, | |
| "step": 253000 | |
| }, | |
| { | |
| "epoch": 25.970026688564978, | |
| "eval_cer": 0.01481042654028436, | |
| "eval_loss": 0.03129878640174866, | |
| "eval_runtime": 22.1837, | |
| "eval_samples_per_second": 4.057, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05540661304736372, | |
| "step": 253000 | |
| }, | |
| { | |
| "epoch": 26.072675015397248, | |
| "grad_norm": 0.23301398754119873, | |
| "learning_rate": 0.00023976548035383667, | |
| "loss": 0.0511, | |
| "step": 254000 | |
| }, | |
| { | |
| "epoch": 26.072675015397248, | |
| "eval_cer": 0.016587677725118485, | |
| "eval_loss": 0.03187458962202072, | |
| "eval_runtime": 22.2688, | |
| "eval_samples_per_second": 4.042, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0580875781948168, | |
| "step": 254000 | |
| }, | |
| { | |
| "epoch": 26.17532334222952, | |
| "grad_norm": 0.2110058218240738, | |
| "learning_rate": 0.00023873688541452376, | |
| "loss": 0.051, | |
| "step": 255000 | |
| }, | |
| { | |
| "epoch": 26.17532334222952, | |
| "eval_cer": 0.01688388625592417, | |
| "eval_loss": 0.03192685917019844, | |
| "eval_runtime": 22.1497, | |
| "eval_samples_per_second": 4.063, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06344950848972297, | |
| "step": 255000 | |
| }, | |
| { | |
| "epoch": 26.277971669061795, | |
| "grad_norm": 0.7681686282157898, | |
| "learning_rate": 0.00023770829047521086, | |
| "loss": 0.0512, | |
| "step": 256000 | |
| }, | |
| { | |
| "epoch": 26.277971669061795, | |
| "eval_cer": 0.017772511848341232, | |
| "eval_loss": 0.03258216753602028, | |
| "eval_runtime": 22.2138, | |
| "eval_samples_per_second": 4.052, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.064343163538874, | |
| "step": 256000 | |
| }, | |
| { | |
| "epoch": 26.380619995894065, | |
| "grad_norm": 0.20167267322540283, | |
| "learning_rate": 0.00023667969553589798, | |
| "loss": 0.0511, | |
| "step": 257000 | |
| }, | |
| { | |
| "epoch": 26.380619995894065, | |
| "eval_cer": 0.015847156398104266, | |
| "eval_loss": 0.03150052949786186, | |
| "eval_runtime": 22.1067, | |
| "eval_samples_per_second": 4.071, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06166219839142091, | |
| "step": 257000 | |
| }, | |
| { | |
| "epoch": 26.48326832272634, | |
| "grad_norm": 0.2660065293312073, | |
| "learning_rate": 0.00023565110059658507, | |
| "loss": 0.0514, | |
| "step": 258000 | |
| }, | |
| { | |
| "epoch": 26.48326832272634, | |
| "eval_cer": 0.017031990521327013, | |
| "eval_loss": 0.031788378953933716, | |
| "eval_runtime": 22.0549, | |
| "eval_samples_per_second": 4.081, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0580875781948168, | |
| "step": 258000 | |
| }, | |
| { | |
| "epoch": 26.585916649558612, | |
| "grad_norm": 0.18703380227088928, | |
| "learning_rate": 0.00023462250565727216, | |
| "loss": 0.0516, | |
| "step": 259000 | |
| }, | |
| { | |
| "epoch": 26.585916649558612, | |
| "eval_cer": 0.016587677725118485, | |
| "eval_loss": 0.03131980076432228, | |
| "eval_runtime": 22.1633, | |
| "eval_samples_per_second": 4.061, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.058981233243967826, | |
| "step": 259000 | |
| }, | |
| { | |
| "epoch": 26.688564976390886, | |
| "grad_norm": 0.25654709339141846, | |
| "learning_rate": 0.00023359391071795926, | |
| "loss": 0.0517, | |
| "step": 260000 | |
| }, | |
| { | |
| "epoch": 26.688564976390886, | |
| "eval_cer": 0.016291469194312798, | |
| "eval_loss": 0.031445086002349854, | |
| "eval_runtime": 22.3924, | |
| "eval_samples_per_second": 4.019, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05987488829311886, | |
| "step": 260000 | |
| }, | |
| { | |
| "epoch": 26.791213303223156, | |
| "grad_norm": 0.2620410919189453, | |
| "learning_rate": 0.00023256531577864638, | |
| "loss": 0.052, | |
| "step": 261000 | |
| }, | |
| { | |
| "epoch": 26.791213303223156, | |
| "eval_cer": 0.017031990521327013, | |
| "eval_loss": 0.03067016415297985, | |
| "eval_runtime": 22.1951, | |
| "eval_samples_per_second": 4.055, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06076854334226988, | |
| "step": 261000 | |
| }, | |
| { | |
| "epoch": 26.89386163005543, | |
| "grad_norm": 0.2175012230873108, | |
| "learning_rate": 0.00023153672083933347, | |
| "loss": 0.0519, | |
| "step": 262000 | |
| }, | |
| { | |
| "epoch": 26.89386163005543, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.031155884265899658, | |
| "eval_runtime": 22.1754, | |
| "eval_samples_per_second": 4.059, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05272564789991063, | |
| "step": 262000 | |
| }, | |
| { | |
| "epoch": 26.996509956887703, | |
| "grad_norm": 0.22425027191638947, | |
| "learning_rate": 0.00023050812590002056, | |
| "loss": 0.052, | |
| "step": 263000 | |
| }, | |
| { | |
| "epoch": 26.996509956887703, | |
| "eval_cer": 0.016587677725118485, | |
| "eval_loss": 0.030132591724395752, | |
| "eval_runtime": 22.1695, | |
| "eval_samples_per_second": 4.06, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05719392314566577, | |
| "step": 263000 | |
| }, | |
| { | |
| "epoch": 27.099158283719976, | |
| "grad_norm": 0.24267776310443878, | |
| "learning_rate": 0.00022947953096070768, | |
| "loss": 0.0506, | |
| "step": 264000 | |
| }, | |
| { | |
| "epoch": 27.099158283719976, | |
| "eval_cer": 0.0173281990521327, | |
| "eval_loss": 0.030183136463165283, | |
| "eval_runtime": 22.1929, | |
| "eval_samples_per_second": 4.055, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0580875781948168, | |
| "step": 264000 | |
| }, | |
| { | |
| "epoch": 27.201806610552246, | |
| "grad_norm": 0.22535988688468933, | |
| "learning_rate": 0.00022845093602139478, | |
| "loss": 0.051, | |
| "step": 265000 | |
| }, | |
| { | |
| "epoch": 27.201806610552246, | |
| "eval_cer": 0.015550947867298577, | |
| "eval_loss": 0.029872052371501923, | |
| "eval_runtime": 21.6648, | |
| "eval_samples_per_second": 4.154, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.05272564789991063, | |
| "step": 265000 | |
| }, | |
| { | |
| "epoch": 27.30445493738452, | |
| "grad_norm": 0.2646799385547638, | |
| "learning_rate": 0.00022742234108208187, | |
| "loss": 0.0511, | |
| "step": 266000 | |
| }, | |
| { | |
| "epoch": 27.30445493738452, | |
| "eval_cer": 0.016143364928909953, | |
| "eval_loss": 0.030452899634838104, | |
| "eval_runtime": 21.9979, | |
| "eval_samples_per_second": 4.091, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0580875781948168, | |
| "step": 266000 | |
| }, | |
| { | |
| "epoch": 27.407103264216794, | |
| "grad_norm": 0.1818641871213913, | |
| "learning_rate": 0.00022639374614276896, | |
| "loss": 0.0507, | |
| "step": 267000 | |
| }, | |
| { | |
| "epoch": 27.407103264216794, | |
| "eval_cer": 0.01806872037914692, | |
| "eval_loss": 0.030228691175580025, | |
| "eval_runtime": 22.1513, | |
| "eval_samples_per_second": 4.063, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06166219839142091, | |
| "step": 267000 | |
| }, | |
| { | |
| "epoch": 27.509751591049067, | |
| "grad_norm": 0.2914126515388489, | |
| "learning_rate": 0.00022536515120345608, | |
| "loss": 0.0506, | |
| "step": 268000 | |
| }, | |
| { | |
| "epoch": 27.509751591049067, | |
| "eval_cer": 0.01688388625592417, | |
| "eval_loss": 0.030619405210018158, | |
| "eval_runtime": 22.1314, | |
| "eval_samples_per_second": 4.067, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06344950848972297, | |
| "step": 268000 | |
| }, | |
| { | |
| "epoch": 27.612399917881337, | |
| "grad_norm": 0.19746644794940948, | |
| "learning_rate": 0.00022433655626414317, | |
| "loss": 0.0508, | |
| "step": 269000 | |
| }, | |
| { | |
| "epoch": 27.612399917881337, | |
| "eval_cer": 0.014218009478672985, | |
| "eval_loss": 0.029848448932170868, | |
| "eval_runtime": 22.0615, | |
| "eval_samples_per_second": 4.08, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05451295799821269, | |
| "step": 269000 | |
| }, | |
| { | |
| "epoch": 27.71504824471361, | |
| "grad_norm": 0.171453595161438, | |
| "learning_rate": 0.00022330796132483027, | |
| "loss": 0.0506, | |
| "step": 270000 | |
| }, | |
| { | |
| "epoch": 27.71504824471361, | |
| "eval_cer": 0.01643957345971564, | |
| "eval_loss": 0.029385404661297798, | |
| "eval_runtime": 22.1939, | |
| "eval_samples_per_second": 4.055, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.058981233243967826, | |
| "step": 270000 | |
| }, | |
| { | |
| "epoch": 27.817696571545884, | |
| "grad_norm": 0.20320715010166168, | |
| "learning_rate": 0.0002222793663855174, | |
| "loss": 0.0513, | |
| "step": 271000 | |
| }, | |
| { | |
| "epoch": 27.817696571545884, | |
| "eval_cer": 0.016735781990521326, | |
| "eval_loss": 0.03053821623325348, | |
| "eval_runtime": 22.1393, | |
| "eval_samples_per_second": 4.065, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06255585344057193, | |
| "step": 271000 | |
| }, | |
| { | |
| "epoch": 27.920344898378158, | |
| "grad_norm": 0.18147552013397217, | |
| "learning_rate": 0.00022125077144620448, | |
| "loss": 0.051, | |
| "step": 272000 | |
| }, | |
| { | |
| "epoch": 27.920344898378158, | |
| "eval_cer": 0.015847156398104266, | |
| "eval_loss": 0.030883438885211945, | |
| "eval_runtime": 22.1075, | |
| "eval_samples_per_second": 4.071, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05630026809651475, | |
| "step": 272000 | |
| }, | |
| { | |
| "epoch": 28.022993225210428, | |
| "grad_norm": 0.3934711813926697, | |
| "learning_rate": 0.00022022217650689157, | |
| "loss": 0.0505, | |
| "step": 273000 | |
| }, | |
| { | |
| "epoch": 28.022993225210428, | |
| "eval_cer": 0.015995260663507108, | |
| "eval_loss": 0.030014000833034515, | |
| "eval_runtime": 21.9105, | |
| "eval_samples_per_second": 4.108, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.05451295799821269, | |
| "step": 273000 | |
| }, | |
| { | |
| "epoch": 28.1256415520427, | |
| "grad_norm": 0.18721525371074677, | |
| "learning_rate": 0.0002191935815675787, | |
| "loss": 0.0495, | |
| "step": 274000 | |
| }, | |
| { | |
| "epoch": 28.1256415520427, | |
| "eval_cer": 0.015106635071090047, | |
| "eval_loss": 0.029803840443491936, | |
| "eval_runtime": 22.2218, | |
| "eval_samples_per_second": 4.05, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 274000 | |
| }, | |
| { | |
| "epoch": 28.228289878874975, | |
| "grad_norm": 0.2836057245731354, | |
| "learning_rate": 0.0002181649866282658, | |
| "loss": 0.0499, | |
| "step": 275000 | |
| }, | |
| { | |
| "epoch": 28.228289878874975, | |
| "eval_cer": 0.014366113744075829, | |
| "eval_loss": 0.02985943853855133, | |
| "eval_runtime": 22.1574, | |
| "eval_samples_per_second": 4.062, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05183199285075961, | |
| "step": 275000 | |
| }, | |
| { | |
| "epoch": 28.33093820570725, | |
| "grad_norm": 0.16741269826889038, | |
| "learning_rate": 0.00021713639168895288, | |
| "loss": 0.0503, | |
| "step": 276000 | |
| }, | |
| { | |
| "epoch": 28.33093820570725, | |
| "eval_cer": 0.018661137440758292, | |
| "eval_loss": 0.02978348545730114, | |
| "eval_runtime": 22.2061, | |
| "eval_samples_per_second": 4.053, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.058981233243967826, | |
| "step": 276000 | |
| }, | |
| { | |
| "epoch": 28.43358653253952, | |
| "grad_norm": 0.5261670351028442, | |
| "learning_rate": 0.00021610779674963997, | |
| "loss": 0.0505, | |
| "step": 277000 | |
| }, | |
| { | |
| "epoch": 28.43358653253952, | |
| "eval_cer": 0.016143364928909953, | |
| "eval_loss": 0.02990272268652916, | |
| "eval_runtime": 22.1117, | |
| "eval_samples_per_second": 4.07, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05540661304736372, | |
| "step": 277000 | |
| }, | |
| { | |
| "epoch": 28.536234859371792, | |
| "grad_norm": 0.18356911838054657, | |
| "learning_rate": 0.0002150792018103271, | |
| "loss": 0.0502, | |
| "step": 278000 | |
| }, | |
| { | |
| "epoch": 28.536234859371792, | |
| "eval_cer": 0.017476303317535545, | |
| "eval_loss": 0.0308319590985775, | |
| "eval_runtime": 22.2031, | |
| "eval_samples_per_second": 4.053, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05540661304736372, | |
| "step": 278000 | |
| }, | |
| { | |
| "epoch": 28.638883186204065, | |
| "grad_norm": 0.2313164621591568, | |
| "learning_rate": 0.0002140506068710142, | |
| "loss": 0.0506, | |
| "step": 279000 | |
| }, | |
| { | |
| "epoch": 28.638883186204065, | |
| "eval_cer": 0.016735781990521326, | |
| "eval_loss": 0.03133101388812065, | |
| "eval_runtime": 22.2245, | |
| "eval_samples_per_second": 4.05, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06255585344057193, | |
| "step": 279000 | |
| }, | |
| { | |
| "epoch": 28.74153151303634, | |
| "grad_norm": 0.23044399917125702, | |
| "learning_rate": 0.00021302201193170128, | |
| "loss": 0.0511, | |
| "step": 280000 | |
| }, | |
| { | |
| "epoch": 28.74153151303634, | |
| "eval_cer": 0.01762440758293839, | |
| "eval_loss": 0.0300216656178236, | |
| "eval_runtime": 22.3006, | |
| "eval_samples_per_second": 4.036, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0580875781948168, | |
| "step": 280000 | |
| }, | |
| { | |
| "epoch": 28.84417983986861, | |
| "grad_norm": 0.35165879130363464, | |
| "learning_rate": 0.0002119934169923884, | |
| "loss": 0.051, | |
| "step": 281000 | |
| }, | |
| { | |
| "epoch": 28.84417983986861, | |
| "eval_cer": 0.0173281990521327, | |
| "eval_loss": 0.029326628893613815, | |
| "eval_runtime": 22.3265, | |
| "eval_samples_per_second": 4.031, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05630026809651475, | |
| "step": 281000 | |
| }, | |
| { | |
| "epoch": 28.946828166700882, | |
| "grad_norm": 0.276239275932312, | |
| "learning_rate": 0.00021096482205307552, | |
| "loss": 0.0499, | |
| "step": 282000 | |
| }, | |
| { | |
| "epoch": 28.946828166700882, | |
| "eval_cer": 0.015847156398104266, | |
| "eval_loss": 0.02999110147356987, | |
| "eval_runtime": 22.3215, | |
| "eval_samples_per_second": 4.032, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05540661304736372, | |
| "step": 282000 | |
| }, | |
| { | |
| "epoch": 29.049476493533156, | |
| "grad_norm": 0.2091301679611206, | |
| "learning_rate": 0.00020993622711376261, | |
| "loss": 0.0494, | |
| "step": 283000 | |
| }, | |
| { | |
| "epoch": 29.049476493533156, | |
| "eval_cer": 0.016735781990521326, | |
| "eval_loss": 0.029503343626856804, | |
| "eval_runtime": 22.2699, | |
| "eval_samples_per_second": 4.041, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05630026809651475, | |
| "step": 283000 | |
| }, | |
| { | |
| "epoch": 29.15212482036543, | |
| "grad_norm": 0.3076883852481842, | |
| "learning_rate": 0.0002089076321744497, | |
| "loss": 0.0493, | |
| "step": 284000 | |
| }, | |
| { | |
| "epoch": 29.15212482036543, | |
| "eval_cer": 0.01806872037914692, | |
| "eval_loss": 0.02992323227226734, | |
| "eval_runtime": 22.4692, | |
| "eval_samples_per_second": 4.005, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0580875781948168, | |
| "step": 284000 | |
| }, | |
| { | |
| "epoch": 29.2547731471977, | |
| "grad_norm": 0.2164120078086853, | |
| "learning_rate": 0.00020787903723513683, | |
| "loss": 0.0492, | |
| "step": 285000 | |
| }, | |
| { | |
| "epoch": 29.2547731471977, | |
| "eval_cer": 0.017772511848341232, | |
| "eval_loss": 0.029783058911561966, | |
| "eval_runtime": 22.2084, | |
| "eval_samples_per_second": 4.053, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05987488829311886, | |
| "step": 285000 | |
| }, | |
| { | |
| "epoch": 29.357421474029973, | |
| "grad_norm": 0.25016504526138306, | |
| "learning_rate": 0.00020685044229582392, | |
| "loss": 0.0497, | |
| "step": 286000 | |
| }, | |
| { | |
| "epoch": 29.357421474029973, | |
| "eval_cer": 0.016587677725118485, | |
| "eval_loss": 0.029911022633314133, | |
| "eval_runtime": 22.0994, | |
| "eval_samples_per_second": 4.073, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05987488829311886, | |
| "step": 286000 | |
| }, | |
| { | |
| "epoch": 29.460069800862247, | |
| "grad_norm": 0.2327207326889038, | |
| "learning_rate": 0.00020582184735651101, | |
| "loss": 0.0496, | |
| "step": 287000 | |
| }, | |
| { | |
| "epoch": 29.460069800862247, | |
| "eval_cer": 0.016291469194312798, | |
| "eval_loss": 0.028591720387339592, | |
| "eval_runtime": 22.1807, | |
| "eval_samples_per_second": 4.058, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05540661304736372, | |
| "step": 287000 | |
| }, | |
| { | |
| "epoch": 29.56271812769452, | |
| "grad_norm": 0.25889137387275696, | |
| "learning_rate": 0.00020479325241719813, | |
| "loss": 0.0497, | |
| "step": 288000 | |
| }, | |
| { | |
| "epoch": 29.56271812769452, | |
| "eval_cer": 0.015995260663507108, | |
| "eval_loss": 0.02914293482899666, | |
| "eval_runtime": 22.2487, | |
| "eval_samples_per_second": 4.045, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 288000 | |
| }, | |
| { | |
| "epoch": 29.66536645452679, | |
| "grad_norm": 0.3326428532600403, | |
| "learning_rate": 0.00020376465747788523, | |
| "loss": 0.0499, | |
| "step": 289000 | |
| }, | |
| { | |
| "epoch": 29.66536645452679, | |
| "eval_cer": 0.015402843601895734, | |
| "eval_loss": 0.028966935351490974, | |
| "eval_runtime": 22.2836, | |
| "eval_samples_per_second": 4.039, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05630026809651475, | |
| "step": 289000 | |
| }, | |
| { | |
| "epoch": 29.768014781359064, | |
| "grad_norm": 0.2747463583946228, | |
| "learning_rate": 0.00020273606253857232, | |
| "loss": 0.0498, | |
| "step": 290000 | |
| }, | |
| { | |
| "epoch": 29.768014781359064, | |
| "eval_cer": 0.015402843601895734, | |
| "eval_loss": 0.027468033134937286, | |
| "eval_runtime": 22.3116, | |
| "eval_samples_per_second": 4.034, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05630026809651475, | |
| "step": 290000 | |
| }, | |
| { | |
| "epoch": 29.870663108191337, | |
| "grad_norm": 0.31372368335723877, | |
| "learning_rate": 0.00020170746759925941, | |
| "loss": 0.0495, | |
| "step": 291000 | |
| }, | |
| { | |
| "epoch": 29.870663108191337, | |
| "eval_cer": 0.01569905213270142, | |
| "eval_loss": 0.029326878488063812, | |
| "eval_runtime": 22.3423, | |
| "eval_samples_per_second": 4.028, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05630026809651475, | |
| "step": 291000 | |
| }, | |
| { | |
| "epoch": 29.97331143502361, | |
| "grad_norm": 0.18704882264137268, | |
| "learning_rate": 0.00020067887265994653, | |
| "loss": 0.0499, | |
| "step": 292000 | |
| }, | |
| { | |
| "epoch": 29.97331143502361, | |
| "eval_cer": 0.01481042654028436, | |
| "eval_loss": 0.028479211032390594, | |
| "eval_runtime": 22.4625, | |
| "eval_samples_per_second": 4.007, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 292000 | |
| }, | |
| { | |
| "epoch": 30.07595976185588, | |
| "grad_norm": 0.1708535999059677, | |
| "learning_rate": 0.00019965027772063363, | |
| "loss": 0.0484, | |
| "step": 293000 | |
| }, | |
| { | |
| "epoch": 30.07595976185588, | |
| "eval_cer": 0.013921800947867298, | |
| "eval_loss": 0.028360920026898384, | |
| "eval_runtime": 22.3021, | |
| "eval_samples_per_second": 4.036, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 293000 | |
| }, | |
| { | |
| "epoch": 30.178608088688154, | |
| "grad_norm": 0.24666380882263184, | |
| "learning_rate": 0.00019862168278132072, | |
| "loss": 0.0489, | |
| "step": 294000 | |
| }, | |
| { | |
| "epoch": 30.178608088688154, | |
| "eval_cer": 0.016291469194312798, | |
| "eval_loss": 0.028621409088373184, | |
| "eval_runtime": 22.2968, | |
| "eval_samples_per_second": 4.036, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05719392314566577, | |
| "step": 294000 | |
| }, | |
| { | |
| "epoch": 30.281256415520428, | |
| "grad_norm": 0.2784259021282196, | |
| "learning_rate": 0.00019759308784200784, | |
| "loss": 0.0493, | |
| "step": 295000 | |
| }, | |
| { | |
| "epoch": 30.281256415520428, | |
| "eval_cer": 0.016143364928909953, | |
| "eval_loss": 0.028690271079540253, | |
| "eval_runtime": 22.252, | |
| "eval_samples_per_second": 4.045, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05451295799821269, | |
| "step": 295000 | |
| }, | |
| { | |
| "epoch": 30.3839047423527, | |
| "grad_norm": 0.2628447711467743, | |
| "learning_rate": 0.00019656449290269493, | |
| "loss": 0.049, | |
| "step": 296000 | |
| }, | |
| { | |
| "epoch": 30.3839047423527, | |
| "eval_cer": 0.016291469194312798, | |
| "eval_loss": 0.028396843001246452, | |
| "eval_runtime": 22.2552, | |
| "eval_samples_per_second": 4.044, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05451295799821269, | |
| "step": 296000 | |
| }, | |
| { | |
| "epoch": 30.48655306918497, | |
| "grad_norm": 0.23084846138954163, | |
| "learning_rate": 0.00019553589796338203, | |
| "loss": 0.049, | |
| "step": 297000 | |
| }, | |
| { | |
| "epoch": 30.48655306918497, | |
| "eval_cer": 0.01481042654028436, | |
| "eval_loss": 0.02870321460068226, | |
| "eval_runtime": 22.2205, | |
| "eval_samples_per_second": 4.05, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05451295799821269, | |
| "step": 297000 | |
| }, | |
| { | |
| "epoch": 30.589201396017245, | |
| "grad_norm": 0.17921754717826843, | |
| "learning_rate": 0.00019450730302406912, | |
| "loss": 0.0491, | |
| "step": 298000 | |
| }, | |
| { | |
| "epoch": 30.589201396017245, | |
| "eval_cer": 0.014069905213270142, | |
| "eval_loss": 0.028362760320305824, | |
| "eval_runtime": 22.2536, | |
| "eval_samples_per_second": 4.044, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05272564789991063, | |
| "step": 298000 | |
| }, | |
| { | |
| "epoch": 30.69184972284952, | |
| "grad_norm": 0.20431461930274963, | |
| "learning_rate": 0.00019347870808475624, | |
| "loss": 0.0493, | |
| "step": 299000 | |
| }, | |
| { | |
| "epoch": 30.69184972284952, | |
| "eval_cer": 0.014366113744075829, | |
| "eval_loss": 0.027963554486632347, | |
| "eval_runtime": 22.2764, | |
| "eval_samples_per_second": 4.04, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05451295799821269, | |
| "step": 299000 | |
| }, | |
| { | |
| "epoch": 30.794498049681792, | |
| "grad_norm": 0.3255954384803772, | |
| "learning_rate": 0.00019245011314544333, | |
| "loss": 0.049, | |
| "step": 300000 | |
| }, | |
| { | |
| "epoch": 30.794498049681792, | |
| "eval_cer": 0.016143364928909953, | |
| "eval_loss": 0.028848888352513313, | |
| "eval_runtime": 22.3367, | |
| "eval_samples_per_second": 4.029, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05540661304736372, | |
| "step": 300000 | |
| }, | |
| { | |
| "epoch": 30.897146376514062, | |
| "grad_norm": 0.30805402994155884, | |
| "learning_rate": 0.00019142151820613043, | |
| "loss": 0.0493, | |
| "step": 301000 | |
| }, | |
| { | |
| "epoch": 30.897146376514062, | |
| "eval_cer": 0.014958530805687204, | |
| "eval_loss": 0.029449112713336945, | |
| "eval_runtime": 22.2797, | |
| "eval_samples_per_second": 4.04, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05630026809651475, | |
| "step": 301000 | |
| }, | |
| { | |
| "epoch": 30.999794703346335, | |
| "grad_norm": 0.2879369854927063, | |
| "learning_rate": 0.00019039292326681755, | |
| "loss": 0.0491, | |
| "step": 302000 | |
| }, | |
| { | |
| "epoch": 30.999794703346335, | |
| "eval_cer": 0.015847156398104266, | |
| "eval_loss": 0.028681093826889992, | |
| "eval_runtime": 22.3072, | |
| "eval_samples_per_second": 4.035, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05451295799821269, | |
| "step": 302000 | |
| }, | |
| { | |
| "epoch": 31.10244303017861, | |
| "grad_norm": 0.30376702547073364, | |
| "learning_rate": 0.00018936432832750464, | |
| "loss": 0.0479, | |
| "step": 303000 | |
| }, | |
| { | |
| "epoch": 31.10244303017861, | |
| "eval_cer": 0.013773696682464455, | |
| "eval_loss": 0.028124256059527397, | |
| "eval_runtime": 22.255, | |
| "eval_samples_per_second": 4.044, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 303000 | |
| }, | |
| { | |
| "epoch": 31.20509135701088, | |
| "grad_norm": 0.325859934091568, | |
| "learning_rate": 0.00018833573338819173, | |
| "loss": 0.0485, | |
| "step": 304000 | |
| }, | |
| { | |
| "epoch": 31.20509135701088, | |
| "eval_cer": 0.013625592417061612, | |
| "eval_loss": 0.028593741357326508, | |
| "eval_runtime": 22.3606, | |
| "eval_samples_per_second": 4.025, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05183199285075961, | |
| "step": 304000 | |
| }, | |
| { | |
| "epoch": 31.307739683843153, | |
| "grad_norm": 0.26860424876213074, | |
| "learning_rate": 0.00018730713844887883, | |
| "loss": 0.0481, | |
| "step": 305000 | |
| }, | |
| { | |
| "epoch": 31.307739683843153, | |
| "eval_cer": 0.015402843601895734, | |
| "eval_loss": 0.028675682842731476, | |
| "eval_runtime": 22.1875, | |
| "eval_samples_per_second": 4.056, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05272564789991063, | |
| "step": 305000 | |
| }, | |
| { | |
| "epoch": 31.410388010675426, | |
| "grad_norm": 0.36149585247039795, | |
| "learning_rate": 0.00018627854350956595, | |
| "loss": 0.0488, | |
| "step": 306000 | |
| }, | |
| { | |
| "epoch": 31.410388010675426, | |
| "eval_cer": 0.015995260663507108, | |
| "eval_loss": 0.029292631894350052, | |
| "eval_runtime": 22.1759, | |
| "eval_samples_per_second": 4.058, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05630026809651475, | |
| "step": 306000 | |
| }, | |
| { | |
| "epoch": 31.5130363375077, | |
| "grad_norm": 0.23009520769119263, | |
| "learning_rate": 0.00018524994857025304, | |
| "loss": 0.0487, | |
| "step": 307000 | |
| }, | |
| { | |
| "epoch": 31.5130363375077, | |
| "eval_cer": 0.014366113744075829, | |
| "eval_loss": 0.02981569990515709, | |
| "eval_runtime": 22.1972, | |
| "eval_samples_per_second": 4.055, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05272564789991063, | |
| "step": 307000 | |
| }, | |
| { | |
| "epoch": 31.61568466433997, | |
| "grad_norm": 0.2854170799255371, | |
| "learning_rate": 0.00018422135363094013, | |
| "loss": 0.0485, | |
| "step": 308000 | |
| }, | |
| { | |
| "epoch": 31.61568466433997, | |
| "eval_cer": 0.014958530805687204, | |
| "eval_loss": 0.030133897438645363, | |
| "eval_runtime": 22.3301, | |
| "eval_samples_per_second": 4.03, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05451295799821269, | |
| "step": 308000 | |
| }, | |
| { | |
| "epoch": 31.718332991172243, | |
| "grad_norm": 0.20701636373996735, | |
| "learning_rate": 0.00018319275869162725, | |
| "loss": 0.0488, | |
| "step": 309000 | |
| }, | |
| { | |
| "epoch": 31.718332991172243, | |
| "eval_cer": 0.016143364928909953, | |
| "eval_loss": 0.02945251390337944, | |
| "eval_runtime": 22.1144, | |
| "eval_samples_per_second": 4.07, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05540661304736372, | |
| "step": 309000 | |
| }, | |
| { | |
| "epoch": 31.820981318004517, | |
| "grad_norm": 0.16278359293937683, | |
| "learning_rate": 0.00018216416375231435, | |
| "loss": 0.0489, | |
| "step": 310000 | |
| }, | |
| { | |
| "epoch": 31.820981318004517, | |
| "eval_cer": 0.015106635071090047, | |
| "eval_loss": 0.02939271740615368, | |
| "eval_runtime": 22.2354, | |
| "eval_samples_per_second": 4.048, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 310000 | |
| }, | |
| { | |
| "epoch": 31.92362964483679, | |
| "grad_norm": 0.20761480927467346, | |
| "learning_rate": 0.00018113556881300144, | |
| "loss": 0.049, | |
| "step": 311000 | |
| }, | |
| { | |
| "epoch": 31.92362964483679, | |
| "eval_cer": 0.015847156398104266, | |
| "eval_loss": 0.028488388285040855, | |
| "eval_runtime": 22.1425, | |
| "eval_samples_per_second": 4.065, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05272564789991063, | |
| "step": 311000 | |
| }, | |
| { | |
| "epoch": 32.02627797166906, | |
| "grad_norm": 0.23206296563148499, | |
| "learning_rate": 0.00018010697387368853, | |
| "loss": 0.0487, | |
| "step": 312000 | |
| }, | |
| { | |
| "epoch": 32.02627797166906, | |
| "eval_cer": 0.017180094786729858, | |
| "eval_loss": 0.028869740664958954, | |
| "eval_runtime": 22.3356, | |
| "eval_samples_per_second": 4.029, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05719392314566577, | |
| "step": 312000 | |
| }, | |
| { | |
| "epoch": 32.128926298501334, | |
| "grad_norm": 0.3341107964515686, | |
| "learning_rate": 0.00017907837893437565, | |
| "loss": 0.0481, | |
| "step": 313000 | |
| }, | |
| { | |
| "epoch": 32.128926298501334, | |
| "eval_cer": 0.018216824644549764, | |
| "eval_loss": 0.029275845736265182, | |
| "eval_runtime": 22.1486, | |
| "eval_samples_per_second": 4.063, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06166219839142091, | |
| "step": 313000 | |
| }, | |
| { | |
| "epoch": 32.23157462533361, | |
| "grad_norm": 0.240467831492424, | |
| "learning_rate": 0.00017804978399506275, | |
| "loss": 0.0481, | |
| "step": 314000 | |
| }, | |
| { | |
| "epoch": 32.23157462533361, | |
| "eval_cer": 0.017031990521327013, | |
| "eval_loss": 0.02915882132947445, | |
| "eval_runtime": 22.1429, | |
| "eval_samples_per_second": 4.065, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05630026809651475, | |
| "step": 314000 | |
| }, | |
| { | |
| "epoch": 32.33422295216588, | |
| "grad_norm": 0.24573862552642822, | |
| "learning_rate": 0.00017702118905574984, | |
| "loss": 0.0473, | |
| "step": 315000 | |
| }, | |
| { | |
| "epoch": 32.33422295216588, | |
| "eval_cer": 0.018364928909952605, | |
| "eval_loss": 0.02838301472365856, | |
| "eval_runtime": 22.1497, | |
| "eval_samples_per_second": 4.063, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05719392314566577, | |
| "step": 315000 | |
| }, | |
| { | |
| "epoch": 32.436871278998154, | |
| "grad_norm": 0.3885030746459961, | |
| "learning_rate": 0.00017599259411643696, | |
| "loss": 0.0486, | |
| "step": 316000 | |
| }, | |
| { | |
| "epoch": 32.436871278998154, | |
| "eval_cer": 0.01762440758293839, | |
| "eval_loss": 0.02944045141339302, | |
| "eval_runtime": 22.1602, | |
| "eval_samples_per_second": 4.061, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0580875781948168, | |
| "step": 316000 | |
| }, | |
| { | |
| "epoch": 32.53951960583043, | |
| "grad_norm": 0.31944355368614197, | |
| "learning_rate": 0.00017496399917712405, | |
| "loss": 0.0483, | |
| "step": 317000 | |
| }, | |
| { | |
| "epoch": 32.53951960583043, | |
| "eval_cer": 0.0173281990521327, | |
| "eval_loss": 0.028936417773365974, | |
| "eval_runtime": 21.9567, | |
| "eval_samples_per_second": 4.099, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.05630026809651475, | |
| "step": 317000 | |
| }, | |
| { | |
| "epoch": 32.642167932662694, | |
| "grad_norm": 0.2156781703233719, | |
| "learning_rate": 0.00017393540423781115, | |
| "loss": 0.0477, | |
| "step": 318000 | |
| }, | |
| { | |
| "epoch": 32.642167932662694, | |
| "eval_cer": 0.018661137440758292, | |
| "eval_loss": 0.029529759660363197, | |
| "eval_runtime": 22.0507, | |
| "eval_samples_per_second": 4.082, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.06076854334226988, | |
| "step": 318000 | |
| }, | |
| { | |
| "epoch": 32.74481625949497, | |
| "grad_norm": 0.23771491646766663, | |
| "learning_rate": 0.00017290680929849824, | |
| "loss": 0.048, | |
| "step": 319000 | |
| }, | |
| { | |
| "epoch": 32.74481625949497, | |
| "eval_cer": 0.016291469194312798, | |
| "eval_loss": 0.029350074008107185, | |
| "eval_runtime": 22.1715, | |
| "eval_samples_per_second": 4.059, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05272564789991063, | |
| "step": 319000 | |
| }, | |
| { | |
| "epoch": 32.84746458632724, | |
| "grad_norm": 0.4010453224182129, | |
| "learning_rate": 0.00017187821435918536, | |
| "loss": 0.0482, | |
| "step": 320000 | |
| }, | |
| { | |
| "epoch": 32.84746458632724, | |
| "eval_cer": 0.016587677725118485, | |
| "eval_loss": 0.0294723492115736, | |
| "eval_runtime": 22.2642, | |
| "eval_samples_per_second": 4.042, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05451295799821269, | |
| "step": 320000 | |
| }, | |
| { | |
| "epoch": 32.950112913159515, | |
| "grad_norm": 0.21573138236999512, | |
| "learning_rate": 0.00017084961941987245, | |
| "loss": 0.0478, | |
| "step": 321000 | |
| }, | |
| { | |
| "epoch": 32.950112913159515, | |
| "eval_cer": 0.01643957345971564, | |
| "eval_loss": 0.02849040925502777, | |
| "eval_runtime": 22.3919, | |
| "eval_samples_per_second": 4.019, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05540661304736372, | |
| "step": 321000 | |
| }, | |
| { | |
| "epoch": 33.05276123999179, | |
| "grad_norm": 0.2565110921859741, | |
| "learning_rate": 0.00016982102448055955, | |
| "loss": 0.0476, | |
| "step": 322000 | |
| }, | |
| { | |
| "epoch": 33.05276123999179, | |
| "eval_cer": 0.01643957345971564, | |
| "eval_loss": 0.028644192963838577, | |
| "eval_runtime": 22.4917, | |
| "eval_samples_per_second": 4.001, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 322000 | |
| }, | |
| { | |
| "epoch": 33.15540956682406, | |
| "grad_norm": 0.17566250264644623, | |
| "learning_rate": 0.00016879242954124667, | |
| "loss": 0.0465, | |
| "step": 323000 | |
| }, | |
| { | |
| "epoch": 33.15540956682406, | |
| "eval_cer": 0.015402843601895734, | |
| "eval_loss": 0.028031960129737854, | |
| "eval_runtime": 22.2193, | |
| "eval_samples_per_second": 4.051, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05183199285075961, | |
| "step": 323000 | |
| }, | |
| { | |
| "epoch": 33.258057893656336, | |
| "grad_norm": 0.21690410375595093, | |
| "learning_rate": 0.00016776383460193376, | |
| "loss": 0.0474, | |
| "step": 324000 | |
| }, | |
| { | |
| "epoch": 33.258057893656336, | |
| "eval_cer": 0.01569905213270142, | |
| "eval_loss": 0.028126152232289314, | |
| "eval_runtime": 22.2514, | |
| "eval_samples_per_second": 4.045, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 324000 | |
| }, | |
| { | |
| "epoch": 33.36070622048861, | |
| "grad_norm": 0.17805682122707367, | |
| "learning_rate": 0.00016673523966262085, | |
| "loss": 0.047, | |
| "step": 325000 | |
| }, | |
| { | |
| "epoch": 33.36070622048861, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.028366120532155037, | |
| "eval_runtime": 22.1824, | |
| "eval_samples_per_second": 4.057, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 325000 | |
| }, | |
| { | |
| "epoch": 33.463354547320876, | |
| "grad_norm": 0.23276859521865845, | |
| "learning_rate": 0.00016570664472330795, | |
| "loss": 0.0478, | |
| "step": 326000 | |
| }, | |
| { | |
| "epoch": 33.463354547320876, | |
| "eval_cer": 0.01643957345971564, | |
| "eval_loss": 0.02805442176759243, | |
| "eval_runtime": 22.0401, | |
| "eval_samples_per_second": 4.083, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 326000 | |
| }, | |
| { | |
| "epoch": 33.56600287415315, | |
| "grad_norm": 0.23791708052158356, | |
| "learning_rate": 0.00016467804978399507, | |
| "loss": 0.0479, | |
| "step": 327000 | |
| }, | |
| { | |
| "epoch": 33.56600287415315, | |
| "eval_cer": 0.015402843601895734, | |
| "eval_loss": 0.028071463108062744, | |
| "eval_runtime": 22.166, | |
| "eval_samples_per_second": 4.06, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05272564789991063, | |
| "step": 327000 | |
| }, | |
| { | |
| "epoch": 33.66865120098542, | |
| "grad_norm": 0.22921526432037354, | |
| "learning_rate": 0.00016364945484468216, | |
| "loss": 0.0482, | |
| "step": 328000 | |
| }, | |
| { | |
| "epoch": 33.66865120098542, | |
| "eval_cer": 0.01643957345971564, | |
| "eval_loss": 0.028569117188453674, | |
| "eval_runtime": 22.3334, | |
| "eval_samples_per_second": 4.03, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05272564789991063, | |
| "step": 328000 | |
| }, | |
| { | |
| "epoch": 33.771299527817696, | |
| "grad_norm": 0.1470087319612503, | |
| "learning_rate": 0.00016262085990536925, | |
| "loss": 0.0477, | |
| "step": 329000 | |
| }, | |
| { | |
| "epoch": 33.771299527817696, | |
| "eval_cer": 0.01569905213270142, | |
| "eval_loss": 0.028614189475774765, | |
| "eval_runtime": 22.1814, | |
| "eval_samples_per_second": 4.057, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 329000 | |
| }, | |
| { | |
| "epoch": 33.87394785464997, | |
| "grad_norm": 0.4257276654243469, | |
| "learning_rate": 0.00016159226496605637, | |
| "loss": 0.0483, | |
| "step": 330000 | |
| }, | |
| { | |
| "epoch": 33.87394785464997, | |
| "eval_cer": 0.01643957345971564, | |
| "eval_loss": 0.028485840186476707, | |
| "eval_runtime": 22.174, | |
| "eval_samples_per_second": 4.059, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 330000 | |
| }, | |
| { | |
| "epoch": 33.97659618148224, | |
| "grad_norm": 0.4129483997821808, | |
| "learning_rate": 0.00016056367002674347, | |
| "loss": 0.0475, | |
| "step": 331000 | |
| }, | |
| { | |
| "epoch": 33.97659618148224, | |
| "eval_cer": 0.01525473933649289, | |
| "eval_loss": 0.028352849185466766, | |
| "eval_runtime": 22.3086, | |
| "eval_samples_per_second": 4.034, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05183199285075961, | |
| "step": 331000 | |
| }, | |
| { | |
| "epoch": 34.07924450831452, | |
| "grad_norm": 0.2835679054260254, | |
| "learning_rate": 0.00015953507508743056, | |
| "loss": 0.0465, | |
| "step": 332000 | |
| }, | |
| { | |
| "epoch": 34.07924450831452, | |
| "eval_cer": 0.014514218009478674, | |
| "eval_loss": 0.028890669345855713, | |
| "eval_runtime": 22.1774, | |
| "eval_samples_per_second": 4.058, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 332000 | |
| }, | |
| { | |
| "epoch": 34.18189283514679, | |
| "grad_norm": 0.2247968167066574, | |
| "learning_rate": 0.00015850648014811765, | |
| "loss": 0.0469, | |
| "step": 333000 | |
| }, | |
| { | |
| "epoch": 34.18189283514679, | |
| "eval_cer": 0.015106635071090047, | |
| "eval_loss": 0.028413381427526474, | |
| "eval_runtime": 22.0968, | |
| "eval_samples_per_second": 4.073, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 333000 | |
| }, | |
| { | |
| "epoch": 34.28454116197906, | |
| "grad_norm": 0.17359092831611633, | |
| "learning_rate": 0.00015747788520880477, | |
| "loss": 0.0468, | |
| "step": 334000 | |
| }, | |
| { | |
| "epoch": 34.28454116197906, | |
| "eval_cer": 0.015550947867298577, | |
| "eval_loss": 0.028406651690602303, | |
| "eval_runtime": 22.1838, | |
| "eval_samples_per_second": 4.057, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 334000 | |
| }, | |
| { | |
| "epoch": 34.38718948881133, | |
| "grad_norm": 0.2905976176261902, | |
| "learning_rate": 0.00015644929026949187, | |
| "loss": 0.0467, | |
| "step": 335000 | |
| }, | |
| { | |
| "epoch": 34.38718948881133, | |
| "eval_cer": 0.01569905213270142, | |
| "eval_loss": 0.028134917840361595, | |
| "eval_runtime": 22.2384, | |
| "eval_samples_per_second": 4.047, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05630026809651475, | |
| "step": 335000 | |
| }, | |
| { | |
| "epoch": 34.489837815643604, | |
| "grad_norm": 0.41442832350730896, | |
| "learning_rate": 0.00015542069533017896, | |
| "loss": 0.0472, | |
| "step": 336000 | |
| }, | |
| { | |
| "epoch": 34.489837815643604, | |
| "eval_cer": 0.016587677725118485, | |
| "eval_loss": 0.028330376371741295, | |
| "eval_runtime": 22.2855, | |
| "eval_samples_per_second": 4.038, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05540661304736372, | |
| "step": 336000 | |
| }, | |
| { | |
| "epoch": 34.59248614247588, | |
| "grad_norm": 0.20874008536338806, | |
| "learning_rate": 0.00015439210039086608, | |
| "loss": 0.0472, | |
| "step": 337000 | |
| }, | |
| { | |
| "epoch": 34.59248614247588, | |
| "eval_cer": 0.01688388625592417, | |
| "eval_loss": 0.028587637469172478, | |
| "eval_runtime": 22.2778, | |
| "eval_samples_per_second": 4.04, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05540661304736372, | |
| "step": 337000 | |
| }, | |
| { | |
| "epoch": 34.69513446930815, | |
| "grad_norm": 0.28286242485046387, | |
| "learning_rate": 0.00015336350545155317, | |
| "loss": 0.047, | |
| "step": 338000 | |
| }, | |
| { | |
| "epoch": 34.69513446930815, | |
| "eval_cer": 0.016587677725118485, | |
| "eval_loss": 0.028073778375983238, | |
| "eval_runtime": 22.2502, | |
| "eval_samples_per_second": 4.045, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05540661304736372, | |
| "step": 338000 | |
| }, | |
| { | |
| "epoch": 34.797782796140424, | |
| "grad_norm": 0.2530520558357239, | |
| "learning_rate": 0.00015233491051224027, | |
| "loss": 0.0469, | |
| "step": 339000 | |
| }, | |
| { | |
| "epoch": 34.797782796140424, | |
| "eval_cer": 0.015402843601895734, | |
| "eval_loss": 0.02780935913324356, | |
| "eval_runtime": 22.1698, | |
| "eval_samples_per_second": 4.06, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05183199285075961, | |
| "step": 339000 | |
| }, | |
| { | |
| "epoch": 34.9004311229727, | |
| "grad_norm": 0.24278897047042847, | |
| "learning_rate": 0.00015130631557292736, | |
| "loss": 0.0474, | |
| "step": 340000 | |
| }, | |
| { | |
| "epoch": 34.9004311229727, | |
| "eval_cer": 0.01569905213270142, | |
| "eval_loss": 0.027951352298259735, | |
| "eval_runtime": 22.2193, | |
| "eval_samples_per_second": 4.051, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05540661304736372, | |
| "step": 340000 | |
| }, | |
| { | |
| "epoch": 35.00307944980497, | |
| "grad_norm": 0.23415499925613403, | |
| "learning_rate": 0.00015027772063361448, | |
| "loss": 0.0469, | |
| "step": 341000 | |
| }, | |
| { | |
| "epoch": 35.00307944980497, | |
| "eval_cer": 0.015995260663507108, | |
| "eval_loss": 0.02772090956568718, | |
| "eval_runtime": 22.09, | |
| "eval_samples_per_second": 4.074, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05719392314566577, | |
| "step": 341000 | |
| }, | |
| { | |
| "epoch": 35.10572777663724, | |
| "grad_norm": 0.4553733468055725, | |
| "learning_rate": 0.00014924912569430157, | |
| "loss": 0.0458, | |
| "step": 342000 | |
| }, | |
| { | |
| "epoch": 35.10572777663724, | |
| "eval_cer": 0.013773696682464455, | |
| "eval_loss": 0.027903633192181587, | |
| "eval_runtime": 22.1864, | |
| "eval_samples_per_second": 4.057, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 342000 | |
| }, | |
| { | |
| "epoch": 35.20837610346951, | |
| "grad_norm": 0.1911894530057907, | |
| "learning_rate": 0.00014822053075498867, | |
| "loss": 0.0464, | |
| "step": 343000 | |
| }, | |
| { | |
| "epoch": 35.20837610346951, | |
| "eval_cer": 0.015402843601895734, | |
| "eval_loss": 0.0284242145717144, | |
| "eval_runtime": 22.2316, | |
| "eval_samples_per_second": 4.048, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 343000 | |
| }, | |
| { | |
| "epoch": 35.311024430301785, | |
| "grad_norm": 0.9561129212379456, | |
| "learning_rate": 0.0001471919358156758, | |
| "loss": 0.0463, | |
| "step": 344000 | |
| }, | |
| { | |
| "epoch": 35.311024430301785, | |
| "eval_cer": 0.01569905213270142, | |
| "eval_loss": 0.02789863385260105, | |
| "eval_runtime": 22.2194, | |
| "eval_samples_per_second": 4.051, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05451295799821269, | |
| "step": 344000 | |
| }, | |
| { | |
| "epoch": 35.41367275713406, | |
| "grad_norm": 0.2921608090400696, | |
| "learning_rate": 0.0001461633408763629, | |
| "loss": 0.0461, | |
| "step": 345000 | |
| }, | |
| { | |
| "epoch": 35.41367275713406, | |
| "eval_cer": 0.015995260663507108, | |
| "eval_loss": 0.028261249884963036, | |
| "eval_runtime": 22.0199, | |
| "eval_samples_per_second": 4.087, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05451295799821269, | |
| "step": 345000 | |
| }, | |
| { | |
| "epoch": 35.51632108396633, | |
| "grad_norm": 0.19372744858264923, | |
| "learning_rate": 0.00014513474593705, | |
| "loss": 0.0463, | |
| "step": 346000 | |
| }, | |
| { | |
| "epoch": 35.51632108396633, | |
| "eval_cer": 0.01688388625592417, | |
| "eval_loss": 0.028103064745664597, | |
| "eval_runtime": 22.1334, | |
| "eval_samples_per_second": 4.066, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.0580875781948168, | |
| "step": 346000 | |
| }, | |
| { | |
| "epoch": 35.618969410798606, | |
| "grad_norm": 0.27431151270866394, | |
| "learning_rate": 0.00014410615099773712, | |
| "loss": 0.0472, | |
| "step": 347000 | |
| }, | |
| { | |
| "epoch": 35.618969410798606, | |
| "eval_cer": 0.016587677725118485, | |
| "eval_loss": 0.028425684198737144, | |
| "eval_runtime": 22.139, | |
| "eval_samples_per_second": 4.065, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05630026809651475, | |
| "step": 347000 | |
| }, | |
| { | |
| "epoch": 35.72161773763088, | |
| "grad_norm": 0.27128133177757263, | |
| "learning_rate": 0.0001430775560584242, | |
| "loss": 0.0466, | |
| "step": 348000 | |
| }, | |
| { | |
| "epoch": 35.72161773763088, | |
| "eval_cer": 0.015847156398104266, | |
| "eval_loss": 0.02764798142015934, | |
| "eval_runtime": 22.2733, | |
| "eval_samples_per_second": 4.041, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05451295799821269, | |
| "step": 348000 | |
| }, | |
| { | |
| "epoch": 35.824266064463146, | |
| "grad_norm": 0.26995572447776794, | |
| "learning_rate": 0.0001420489611191113, | |
| "loss": 0.0467, | |
| "step": 349000 | |
| }, | |
| { | |
| "epoch": 35.824266064463146, | |
| "eval_cer": 0.01569905213270142, | |
| "eval_loss": 0.027596063911914825, | |
| "eval_runtime": 22.0467, | |
| "eval_samples_per_second": 4.082, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05272564789991063, | |
| "step": 349000 | |
| }, | |
| { | |
| "epoch": 35.92691439129542, | |
| "grad_norm": 0.31239375472068787, | |
| "learning_rate": 0.0001410203661797984, | |
| "loss": 0.0469, | |
| "step": 350000 | |
| }, | |
| { | |
| "epoch": 35.92691439129542, | |
| "eval_cer": 0.015402843601895734, | |
| "eval_loss": 0.028232304379343987, | |
| "eval_runtime": 22.1495, | |
| "eval_samples_per_second": 4.063, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05183199285075961, | |
| "step": 350000 | |
| }, | |
| { | |
| "epoch": 36.02956271812769, | |
| "grad_norm": 0.344926655292511, | |
| "learning_rate": 0.00013999177124048552, | |
| "loss": 0.0464, | |
| "step": 351000 | |
| }, | |
| { | |
| "epoch": 36.02956271812769, | |
| "eval_cer": 0.013773696682464455, | |
| "eval_loss": 0.026858482509851456, | |
| "eval_runtime": 22.2106, | |
| "eval_samples_per_second": 4.052, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 351000 | |
| }, | |
| { | |
| "epoch": 36.132211044959966, | |
| "grad_norm": 0.1864170879125595, | |
| "learning_rate": 0.0001389631763011726, | |
| "loss": 0.0449, | |
| "step": 352000 | |
| }, | |
| { | |
| "epoch": 36.132211044959966, | |
| "eval_cer": 0.014366113744075829, | |
| "eval_loss": 0.02711603231728077, | |
| "eval_runtime": 22.2552, | |
| "eval_samples_per_second": 4.044, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 352000 | |
| }, | |
| { | |
| "epoch": 36.23485937179224, | |
| "grad_norm": 0.1876976191997528, | |
| "learning_rate": 0.0001379345813618597, | |
| "loss": 0.0466, | |
| "step": 353000 | |
| }, | |
| { | |
| "epoch": 36.23485937179224, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.027281379327178, | |
| "eval_runtime": 22.2442, | |
| "eval_samples_per_second": 4.046, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 353000 | |
| }, | |
| { | |
| "epoch": 36.33750769862451, | |
| "grad_norm": 0.21546737849712372, | |
| "learning_rate": 0.00013690598642254683, | |
| "loss": 0.0458, | |
| "step": 354000 | |
| }, | |
| { | |
| "epoch": 36.33750769862451, | |
| "eval_cer": 0.015106635071090047, | |
| "eval_loss": 0.02689436264336109, | |
| "eval_runtime": 22.259, | |
| "eval_samples_per_second": 4.043, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 354000 | |
| }, | |
| { | |
| "epoch": 36.44015602545679, | |
| "grad_norm": 0.31887394189834595, | |
| "learning_rate": 0.00013587739148323392, | |
| "loss": 0.0459, | |
| "step": 355000 | |
| }, | |
| { | |
| "epoch": 36.44015602545679, | |
| "eval_cer": 0.014218009478672985, | |
| "eval_loss": 0.027051741257309914, | |
| "eval_runtime": 22.2336, | |
| "eval_samples_per_second": 4.048, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 355000 | |
| }, | |
| { | |
| "epoch": 36.54280435228906, | |
| "grad_norm": 0.24578991532325745, | |
| "learning_rate": 0.000134848796543921, | |
| "loss": 0.0469, | |
| "step": 356000 | |
| }, | |
| { | |
| "epoch": 36.54280435228906, | |
| "eval_cer": 0.016291469194312798, | |
| "eval_loss": 0.026666434481739998, | |
| "eval_runtime": 22.215, | |
| "eval_samples_per_second": 4.051, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 356000 | |
| }, | |
| { | |
| "epoch": 36.64545267912133, | |
| "grad_norm": 0.3394581377506256, | |
| "learning_rate": 0.0001338202016046081, | |
| "loss": 0.0458, | |
| "step": 357000 | |
| }, | |
| { | |
| "epoch": 36.64545267912133, | |
| "eval_cer": 0.01481042654028436, | |
| "eval_loss": 0.02698771469295025, | |
| "eval_runtime": 22.1744, | |
| "eval_samples_per_second": 4.059, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 357000 | |
| }, | |
| { | |
| "epoch": 36.7481010059536, | |
| "grad_norm": 0.2731720805168152, | |
| "learning_rate": 0.00013279160666529523, | |
| "loss": 0.0462, | |
| "step": 358000 | |
| }, | |
| { | |
| "epoch": 36.7481010059536, | |
| "eval_cer": 0.015402843601895734, | |
| "eval_loss": 0.02696968987584114, | |
| "eval_runtime": 22.1592, | |
| "eval_samples_per_second": 4.062, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 358000 | |
| }, | |
| { | |
| "epoch": 36.850749332785874, | |
| "grad_norm": 0.2410440891981125, | |
| "learning_rate": 0.00013176301172598232, | |
| "loss": 0.0466, | |
| "step": 359000 | |
| }, | |
| { | |
| "epoch": 36.850749332785874, | |
| "eval_cer": 0.016143364928909953, | |
| "eval_loss": 0.027307961136102676, | |
| "eval_runtime": 22.243, | |
| "eval_samples_per_second": 4.046, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 359000 | |
| }, | |
| { | |
| "epoch": 36.95339765961815, | |
| "grad_norm": 0.28017568588256836, | |
| "learning_rate": 0.0001307344167866694, | |
| "loss": 0.046, | |
| "step": 360000 | |
| }, | |
| { | |
| "epoch": 36.95339765961815, | |
| "eval_cer": 0.014958530805687204, | |
| "eval_loss": 0.02654258720576763, | |
| "eval_runtime": 22.4072, | |
| "eval_samples_per_second": 4.017, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05183199285075961, | |
| "step": 360000 | |
| }, | |
| { | |
| "epoch": 37.05604598645042, | |
| "grad_norm": 0.23387791216373444, | |
| "learning_rate": 0.00012970582184735653, | |
| "loss": 0.0456, | |
| "step": 361000 | |
| }, | |
| { | |
| "epoch": 37.05604598645042, | |
| "eval_cer": 0.014366113744075829, | |
| "eval_loss": 0.026880960911512375, | |
| "eval_runtime": 22.1133, | |
| "eval_samples_per_second": 4.07, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 361000 | |
| }, | |
| { | |
| "epoch": 37.158694313282695, | |
| "grad_norm": 0.3157537579536438, | |
| "learning_rate": 0.00012867722690804363, | |
| "loss": 0.0452, | |
| "step": 362000 | |
| }, | |
| { | |
| "epoch": 37.158694313282695, | |
| "eval_cer": 0.015106635071090047, | |
| "eval_loss": 0.026554275304079056, | |
| "eval_runtime": 22.2184, | |
| "eval_samples_per_second": 4.051, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 362000 | |
| }, | |
| { | |
| "epoch": 37.26134264011497, | |
| "grad_norm": 0.301200807094574, | |
| "learning_rate": 0.00012764863196873072, | |
| "loss": 0.0456, | |
| "step": 363000 | |
| }, | |
| { | |
| "epoch": 37.26134264011497, | |
| "eval_cer": 0.01481042654028436, | |
| "eval_loss": 0.02748226933181286, | |
| "eval_runtime": 22.1594, | |
| "eval_samples_per_second": 4.061, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 363000 | |
| }, | |
| { | |
| "epoch": 37.36399096694724, | |
| "grad_norm": 0.1987874060869217, | |
| "learning_rate": 0.0001266200370294178, | |
| "loss": 0.0454, | |
| "step": 364000 | |
| }, | |
| { | |
| "epoch": 37.36399096694724, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.02689102478325367, | |
| "eval_runtime": 22.289, | |
| "eval_samples_per_second": 4.038, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 364000 | |
| }, | |
| { | |
| "epoch": 37.46663929377951, | |
| "grad_norm": 0.2465968132019043, | |
| "learning_rate": 0.00012559144209010493, | |
| "loss": 0.0453, | |
| "step": 365000 | |
| }, | |
| { | |
| "epoch": 37.46663929377951, | |
| "eval_cer": 0.012588862559241706, | |
| "eval_loss": 0.026751089841127396, | |
| "eval_runtime": 21.9569, | |
| "eval_samples_per_second": 4.099, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.045576407506702415, | |
| "step": 365000 | |
| }, | |
| { | |
| "epoch": 37.56928762061178, | |
| "grad_norm": 0.39610666036605835, | |
| "learning_rate": 0.00012456284715079203, | |
| "loss": 0.046, | |
| "step": 366000 | |
| }, | |
| { | |
| "epoch": 37.56928762061178, | |
| "eval_cer": 0.014366113744075829, | |
| "eval_loss": 0.026830825954675674, | |
| "eval_runtime": 22.2783, | |
| "eval_samples_per_second": 4.04, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 366000 | |
| }, | |
| { | |
| "epoch": 37.671935947444055, | |
| "grad_norm": 0.26581278443336487, | |
| "learning_rate": 0.00012353425221147912, | |
| "loss": 0.046, | |
| "step": 367000 | |
| }, | |
| { | |
| "epoch": 37.671935947444055, | |
| "eval_cer": 0.014514218009478674, | |
| "eval_loss": 0.0267048142850399, | |
| "eval_runtime": 22.1983, | |
| "eval_samples_per_second": 4.054, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 367000 | |
| }, | |
| { | |
| "epoch": 37.77458427427633, | |
| "grad_norm": 0.22986672818660736, | |
| "learning_rate": 0.00012250565727216624, | |
| "loss": 0.0459, | |
| "step": 368000 | |
| }, | |
| { | |
| "epoch": 37.77458427427633, | |
| "eval_cer": 0.013181279620853081, | |
| "eval_loss": 0.026945000514388084, | |
| "eval_runtime": 22.1927, | |
| "eval_samples_per_second": 4.055, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.045576407506702415, | |
| "step": 368000 | |
| }, | |
| { | |
| "epoch": 37.8772326011086, | |
| "grad_norm": 0.2647237479686737, | |
| "learning_rate": 0.00012147706233285333, | |
| "loss": 0.0454, | |
| "step": 369000 | |
| }, | |
| { | |
| "epoch": 37.8772326011086, | |
| "eval_cer": 0.014366113744075829, | |
| "eval_loss": 0.027104683220386505, | |
| "eval_runtime": 22.1006, | |
| "eval_samples_per_second": 4.072, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 369000 | |
| }, | |
| { | |
| "epoch": 37.979880927940876, | |
| "grad_norm": 0.2698921263217926, | |
| "learning_rate": 0.00012044846739354043, | |
| "loss": 0.0459, | |
| "step": 370000 | |
| }, | |
| { | |
| "epoch": 37.979880927940876, | |
| "eval_cer": 0.014366113744075829, | |
| "eval_loss": 0.027364252135157585, | |
| "eval_runtime": 22.1262, | |
| "eval_samples_per_second": 4.068, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 370000 | |
| }, | |
| { | |
| "epoch": 38.08252925477315, | |
| "grad_norm": 0.2091304063796997, | |
| "learning_rate": 0.00011941987245422753, | |
| "loss": 0.0453, | |
| "step": 371000 | |
| }, | |
| { | |
| "epoch": 38.08252925477315, | |
| "eval_cer": 0.013329383886255925, | |
| "eval_loss": 0.026948757469654083, | |
| "eval_runtime": 22.0181, | |
| "eval_samples_per_second": 4.088, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 371000 | |
| }, | |
| { | |
| "epoch": 38.18517758160542, | |
| "grad_norm": 0.21031425893306732, | |
| "learning_rate": 0.00011839127751491462, | |
| "loss": 0.0446, | |
| "step": 372000 | |
| }, | |
| { | |
| "epoch": 38.18517758160542, | |
| "eval_cer": 0.014069905213270142, | |
| "eval_loss": 0.027428090572357178, | |
| "eval_runtime": 22.1193, | |
| "eval_samples_per_second": 4.069, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05272564789991063, | |
| "step": 372000 | |
| }, | |
| { | |
| "epoch": 38.28782590843769, | |
| "grad_norm": 0.16268426179885864, | |
| "learning_rate": 0.00011736268257560173, | |
| "loss": 0.0452, | |
| "step": 373000 | |
| }, | |
| { | |
| "epoch": 38.28782590843769, | |
| "eval_cer": 0.013181279620853081, | |
| "eval_loss": 0.0269022174179554, | |
| "eval_runtime": 22.3714, | |
| "eval_samples_per_second": 4.023, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.049151027703306524, | |
| "step": 373000 | |
| }, | |
| { | |
| "epoch": 38.39047423526996, | |
| "grad_norm": 0.18285077810287476, | |
| "learning_rate": 0.00011633408763628884, | |
| "loss": 0.0456, | |
| "step": 374000 | |
| }, | |
| { | |
| "epoch": 38.39047423526996, | |
| "eval_cer": 0.01481042654028436, | |
| "eval_loss": 0.026672353968024254, | |
| "eval_runtime": 22.4441, | |
| "eval_samples_per_second": 4.01, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 374000 | |
| }, | |
| { | |
| "epoch": 38.493122562102236, | |
| "grad_norm": 0.14922891557216644, | |
| "learning_rate": 0.00011530549269697593, | |
| "loss": 0.0449, | |
| "step": 375000 | |
| }, | |
| { | |
| "epoch": 38.493122562102236, | |
| "eval_cer": 0.01525473933649289, | |
| "eval_loss": 0.026156587526202202, | |
| "eval_runtime": 22.2935, | |
| "eval_samples_per_second": 4.037, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 375000 | |
| }, | |
| { | |
| "epoch": 38.59577088893451, | |
| "grad_norm": 0.2601664662361145, | |
| "learning_rate": 0.00011427689775766304, | |
| "loss": 0.0453, | |
| "step": 376000 | |
| }, | |
| { | |
| "epoch": 38.59577088893451, | |
| "eval_cer": 0.015106635071090047, | |
| "eval_loss": 0.026526469737291336, | |
| "eval_runtime": 22.4581, | |
| "eval_samples_per_second": 4.007, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 376000 | |
| }, | |
| { | |
| "epoch": 38.69841921576678, | |
| "grad_norm": 0.23301970958709717, | |
| "learning_rate": 0.00011324830281835013, | |
| "loss": 0.0454, | |
| "step": 377000 | |
| }, | |
| { | |
| "epoch": 38.69841921576678, | |
| "eval_cer": 0.014514218009478674, | |
| "eval_loss": 0.026089007034897804, | |
| "eval_runtime": 22.0613, | |
| "eval_samples_per_second": 4.08, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 377000 | |
| }, | |
| { | |
| "epoch": 38.80106754259906, | |
| "grad_norm": 0.1848640739917755, | |
| "learning_rate": 0.00011221970787903724, | |
| "loss": 0.0455, | |
| "step": 378000 | |
| }, | |
| { | |
| "epoch": 38.80106754259906, | |
| "eval_cer": 0.015402843601895734, | |
| "eval_loss": 0.02705644629895687, | |
| "eval_runtime": 22.0628, | |
| "eval_samples_per_second": 4.079, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05183199285075961, | |
| "step": 378000 | |
| }, | |
| { | |
| "epoch": 38.90371586943133, | |
| "grad_norm": 0.2265135794878006, | |
| "learning_rate": 0.00011119111293972433, | |
| "loss": 0.0449, | |
| "step": 379000 | |
| }, | |
| { | |
| "epoch": 38.90371586943133, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.026829397305846214, | |
| "eval_runtime": 22.1974, | |
| "eval_samples_per_second": 4.055, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05272564789991063, | |
| "step": 379000 | |
| }, | |
| { | |
| "epoch": 39.006364196263604, | |
| "grad_norm": 0.18336538970470428, | |
| "learning_rate": 0.00011016251800041144, | |
| "loss": 0.045, | |
| "step": 380000 | |
| }, | |
| { | |
| "epoch": 39.006364196263604, | |
| "eval_cer": 0.014514218009478674, | |
| "eval_loss": 0.02715076506137848, | |
| "eval_runtime": 22.2325, | |
| "eval_samples_per_second": 4.048, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 380000 | |
| }, | |
| { | |
| "epoch": 39.10901252309587, | |
| "grad_norm": 0.2116203010082245, | |
| "learning_rate": 0.00010913392306109854, | |
| "loss": 0.0442, | |
| "step": 381000 | |
| }, | |
| { | |
| "epoch": 39.10901252309587, | |
| "eval_cer": 0.01569905213270142, | |
| "eval_loss": 0.026725102216005325, | |
| "eval_runtime": 22.0106, | |
| "eval_samples_per_second": 4.089, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 381000 | |
| }, | |
| { | |
| "epoch": 39.211660849928144, | |
| "grad_norm": 0.1735929548740387, | |
| "learning_rate": 0.00010810532812178564, | |
| "loss": 0.0446, | |
| "step": 382000 | |
| }, | |
| { | |
| "epoch": 39.211660849928144, | |
| "eval_cer": 0.017031990521327013, | |
| "eval_loss": 0.026632068678736687, | |
| "eval_runtime": 22.2735, | |
| "eval_samples_per_second": 4.041, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05540661304736372, | |
| "step": 382000 | |
| }, | |
| { | |
| "epoch": 39.31430917676042, | |
| "grad_norm": 0.29244282841682434, | |
| "learning_rate": 0.00010707673318247274, | |
| "loss": 0.0443, | |
| "step": 383000 | |
| }, | |
| { | |
| "epoch": 39.31430917676042, | |
| "eval_cer": 0.01525473933649289, | |
| "eval_loss": 0.02649509161710739, | |
| "eval_runtime": 22.223, | |
| "eval_samples_per_second": 4.05, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05183199285075961, | |
| "step": 383000 | |
| }, | |
| { | |
| "epoch": 39.41695750359269, | |
| "grad_norm": 0.21645478904247284, | |
| "learning_rate": 0.00010604813824315984, | |
| "loss": 0.0448, | |
| "step": 384000 | |
| }, | |
| { | |
| "epoch": 39.41695750359269, | |
| "eval_cer": 0.015995260663507108, | |
| "eval_loss": 0.02655700594186783, | |
| "eval_runtime": 22.1333, | |
| "eval_samples_per_second": 4.066, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05540661304736372, | |
| "step": 384000 | |
| }, | |
| { | |
| "epoch": 39.519605830424965, | |
| "grad_norm": 0.24354924261569977, | |
| "learning_rate": 0.00010501954330384694, | |
| "loss": 0.0447, | |
| "step": 385000 | |
| }, | |
| { | |
| "epoch": 39.519605830424965, | |
| "eval_cer": 0.015847156398104266, | |
| "eval_loss": 0.026025088503956795, | |
| "eval_runtime": 22.2171, | |
| "eval_samples_per_second": 4.051, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 385000 | |
| }, | |
| { | |
| "epoch": 39.62225415725724, | |
| "grad_norm": 0.15328273177146912, | |
| "learning_rate": 0.00010399094836453404, | |
| "loss": 0.0452, | |
| "step": 386000 | |
| }, | |
| { | |
| "epoch": 39.62225415725724, | |
| "eval_cer": 0.015847156398104266, | |
| "eval_loss": 0.025482947006821632, | |
| "eval_runtime": 22.319, | |
| "eval_samples_per_second": 4.032, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 386000 | |
| }, | |
| { | |
| "epoch": 39.72490248408951, | |
| "grad_norm": 0.25354963541030884, | |
| "learning_rate": 0.00010296235342522114, | |
| "loss": 0.0447, | |
| "step": 387000 | |
| }, | |
| { | |
| "epoch": 39.72490248408951, | |
| "eval_cer": 0.014958530805687204, | |
| "eval_loss": 0.026158807799220085, | |
| "eval_runtime": 22.2432, | |
| "eval_samples_per_second": 4.046, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 387000 | |
| }, | |
| { | |
| "epoch": 39.827550810921785, | |
| "grad_norm": 0.18456153571605682, | |
| "learning_rate": 0.00010193375848590825, | |
| "loss": 0.0452, | |
| "step": 388000 | |
| }, | |
| { | |
| "epoch": 39.827550810921785, | |
| "eval_cer": 0.01569905213270142, | |
| "eval_loss": 0.026585763320326805, | |
| "eval_runtime": 22.3419, | |
| "eval_samples_per_second": 4.028, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 388000 | |
| }, | |
| { | |
| "epoch": 39.93019913775405, | |
| "grad_norm": 1.1289212703704834, | |
| "learning_rate": 0.00010090516354659534, | |
| "loss": 0.0451, | |
| "step": 389000 | |
| }, | |
| { | |
| "epoch": 39.93019913775405, | |
| "eval_cer": 0.016143364928909953, | |
| "eval_loss": 0.026965312659740448, | |
| "eval_runtime": 22.2111, | |
| "eval_samples_per_second": 4.052, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05272564789991063, | |
| "step": 389000 | |
| }, | |
| { | |
| "epoch": 40.032847464586325, | |
| "grad_norm": 0.2642553448677063, | |
| "learning_rate": 9.987656860728245e-05, | |
| "loss": 0.0441, | |
| "step": 390000 | |
| }, | |
| { | |
| "epoch": 40.032847464586325, | |
| "eval_cer": 0.014069905213270142, | |
| "eval_loss": 0.02652685157954693, | |
| "eval_runtime": 21.9945, | |
| "eval_samples_per_second": 4.092, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 390000 | |
| }, | |
| { | |
| "epoch": 40.1354957914186, | |
| "grad_norm": 0.2834232449531555, | |
| "learning_rate": 9.884797366796956e-05, | |
| "loss": 0.0441, | |
| "step": 391000 | |
| }, | |
| { | |
| "epoch": 40.1354957914186, | |
| "eval_cer": 0.013773696682464455, | |
| "eval_loss": 0.026538578793406487, | |
| "eval_runtime": 21.6963, | |
| "eval_samples_per_second": 4.148, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.049151027703306524, | |
| "step": 391000 | |
| }, | |
| { | |
| "epoch": 40.23814411825087, | |
| "grad_norm": 0.366251140832901, | |
| "learning_rate": 9.781937872865666e-05, | |
| "loss": 0.044, | |
| "step": 392000 | |
| }, | |
| { | |
| "epoch": 40.23814411825087, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.025871722027659416, | |
| "eval_runtime": 21.9657, | |
| "eval_samples_per_second": 4.097, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 392000 | |
| }, | |
| { | |
| "epoch": 40.340792445083146, | |
| "grad_norm": 0.1964322179555893, | |
| "learning_rate": 9.679078378934376e-05, | |
| "loss": 0.0442, | |
| "step": 393000 | |
| }, | |
| { | |
| "epoch": 40.340792445083146, | |
| "eval_cer": 0.013625592417061612, | |
| "eval_loss": 0.025851983577013016, | |
| "eval_runtime": 21.9412, | |
| "eval_samples_per_second": 4.102, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 393000 | |
| }, | |
| { | |
| "epoch": 40.44344077191542, | |
| "grad_norm": 0.19335012137889862, | |
| "learning_rate": 9.576218885003086e-05, | |
| "loss": 0.0444, | |
| "step": 394000 | |
| }, | |
| { | |
| "epoch": 40.44344077191542, | |
| "eval_cer": 0.014218009478672985, | |
| "eval_loss": 0.025898663327097893, | |
| "eval_runtime": 22.0686, | |
| "eval_samples_per_second": 4.078, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 394000 | |
| }, | |
| { | |
| "epoch": 40.54608909874769, | |
| "grad_norm": 0.2371419221162796, | |
| "learning_rate": 9.473359391071797e-05, | |
| "loss": 0.0448, | |
| "step": 395000 | |
| }, | |
| { | |
| "epoch": 40.54608909874769, | |
| "eval_cer": 0.015402843601895734, | |
| "eval_loss": 0.02598469704389572, | |
| "eval_runtime": 22.2182, | |
| "eval_samples_per_second": 4.051, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 395000 | |
| }, | |
| { | |
| "epoch": 40.64873742557997, | |
| "grad_norm": 0.190704807639122, | |
| "learning_rate": 9.370499897140506e-05, | |
| "loss": 0.0446, | |
| "step": 396000 | |
| }, | |
| { | |
| "epoch": 40.64873742557997, | |
| "eval_cer": 0.013181279620853081, | |
| "eval_loss": 0.025896675884723663, | |
| "eval_runtime": 22.4373, | |
| "eval_samples_per_second": 4.011, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 396000 | |
| }, | |
| { | |
| "epoch": 40.75138575241223, | |
| "grad_norm": 0.24071630835533142, | |
| "learning_rate": 9.267640403209217e-05, | |
| "loss": 0.0445, | |
| "step": 397000 | |
| }, | |
| { | |
| "epoch": 40.75138575241223, | |
| "eval_cer": 0.015106635071090047, | |
| "eval_loss": 0.025766143575310707, | |
| "eval_runtime": 22.1683, | |
| "eval_samples_per_second": 4.06, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 397000 | |
| }, | |
| { | |
| "epoch": 40.85403407924451, | |
| "grad_norm": 0.37343931198120117, | |
| "learning_rate": 9.164780909277926e-05, | |
| "loss": 0.044, | |
| "step": 398000 | |
| }, | |
| { | |
| "epoch": 40.85403407924451, | |
| "eval_cer": 0.015106635071090047, | |
| "eval_loss": 0.025763213634490967, | |
| "eval_runtime": 22.2275, | |
| "eval_samples_per_second": 4.049, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 398000 | |
| }, | |
| { | |
| "epoch": 40.95668240607678, | |
| "grad_norm": 0.2885558605194092, | |
| "learning_rate": 9.061921415346637e-05, | |
| "loss": 0.0446, | |
| "step": 399000 | |
| }, | |
| { | |
| "epoch": 40.95668240607678, | |
| "eval_cer": 0.01569905213270142, | |
| "eval_loss": 0.02574954554438591, | |
| "eval_runtime": 22.2623, | |
| "eval_samples_per_second": 4.043, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05183199285075961, | |
| "step": 399000 | |
| }, | |
| { | |
| "epoch": 41.05933073290905, | |
| "grad_norm": 0.2904144525527954, | |
| "learning_rate": 8.959061921415348e-05, | |
| "loss": 0.0437, | |
| "step": 400000 | |
| }, | |
| { | |
| "epoch": 41.05933073290905, | |
| "eval_cer": 0.016291469194312798, | |
| "eval_loss": 0.026339180767536163, | |
| "eval_runtime": 22.3982, | |
| "eval_samples_per_second": 4.018, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 400000 | |
| }, | |
| { | |
| "epoch": 41.16197905974133, | |
| "grad_norm": 0.47024524211883545, | |
| "learning_rate": 8.856202427484057e-05, | |
| "loss": 0.0441, | |
| "step": 401000 | |
| }, | |
| { | |
| "epoch": 41.16197905974133, | |
| "eval_cer": 0.015402843601895734, | |
| "eval_loss": 0.026481064036488533, | |
| "eval_runtime": 22.1309, | |
| "eval_samples_per_second": 4.067, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.049151027703306524, | |
| "step": 401000 | |
| }, | |
| { | |
| "epoch": 41.2646273865736, | |
| "grad_norm": 0.20567986369132996, | |
| "learning_rate": 8.753342933552768e-05, | |
| "loss": 0.0438, | |
| "step": 402000 | |
| }, | |
| { | |
| "epoch": 41.2646273865736, | |
| "eval_cer": 0.01569905213270142, | |
| "eval_loss": 0.026423340663313866, | |
| "eval_runtime": 22.1647, | |
| "eval_samples_per_second": 4.061, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 402000 | |
| }, | |
| { | |
| "epoch": 41.367275713405874, | |
| "grad_norm": 0.25154054164886475, | |
| "learning_rate": 8.650483439621477e-05, | |
| "loss": 0.0439, | |
| "step": 403000 | |
| }, | |
| { | |
| "epoch": 41.367275713405874, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.025870798155665398, | |
| "eval_runtime": 22.2814, | |
| "eval_samples_per_second": 4.039, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.045576407506702415, | |
| "step": 403000 | |
| }, | |
| { | |
| "epoch": 41.46992404023814, | |
| "grad_norm": 0.22818030416965485, | |
| "learning_rate": 8.547623945690188e-05, | |
| "loss": 0.0437, | |
| "step": 404000 | |
| }, | |
| { | |
| "epoch": 41.46992404023814, | |
| "eval_cer": 0.015847156398104266, | |
| "eval_loss": 0.025692187249660492, | |
| "eval_runtime": 22.2761, | |
| "eval_samples_per_second": 4.04, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05272564789991063, | |
| "step": 404000 | |
| }, | |
| { | |
| "epoch": 41.572572367070414, | |
| "grad_norm": 0.20878200232982635, | |
| "learning_rate": 8.444764451758897e-05, | |
| "loss": 0.044, | |
| "step": 405000 | |
| }, | |
| { | |
| "epoch": 41.572572367070414, | |
| "eval_cer": 0.014514218009478674, | |
| "eval_loss": 0.025789031758904457, | |
| "eval_runtime": 22.1041, | |
| "eval_samples_per_second": 4.072, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 405000 | |
| }, | |
| { | |
| "epoch": 41.67522069390269, | |
| "grad_norm": 0.37076711654663086, | |
| "learning_rate": 8.341904957827608e-05, | |
| "loss": 0.0438, | |
| "step": 406000 | |
| }, | |
| { | |
| "epoch": 41.67522069390269, | |
| "eval_cer": 0.01569905213270142, | |
| "eval_loss": 0.025606298819184303, | |
| "eval_runtime": 22.1493, | |
| "eval_samples_per_second": 4.063, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 406000 | |
| }, | |
| { | |
| "epoch": 41.77786902073496, | |
| "grad_norm": 0.3799729347229004, | |
| "learning_rate": 8.239045463896318e-05, | |
| "loss": 0.0439, | |
| "step": 407000 | |
| }, | |
| { | |
| "epoch": 41.77786902073496, | |
| "eval_cer": 0.015106635071090047, | |
| "eval_loss": 0.02529660426080227, | |
| "eval_runtime": 22.1548, | |
| "eval_samples_per_second": 4.062, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 407000 | |
| }, | |
| { | |
| "epoch": 41.880517347567235, | |
| "grad_norm": 0.1976720541715622, | |
| "learning_rate": 8.136185969965028e-05, | |
| "loss": 0.0442, | |
| "step": 408000 | |
| }, | |
| { | |
| "epoch": 41.880517347567235, | |
| "eval_cer": 0.013181279620853081, | |
| "eval_loss": 0.025364946573972702, | |
| "eval_runtime": 22.2419, | |
| "eval_samples_per_second": 4.046, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.045576407506702415, | |
| "step": 408000 | |
| }, | |
| { | |
| "epoch": 41.98316567439951, | |
| "grad_norm": 0.35871848464012146, | |
| "learning_rate": 8.033326476033738e-05, | |
| "loss": 0.0438, | |
| "step": 409000 | |
| }, | |
| { | |
| "epoch": 41.98316567439951, | |
| "eval_cer": 0.016143364928909953, | |
| "eval_loss": 0.02546422928571701, | |
| "eval_runtime": 22.0995, | |
| "eval_samples_per_second": 4.072, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05361930294906166, | |
| "step": 409000 | |
| }, | |
| { | |
| "epoch": 42.08581400123178, | |
| "grad_norm": 0.3228365480899811, | |
| "learning_rate": 7.930466982102448e-05, | |
| "loss": 0.0435, | |
| "step": 410000 | |
| }, | |
| { | |
| "epoch": 42.08581400123178, | |
| "eval_cer": 0.013921800947867298, | |
| "eval_loss": 0.025534870103001595, | |
| "eval_runtime": 22.0953, | |
| "eval_samples_per_second": 4.073, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 410000 | |
| }, | |
| { | |
| "epoch": 42.188462328064055, | |
| "grad_norm": 0.32005515694618225, | |
| "learning_rate": 7.827607488171158e-05, | |
| "loss": 0.0432, | |
| "step": 411000 | |
| }, | |
| { | |
| "epoch": 42.188462328064055, | |
| "eval_cer": 0.01229265402843602, | |
| "eval_loss": 0.024984827265143394, | |
| "eval_runtime": 22.11, | |
| "eval_samples_per_second": 4.071, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.044682752457551385, | |
| "step": 411000 | |
| }, | |
| { | |
| "epoch": 42.29111065489632, | |
| "grad_norm": 0.2567862570285797, | |
| "learning_rate": 7.724747994239868e-05, | |
| "loss": 0.0435, | |
| "step": 412000 | |
| }, | |
| { | |
| "epoch": 42.29111065489632, | |
| "eval_cer": 0.01525473933649289, | |
| "eval_loss": 0.025618551298975945, | |
| "eval_runtime": 22.2013, | |
| "eval_samples_per_second": 4.054, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05183199285075961, | |
| "step": 412000 | |
| }, | |
| { | |
| "epoch": 42.393758981728595, | |
| "grad_norm": 0.39079025387763977, | |
| "learning_rate": 7.621888500308578e-05, | |
| "loss": 0.0434, | |
| "step": 413000 | |
| }, | |
| { | |
| "epoch": 42.393758981728595, | |
| "eval_cer": 0.013773696682464455, | |
| "eval_loss": 0.025224734097719193, | |
| "eval_runtime": 21.8997, | |
| "eval_samples_per_second": 4.11, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.045576407506702415, | |
| "step": 413000 | |
| }, | |
| { | |
| "epoch": 42.49640730856087, | |
| "grad_norm": 0.20112274587154388, | |
| "learning_rate": 7.519029006377289e-05, | |
| "loss": 0.0434, | |
| "step": 414000 | |
| }, | |
| { | |
| "epoch": 42.49640730856087, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.02557324431836605, | |
| "eval_runtime": 22.1285, | |
| "eval_samples_per_second": 4.067, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.049151027703306524, | |
| "step": 414000 | |
| }, | |
| { | |
| "epoch": 42.59905563539314, | |
| "grad_norm": 0.29187527298927307, | |
| "learning_rate": 7.416169512445998e-05, | |
| "loss": 0.0433, | |
| "step": 415000 | |
| }, | |
| { | |
| "epoch": 42.59905563539314, | |
| "eval_cer": 0.014366113744075829, | |
| "eval_loss": 0.02564132958650589, | |
| "eval_runtime": 22.0916, | |
| "eval_samples_per_second": 4.074, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 415000 | |
| }, | |
| { | |
| "epoch": 42.701703962225416, | |
| "grad_norm": 0.20293624699115753, | |
| "learning_rate": 7.313310018514709e-05, | |
| "loss": 0.0437, | |
| "step": 416000 | |
| }, | |
| { | |
| "epoch": 42.701703962225416, | |
| "eval_cer": 0.01481042654028436, | |
| "eval_loss": 0.025618135929107666, | |
| "eval_runtime": 22.0252, | |
| "eval_samples_per_second": 4.086, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.049151027703306524, | |
| "step": 416000 | |
| }, | |
| { | |
| "epoch": 42.80435228905769, | |
| "grad_norm": 0.2777237296104431, | |
| "learning_rate": 7.210450524583418e-05, | |
| "loss": 0.0438, | |
| "step": 417000 | |
| }, | |
| { | |
| "epoch": 42.80435228905769, | |
| "eval_cer": 0.01569905213270142, | |
| "eval_loss": 0.025969378650188446, | |
| "eval_runtime": 22.1694, | |
| "eval_samples_per_second": 4.06, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05272564789991063, | |
| "step": 417000 | |
| }, | |
| { | |
| "epoch": 42.90700061588996, | |
| "grad_norm": 0.29848456382751465, | |
| "learning_rate": 7.107591030652129e-05, | |
| "loss": 0.0436, | |
| "step": 418000 | |
| }, | |
| { | |
| "epoch": 42.90700061588996, | |
| "eval_cer": 0.015106635071090047, | |
| "eval_loss": 0.02571621723473072, | |
| "eval_runtime": 22.2799, | |
| "eval_samples_per_second": 4.04, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.049151027703306524, | |
| "step": 418000 | |
| }, | |
| { | |
| "epoch": 43.00964894272224, | |
| "grad_norm": 0.21752919256687164, | |
| "learning_rate": 7.004731536720838e-05, | |
| "loss": 0.0435, | |
| "step": 419000 | |
| }, | |
| { | |
| "epoch": 43.00964894272224, | |
| "eval_cer": 0.014958530805687204, | |
| "eval_loss": 0.025668691843748093, | |
| "eval_runtime": 22.1447, | |
| "eval_samples_per_second": 4.064, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.05093833780160858, | |
| "step": 419000 | |
| }, | |
| { | |
| "epoch": 43.1122972695545, | |
| "grad_norm": 0.3110567033290863, | |
| "learning_rate": 6.901872042789549e-05, | |
| "loss": 0.0431, | |
| "step": 420000 | |
| }, | |
| { | |
| "epoch": 43.1122972695545, | |
| "eval_cer": 0.014069905213270142, | |
| "eval_loss": 0.025269243866205215, | |
| "eval_runtime": 22.5329, | |
| "eval_samples_per_second": 3.994, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.045576407506702415, | |
| "step": 420000 | |
| }, | |
| { | |
| "epoch": 43.21494559638678, | |
| "grad_norm": 0.26389312744140625, | |
| "learning_rate": 6.79901254885826e-05, | |
| "loss": 0.0429, | |
| "step": 421000 | |
| }, | |
| { | |
| "epoch": 43.21494559638678, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.025585921481251717, | |
| "eval_runtime": 22.2351, | |
| "eval_samples_per_second": 4.048, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 421000 | |
| }, | |
| { | |
| "epoch": 43.31759392321905, | |
| "grad_norm": 0.2228713184595108, | |
| "learning_rate": 6.69615305492697e-05, | |
| "loss": 0.043, | |
| "step": 422000 | |
| }, | |
| { | |
| "epoch": 43.31759392321905, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.02543068863451481, | |
| "eval_runtime": 22.1864, | |
| "eval_samples_per_second": 4.057, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.049151027703306524, | |
| "step": 422000 | |
| }, | |
| { | |
| "epoch": 43.420242250051324, | |
| "grad_norm": 0.1402529925107956, | |
| "learning_rate": 6.593293560995681e-05, | |
| "loss": 0.0439, | |
| "step": 423000 | |
| }, | |
| { | |
| "epoch": 43.420242250051324, | |
| "eval_cer": 0.014514218009478674, | |
| "eval_loss": 0.025465745478868484, | |
| "eval_runtime": 22.3826, | |
| "eval_samples_per_second": 4.021, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 423000 | |
| }, | |
| { | |
| "epoch": 43.5228905768836, | |
| "grad_norm": 0.3183715045452118, | |
| "learning_rate": 6.49043406706439e-05, | |
| "loss": 0.0431, | |
| "step": 424000 | |
| }, | |
| { | |
| "epoch": 43.5228905768836, | |
| "eval_cer": 0.014514218009478674, | |
| "eval_loss": 0.025318369269371033, | |
| "eval_runtime": 22.2919, | |
| "eval_samples_per_second": 4.037, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 424000 | |
| }, | |
| { | |
| "epoch": 43.62553890371587, | |
| "grad_norm": 0.29827529191970825, | |
| "learning_rate": 6.387574573133101e-05, | |
| "loss": 0.0434, | |
| "step": 425000 | |
| }, | |
| { | |
| "epoch": 43.62553890371587, | |
| "eval_cer": 0.01481042654028436, | |
| "eval_loss": 0.025606686249375343, | |
| "eval_runtime": 22.0625, | |
| "eval_samples_per_second": 4.079, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 425000 | |
| }, | |
| { | |
| "epoch": 43.728187230548144, | |
| "grad_norm": 0.16081830859184265, | |
| "learning_rate": 6.28471507920181e-05, | |
| "loss": 0.0431, | |
| "step": 426000 | |
| }, | |
| { | |
| "epoch": 43.728187230548144, | |
| "eval_cer": 0.014958530805687204, | |
| "eval_loss": 0.025782672688364983, | |
| "eval_runtime": 22.1883, | |
| "eval_samples_per_second": 4.056, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 426000 | |
| }, | |
| { | |
| "epoch": 43.83083555738042, | |
| "grad_norm": 0.19551779329776764, | |
| "learning_rate": 6.181855585270521e-05, | |
| "loss": 0.0437, | |
| "step": 427000 | |
| }, | |
| { | |
| "epoch": 43.83083555738042, | |
| "eval_cer": 0.013773696682464455, | |
| "eval_loss": 0.025874827057123184, | |
| "eval_runtime": 22.0445, | |
| "eval_samples_per_second": 4.083, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 427000 | |
| }, | |
| { | |
| "epoch": 43.933483884212684, | |
| "grad_norm": 0.3405396044254303, | |
| "learning_rate": 6.07899609133923e-05, | |
| "loss": 0.0431, | |
| "step": 428000 | |
| }, | |
| { | |
| "epoch": 43.933483884212684, | |
| "eval_cer": 0.014366113744075829, | |
| "eval_loss": 0.025233900174498558, | |
| "eval_runtime": 22.1498, | |
| "eval_samples_per_second": 4.063, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 428000 | |
| }, | |
| { | |
| "epoch": 44.03613221104496, | |
| "grad_norm": 0.1871403306722641, | |
| "learning_rate": 5.97613659740794e-05, | |
| "loss": 0.0424, | |
| "step": 429000 | |
| }, | |
| { | |
| "epoch": 44.03613221104496, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.025231176987290382, | |
| "eval_runtime": 22.1695, | |
| "eval_samples_per_second": 4.06, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 429000 | |
| }, | |
| { | |
| "epoch": 44.13878053787723, | |
| "grad_norm": 0.272748202085495, | |
| "learning_rate": 5.873277103476652e-05, | |
| "loss": 0.0425, | |
| "step": 430000 | |
| }, | |
| { | |
| "epoch": 44.13878053787723, | |
| "eval_cer": 0.012885071090047393, | |
| "eval_loss": 0.025549624115228653, | |
| "eval_runtime": 22.1325, | |
| "eval_samples_per_second": 4.066, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 430000 | |
| }, | |
| { | |
| "epoch": 44.241428864709505, | |
| "grad_norm": 0.15834620594978333, | |
| "learning_rate": 5.7704176095453617e-05, | |
| "loss": 0.0428, | |
| "step": 431000 | |
| }, | |
| { | |
| "epoch": 44.241428864709505, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.02536383457481861, | |
| "eval_runtime": 22.0624, | |
| "eval_samples_per_second": 4.079, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.049151027703306524, | |
| "step": 431000 | |
| }, | |
| { | |
| "epoch": 44.34407719154178, | |
| "grad_norm": 0.21297834813594818, | |
| "learning_rate": 5.6675581156140717e-05, | |
| "loss": 0.0426, | |
| "step": 432000 | |
| }, | |
| { | |
| "epoch": 44.34407719154178, | |
| "eval_cer": 0.013033175355450236, | |
| "eval_loss": 0.025508729740977287, | |
| "eval_runtime": 22.2302, | |
| "eval_samples_per_second": 4.049, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 432000 | |
| }, | |
| { | |
| "epoch": 44.44672551837405, | |
| "grad_norm": 0.3164765536785126, | |
| "learning_rate": 5.5646986216827816e-05, | |
| "loss": 0.0428, | |
| "step": 433000 | |
| }, | |
| { | |
| "epoch": 44.44672551837405, | |
| "eval_cer": 0.014069905213270142, | |
| "eval_loss": 0.025067314505577087, | |
| "eval_runtime": 22.074, | |
| "eval_samples_per_second": 4.077, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 433000 | |
| }, | |
| { | |
| "epoch": 44.549373845206325, | |
| "grad_norm": 0.14805859327316284, | |
| "learning_rate": 5.4618391277514916e-05, | |
| "loss": 0.0435, | |
| "step": 434000 | |
| }, | |
| { | |
| "epoch": 44.549373845206325, | |
| "eval_cer": 0.014218009478672985, | |
| "eval_loss": 0.02515277825295925, | |
| "eval_runtime": 22.2077, | |
| "eval_samples_per_second": 4.053, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 434000 | |
| }, | |
| { | |
| "epoch": 44.6520221720386, | |
| "grad_norm": 0.19713029265403748, | |
| "learning_rate": 5.3589796338202016e-05, | |
| "loss": 0.0429, | |
| "step": 435000 | |
| }, | |
| { | |
| "epoch": 44.6520221720386, | |
| "eval_cer": 0.014514218009478674, | |
| "eval_loss": 0.02521173469722271, | |
| "eval_runtime": 22.2596, | |
| "eval_samples_per_second": 4.043, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 435000 | |
| }, | |
| { | |
| "epoch": 44.754670498870865, | |
| "grad_norm": 0.2486603856086731, | |
| "learning_rate": 5.2561201398889116e-05, | |
| "loss": 0.0428, | |
| "step": 436000 | |
| }, | |
| { | |
| "epoch": 44.754670498870865, | |
| "eval_cer": 0.01481042654028436, | |
| "eval_loss": 0.025502758100628853, | |
| "eval_runtime": 22.2902, | |
| "eval_samples_per_second": 4.038, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 436000 | |
| }, | |
| { | |
| "epoch": 44.85731882570314, | |
| "grad_norm": 0.29049795866012573, | |
| "learning_rate": 5.153260645957622e-05, | |
| "loss": 0.0425, | |
| "step": 437000 | |
| }, | |
| { | |
| "epoch": 44.85731882570314, | |
| "eval_cer": 0.01525473933649289, | |
| "eval_loss": 0.02529684267938137, | |
| "eval_runtime": 22.181, | |
| "eval_samples_per_second": 4.058, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 437000 | |
| }, | |
| { | |
| "epoch": 44.95996715253541, | |
| "grad_norm": 0.2436273992061615, | |
| "learning_rate": 5.050401152026332e-05, | |
| "loss": 0.0428, | |
| "step": 438000 | |
| }, | |
| { | |
| "epoch": 44.95996715253541, | |
| "eval_cer": 0.014366113744075829, | |
| "eval_loss": 0.024828782305121422, | |
| "eval_runtime": 22.2395, | |
| "eval_samples_per_second": 4.047, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 438000 | |
| }, | |
| { | |
| "epoch": 45.062615479367686, | |
| "grad_norm": 0.24658174812793732, | |
| "learning_rate": 4.947541658095042e-05, | |
| "loss": 0.0424, | |
| "step": 439000 | |
| }, | |
| { | |
| "epoch": 45.062615479367686, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.025087928399443626, | |
| "eval_runtime": 22.0216, | |
| "eval_samples_per_second": 4.087, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 439000 | |
| }, | |
| { | |
| "epoch": 45.16526380619996, | |
| "grad_norm": 0.39545565843582153, | |
| "learning_rate": 4.844682164163752e-05, | |
| "loss": 0.0423, | |
| "step": 440000 | |
| }, | |
| { | |
| "epoch": 45.16526380619996, | |
| "eval_cer": 0.013773696682464455, | |
| "eval_loss": 0.0248849056661129, | |
| "eval_runtime": 22.0732, | |
| "eval_samples_per_second": 4.077, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.045576407506702415, | |
| "step": 440000 | |
| }, | |
| { | |
| "epoch": 45.26791213303223, | |
| "grad_norm": 0.3729030191898346, | |
| "learning_rate": 4.741822670232462e-05, | |
| "loss": 0.0421, | |
| "step": 441000 | |
| }, | |
| { | |
| "epoch": 45.26791213303223, | |
| "eval_cer": 0.013773696682464455, | |
| "eval_loss": 0.024852894246578217, | |
| "eval_runtime": 22.1136, | |
| "eval_samples_per_second": 4.07, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.045576407506702415, | |
| "step": 441000 | |
| }, | |
| { | |
| "epoch": 45.37056045986451, | |
| "grad_norm": 0.2696306109428406, | |
| "learning_rate": 4.638963176301172e-05, | |
| "loss": 0.0424, | |
| "step": 442000 | |
| }, | |
| { | |
| "epoch": 45.37056045986451, | |
| "eval_cer": 0.014218009478672985, | |
| "eval_loss": 0.024875616654753685, | |
| "eval_runtime": 22.2425, | |
| "eval_samples_per_second": 4.046, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 442000 | |
| }, | |
| { | |
| "epoch": 45.47320878669678, | |
| "grad_norm": 0.18253710865974426, | |
| "learning_rate": 4.536103682369882e-05, | |
| "loss": 0.0427, | |
| "step": 443000 | |
| }, | |
| { | |
| "epoch": 45.47320878669678, | |
| "eval_cer": 0.013921800947867298, | |
| "eval_loss": 0.02515345811843872, | |
| "eval_runtime": 22.4087, | |
| "eval_samples_per_second": 4.016, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.045576407506702415, | |
| "step": 443000 | |
| }, | |
| { | |
| "epoch": 45.57585711352905, | |
| "grad_norm": 0.5241480469703674, | |
| "learning_rate": 4.433244188438593e-05, | |
| "loss": 0.0424, | |
| "step": 444000 | |
| }, | |
| { | |
| "epoch": 45.57585711352905, | |
| "eval_cer": 0.014218009478672985, | |
| "eval_loss": 0.024990374222397804, | |
| "eval_runtime": 22.5101, | |
| "eval_samples_per_second": 3.998, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 444000 | |
| }, | |
| { | |
| "epoch": 45.67850544036132, | |
| "grad_norm": 0.26579299569129944, | |
| "learning_rate": 4.330384694507303e-05, | |
| "loss": 0.0422, | |
| "step": 445000 | |
| }, | |
| { | |
| "epoch": 45.67850544036132, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.025033339858055115, | |
| "eval_runtime": 22.2997, | |
| "eval_samples_per_second": 4.036, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 445000 | |
| }, | |
| { | |
| "epoch": 45.781153767193594, | |
| "grad_norm": 0.15677900612354279, | |
| "learning_rate": 4.2275252005760136e-05, | |
| "loss": 0.0424, | |
| "step": 446000 | |
| }, | |
| { | |
| "epoch": 45.781153767193594, | |
| "eval_cer": 0.015402843601895734, | |
| "eval_loss": 0.025280646979808807, | |
| "eval_runtime": 22.1999, | |
| "eval_samples_per_second": 4.054, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.050044682752457555, | |
| "step": 446000 | |
| }, | |
| { | |
| "epoch": 45.88380209402587, | |
| "grad_norm": 0.25194716453552246, | |
| "learning_rate": 4.1246657066447236e-05, | |
| "loss": 0.0428, | |
| "step": 447000 | |
| }, | |
| { | |
| "epoch": 45.88380209402587, | |
| "eval_cer": 0.013625592417061612, | |
| "eval_loss": 0.02510838583111763, | |
| "eval_runtime": 22.1534, | |
| "eval_samples_per_second": 4.063, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.043789097408400354, | |
| "step": 447000 | |
| }, | |
| { | |
| "epoch": 45.98645042085814, | |
| "grad_norm": 0.24914862215518951, | |
| "learning_rate": 4.0218062127134336e-05, | |
| "loss": 0.0428, | |
| "step": 448000 | |
| }, | |
| { | |
| "epoch": 45.98645042085814, | |
| "eval_cer": 0.015106635071090047, | |
| "eval_loss": 0.02476254291832447, | |
| "eval_runtime": 22.087, | |
| "eval_samples_per_second": 4.075, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.049151027703306524, | |
| "step": 448000 | |
| }, | |
| { | |
| "epoch": 46.089098747690414, | |
| "grad_norm": 0.318974107503891, | |
| "learning_rate": 3.9189467187821436e-05, | |
| "loss": 0.0421, | |
| "step": 449000 | |
| }, | |
| { | |
| "epoch": 46.089098747690414, | |
| "eval_cer": 0.014218009478672985, | |
| "eval_loss": 0.02481299825012684, | |
| "eval_runtime": 21.9633, | |
| "eval_samples_per_second": 4.098, | |
| "eval_steps_per_second": 0.046, | |
| "eval_wer": 0.043789097408400354, | |
| "step": 449000 | |
| }, | |
| { | |
| "epoch": 46.19174707452269, | |
| "grad_norm": 0.21947523951530457, | |
| "learning_rate": 3.816087224850854e-05, | |
| "loss": 0.0421, | |
| "step": 450000 | |
| }, | |
| { | |
| "epoch": 46.19174707452269, | |
| "eval_cer": 0.013625592417061612, | |
| "eval_loss": 0.025007640942931175, | |
| "eval_runtime": 22.0576, | |
| "eval_samples_per_second": 4.08, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.044682752457551385, | |
| "step": 450000 | |
| }, | |
| { | |
| "epoch": 46.29439540135496, | |
| "grad_norm": 0.2869652807712555, | |
| "learning_rate": 3.713227730919564e-05, | |
| "loss": 0.0417, | |
| "step": 451000 | |
| }, | |
| { | |
| "epoch": 46.29439540135496, | |
| "eval_cer": 0.014069905213270142, | |
| "eval_loss": 0.025139357894659042, | |
| "eval_runtime": 22.1428, | |
| "eval_samples_per_second": 4.065, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 451000 | |
| }, | |
| { | |
| "epoch": 46.39704372818723, | |
| "grad_norm": 0.24226853251457214, | |
| "learning_rate": 3.610368236988274e-05, | |
| "loss": 0.0423, | |
| "step": 452000 | |
| }, | |
| { | |
| "epoch": 46.39704372818723, | |
| "eval_cer": 0.014218009478672985, | |
| "eval_loss": 0.02465611696243286, | |
| "eval_runtime": 22.2059, | |
| "eval_samples_per_second": 4.053, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 452000 | |
| }, | |
| { | |
| "epoch": 46.4996920550195, | |
| "grad_norm": 0.508613646030426, | |
| "learning_rate": 3.507508743056984e-05, | |
| "loss": 0.0422, | |
| "step": 453000 | |
| }, | |
| { | |
| "epoch": 46.4996920550195, | |
| "eval_cer": 0.014958530805687204, | |
| "eval_loss": 0.024790233001112938, | |
| "eval_runtime": 22.0074, | |
| "eval_samples_per_second": 4.09, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 453000 | |
| }, | |
| { | |
| "epoch": 46.602340381851775, | |
| "grad_norm": 0.22070400416851044, | |
| "learning_rate": 3.404649249125694e-05, | |
| "loss": 0.0415, | |
| "step": 454000 | |
| }, | |
| { | |
| "epoch": 46.602340381851775, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.024966726079583168, | |
| "eval_runtime": 22.0647, | |
| "eval_samples_per_second": 4.079, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 454000 | |
| }, | |
| { | |
| "epoch": 46.70498870868405, | |
| "grad_norm": 0.19661836326122284, | |
| "learning_rate": 3.301789755194404e-05, | |
| "loss": 0.0424, | |
| "step": 455000 | |
| }, | |
| { | |
| "epoch": 46.70498870868405, | |
| "eval_cer": 0.013625592417061612, | |
| "eval_loss": 0.024841254577040672, | |
| "eval_runtime": 22.1576, | |
| "eval_samples_per_second": 4.062, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.043789097408400354, | |
| "step": 455000 | |
| }, | |
| { | |
| "epoch": 46.80763703551632, | |
| "grad_norm": 0.475782185792923, | |
| "learning_rate": 3.198930261263114e-05, | |
| "loss": 0.0424, | |
| "step": 456000 | |
| }, | |
| { | |
| "epoch": 46.80763703551632, | |
| "eval_cer": 0.014514218009478674, | |
| "eval_loss": 0.02477007918059826, | |
| "eval_runtime": 22.1397, | |
| "eval_samples_per_second": 4.065, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 456000 | |
| }, | |
| { | |
| "epoch": 46.910285362348596, | |
| "grad_norm": 0.36696240305900574, | |
| "learning_rate": 3.096070767331825e-05, | |
| "loss": 0.0426, | |
| "step": 457000 | |
| }, | |
| { | |
| "epoch": 46.910285362348596, | |
| "eval_cer": 0.013625592417061612, | |
| "eval_loss": 0.024692127481102943, | |
| "eval_runtime": 22.0727, | |
| "eval_samples_per_second": 4.077, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.043789097408400354, | |
| "step": 457000 | |
| }, | |
| { | |
| "epoch": 47.01293368918087, | |
| "grad_norm": 0.1705227941274643, | |
| "learning_rate": 2.993211273400535e-05, | |
| "loss": 0.0422, | |
| "step": 458000 | |
| }, | |
| { | |
| "epoch": 47.01293368918087, | |
| "eval_cer": 0.014514218009478674, | |
| "eval_loss": 0.024774568155407906, | |
| "eval_runtime": 22.1972, | |
| "eval_samples_per_second": 4.055, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 458000 | |
| }, | |
| { | |
| "epoch": 47.115582016013136, | |
| "grad_norm": 0.21439406275749207, | |
| "learning_rate": 2.8903517794692452e-05, | |
| "loss": 0.0418, | |
| "step": 459000 | |
| }, | |
| { | |
| "epoch": 47.115582016013136, | |
| "eval_cer": 0.015550947867298577, | |
| "eval_loss": 0.024719279259443283, | |
| "eval_runtime": 22.0766, | |
| "eval_samples_per_second": 4.077, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 459000 | |
| }, | |
| { | |
| "epoch": 47.21823034284541, | |
| "grad_norm": 0.34435534477233887, | |
| "learning_rate": 2.7874922855379552e-05, | |
| "loss": 0.0415, | |
| "step": 460000 | |
| }, | |
| { | |
| "epoch": 47.21823034284541, | |
| "eval_cer": 0.015550947867298577, | |
| "eval_loss": 0.024668768048286438, | |
| "eval_runtime": 22.0611, | |
| "eval_samples_per_second": 4.08, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.049151027703306524, | |
| "step": 460000 | |
| }, | |
| { | |
| "epoch": 47.32087866967768, | |
| "grad_norm": 0.25130486488342285, | |
| "learning_rate": 2.6846327916066652e-05, | |
| "loss": 0.0417, | |
| "step": 461000 | |
| }, | |
| { | |
| "epoch": 47.32087866967768, | |
| "eval_cer": 0.015402843601895734, | |
| "eval_loss": 0.024554278701543808, | |
| "eval_runtime": 22.0634, | |
| "eval_samples_per_second": 4.079, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.049151027703306524, | |
| "step": 461000 | |
| }, | |
| { | |
| "epoch": 47.423526996509956, | |
| "grad_norm": 0.29562491178512573, | |
| "learning_rate": 2.5817732976753755e-05, | |
| "loss": 0.0414, | |
| "step": 462000 | |
| }, | |
| { | |
| "epoch": 47.423526996509956, | |
| "eval_cer": 0.014366113744075829, | |
| "eval_loss": 0.02472042851150036, | |
| "eval_runtime": 22.1921, | |
| "eval_samples_per_second": 4.055, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.043789097408400354, | |
| "step": 462000 | |
| }, | |
| { | |
| "epoch": 47.52617532334223, | |
| "grad_norm": 0.4088131785392761, | |
| "learning_rate": 2.4789138037440855e-05, | |
| "loss": 0.0421, | |
| "step": 463000 | |
| }, | |
| { | |
| "epoch": 47.52617532334223, | |
| "eval_cer": 0.014218009478672985, | |
| "eval_loss": 0.024495158344507217, | |
| "eval_runtime": 22.2123, | |
| "eval_samples_per_second": 4.052, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.044682752457551385, | |
| "step": 463000 | |
| }, | |
| { | |
| "epoch": 47.6288236501745, | |
| "grad_norm": 0.27792465686798096, | |
| "learning_rate": 2.3760543098127955e-05, | |
| "loss": 0.0419, | |
| "step": 464000 | |
| }, | |
| { | |
| "epoch": 47.6288236501745, | |
| "eval_cer": 0.01481042654028436, | |
| "eval_loss": 0.02450607530772686, | |
| "eval_runtime": 22.2255, | |
| "eval_samples_per_second": 4.049, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 464000 | |
| }, | |
| { | |
| "epoch": 47.73147197700678, | |
| "grad_norm": 0.19159696996212006, | |
| "learning_rate": 2.2731948158815062e-05, | |
| "loss": 0.042, | |
| "step": 465000 | |
| }, | |
| { | |
| "epoch": 47.73147197700678, | |
| "eval_cer": 0.014218009478672985, | |
| "eval_loss": 0.02452634647488594, | |
| "eval_runtime": 22.143, | |
| "eval_samples_per_second": 4.064, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.044682752457551385, | |
| "step": 465000 | |
| }, | |
| { | |
| "epoch": 47.83412030383905, | |
| "grad_norm": 0.2656868100166321, | |
| "learning_rate": 2.1703353219502162e-05, | |
| "loss": 0.0412, | |
| "step": 466000 | |
| }, | |
| { | |
| "epoch": 47.83412030383905, | |
| "eval_cer": 0.015550947867298577, | |
| "eval_loss": 0.024535449221730232, | |
| "eval_runtime": 22.2342, | |
| "eval_samples_per_second": 4.048, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.049151027703306524, | |
| "step": 466000 | |
| }, | |
| { | |
| "epoch": 47.93676863067132, | |
| "grad_norm": 0.15053987503051758, | |
| "learning_rate": 2.0674758280189262e-05, | |
| "loss": 0.0416, | |
| "step": 467000 | |
| }, | |
| { | |
| "epoch": 47.93676863067132, | |
| "eval_cer": 0.015106635071090047, | |
| "eval_loss": 0.024381397292017937, | |
| "eval_runtime": 22.2314, | |
| "eval_samples_per_second": 4.048, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 467000 | |
| }, | |
| { | |
| "epoch": 48.03941695750359, | |
| "grad_norm": 0.2812643051147461, | |
| "learning_rate": 1.9646163340876362e-05, | |
| "loss": 0.0417, | |
| "step": 468000 | |
| }, | |
| { | |
| "epoch": 48.03941695750359, | |
| "eval_cer": 0.01481042654028436, | |
| "eval_loss": 0.02442990057170391, | |
| "eval_runtime": 22.3372, | |
| "eval_samples_per_second": 4.029, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 468000 | |
| }, | |
| { | |
| "epoch": 48.142065284335864, | |
| "grad_norm": 0.31642502546310425, | |
| "learning_rate": 1.8617568401563465e-05, | |
| "loss": 0.0419, | |
| "step": 469000 | |
| }, | |
| { | |
| "epoch": 48.142065284335864, | |
| "eval_cer": 0.015106635071090047, | |
| "eval_loss": 0.024683654308319092, | |
| "eval_runtime": 22.1497, | |
| "eval_samples_per_second": 4.063, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 469000 | |
| }, | |
| { | |
| "epoch": 48.24471361116814, | |
| "grad_norm": 0.24842867255210876, | |
| "learning_rate": 1.7588973462250565e-05, | |
| "loss": 0.0413, | |
| "step": 470000 | |
| }, | |
| { | |
| "epoch": 48.24471361116814, | |
| "eval_cer": 0.015106635071090047, | |
| "eval_loss": 0.0246568750590086, | |
| "eval_runtime": 22.1198, | |
| "eval_samples_per_second": 4.069, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 470000 | |
| }, | |
| { | |
| "epoch": 48.34736193800041, | |
| "grad_norm": 0.22725574672222137, | |
| "learning_rate": 1.6560378522937665e-05, | |
| "loss": 0.0417, | |
| "step": 471000 | |
| }, | |
| { | |
| "epoch": 48.34736193800041, | |
| "eval_cer": 0.015106635071090047, | |
| "eval_loss": 0.024483025074005127, | |
| "eval_runtime": 22.0753, | |
| "eval_samples_per_second": 4.077, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 471000 | |
| }, | |
| { | |
| "epoch": 48.450010264832684, | |
| "grad_norm": 0.415797621011734, | |
| "learning_rate": 1.553178358362477e-05, | |
| "loss": 0.0412, | |
| "step": 472000 | |
| }, | |
| { | |
| "epoch": 48.450010264832684, | |
| "eval_cer": 0.014366113744075829, | |
| "eval_loss": 0.024604879319667816, | |
| "eval_runtime": 22.2084, | |
| "eval_samples_per_second": 4.053, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.045576407506702415, | |
| "step": 472000 | |
| }, | |
| { | |
| "epoch": 48.55265859166496, | |
| "grad_norm": 0.4157191216945648, | |
| "learning_rate": 1.450318864431187e-05, | |
| "loss": 0.0416, | |
| "step": 473000 | |
| }, | |
| { | |
| "epoch": 48.55265859166496, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.024540413171052933, | |
| "eval_runtime": 21.9902, | |
| "eval_samples_per_second": 4.093, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 473000 | |
| }, | |
| { | |
| "epoch": 48.65530691849723, | |
| "grad_norm": 0.20073458552360535, | |
| "learning_rate": 1.3474593704998972e-05, | |
| "loss": 0.0413, | |
| "step": 474000 | |
| }, | |
| { | |
| "epoch": 48.65530691849723, | |
| "eval_cer": 0.014958530805687204, | |
| "eval_loss": 0.02452738583087921, | |
| "eval_runtime": 22.0306, | |
| "eval_samples_per_second": 4.085, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 474000 | |
| }, | |
| { | |
| "epoch": 48.7579552453295, | |
| "grad_norm": 0.3352334499359131, | |
| "learning_rate": 1.2445998765686073e-05, | |
| "loss": 0.0413, | |
| "step": 475000 | |
| }, | |
| { | |
| "epoch": 48.7579552453295, | |
| "eval_cer": 0.014958530805687204, | |
| "eval_loss": 0.024565977975726128, | |
| "eval_runtime": 22.1491, | |
| "eval_samples_per_second": 4.063, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 475000 | |
| }, | |
| { | |
| "epoch": 48.86060357216177, | |
| "grad_norm": 0.20034602284431458, | |
| "learning_rate": 1.1417403826373175e-05, | |
| "loss": 0.0418, | |
| "step": 476000 | |
| }, | |
| { | |
| "epoch": 48.86060357216177, | |
| "eval_cer": 0.015106635071090047, | |
| "eval_loss": 0.024530308321118355, | |
| "eval_runtime": 22.0802, | |
| "eval_samples_per_second": 4.076, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 476000 | |
| }, | |
| { | |
| "epoch": 48.963251898994045, | |
| "grad_norm": 0.204274982213974, | |
| "learning_rate": 1.0388808887060275e-05, | |
| "loss": 0.0417, | |
| "step": 477000 | |
| }, | |
| { | |
| "epoch": 48.963251898994045, | |
| "eval_cer": 0.015106635071090047, | |
| "eval_loss": 0.02437027543783188, | |
| "eval_runtime": 22.051, | |
| "eval_samples_per_second": 4.081, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 477000 | |
| }, | |
| { | |
| "epoch": 49.06590022582632, | |
| "grad_norm": 0.27974188327789307, | |
| "learning_rate": 9.360213947747377e-06, | |
| "loss": 0.0417, | |
| "step": 478000 | |
| }, | |
| { | |
| "epoch": 49.06590022582632, | |
| "eval_cer": 0.015402843601895734, | |
| "eval_loss": 0.024484841153025627, | |
| "eval_runtime": 22.1209, | |
| "eval_samples_per_second": 4.069, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04825737265415549, | |
| "step": 478000 | |
| }, | |
| { | |
| "epoch": 49.16854855265859, | |
| "grad_norm": 0.15514741837978363, | |
| "learning_rate": 8.33161900843448e-06, | |
| "loss": 0.0413, | |
| "step": 479000 | |
| }, | |
| { | |
| "epoch": 49.16854855265859, | |
| "eval_cer": 0.014958530805687204, | |
| "eval_loss": 0.024350464344024658, | |
| "eval_runtime": 21.9927, | |
| "eval_samples_per_second": 4.092, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 479000 | |
| }, | |
| { | |
| "epoch": 49.271196879490866, | |
| "grad_norm": 0.13631823658943176, | |
| "learning_rate": 7.30302406912158e-06, | |
| "loss": 0.0413, | |
| "step": 480000 | |
| }, | |
| { | |
| "epoch": 49.271196879490866, | |
| "eval_cer": 0.015106635071090047, | |
| "eval_loss": 0.024330323562026024, | |
| "eval_runtime": 22.2136, | |
| "eval_samples_per_second": 4.052, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 480000 | |
| }, | |
| { | |
| "epoch": 49.37384520632314, | |
| "grad_norm": 0.18198832869529724, | |
| "learning_rate": 6.274429129808681e-06, | |
| "loss": 0.0409, | |
| "step": 481000 | |
| }, | |
| { | |
| "epoch": 49.37384520632314, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.024377064779400826, | |
| "eval_runtime": 22.1427, | |
| "eval_samples_per_second": 4.065, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 481000 | |
| }, | |
| { | |
| "epoch": 49.47649353315541, | |
| "grad_norm": 0.20921219885349274, | |
| "learning_rate": 5.245834190495783e-06, | |
| "loss": 0.0413, | |
| "step": 482000 | |
| }, | |
| { | |
| "epoch": 49.47649353315541, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.024373607710003853, | |
| "eval_runtime": 22.062, | |
| "eval_samples_per_second": 4.079, | |
| "eval_steps_per_second": 0.045, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 482000 | |
| }, | |
| { | |
| "epoch": 49.57914185998768, | |
| "grad_norm": 0.24988599121570587, | |
| "learning_rate": 4.217239251182884e-06, | |
| "loss": 0.0414, | |
| "step": 483000 | |
| }, | |
| { | |
| "epoch": 49.57914185998768, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.024304602295160294, | |
| "eval_runtime": 12.6617, | |
| "eval_samples_per_second": 7.108, | |
| "eval_steps_per_second": 0.079, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 483000 | |
| }, | |
| { | |
| "epoch": 49.68179018681995, | |
| "grad_norm": 0.22157305479049683, | |
| "learning_rate": 3.1886443118699856e-06, | |
| "loss": 0.0407, | |
| "step": 484000 | |
| }, | |
| { | |
| "epoch": 49.68179018681995, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.02429259568452835, | |
| "eval_runtime": 12.7319, | |
| "eval_samples_per_second": 7.069, | |
| "eval_steps_per_second": 0.079, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 484000 | |
| }, | |
| { | |
| "epoch": 49.784438513652226, | |
| "grad_norm": 0.18318428099155426, | |
| "learning_rate": 2.1600493725570872e-06, | |
| "loss": 0.0413, | |
| "step": 485000 | |
| }, | |
| { | |
| "epoch": 49.784438513652226, | |
| "eval_cer": 0.014662322274881517, | |
| "eval_loss": 0.024326322600245476, | |
| "eval_runtime": 12.7187, | |
| "eval_samples_per_second": 7.076, | |
| "eval_steps_per_second": 0.079, | |
| "eval_wer": 0.04647006255585344, | |
| "step": 485000 | |
| }, | |
| { | |
| "epoch": 49.8870868404845, | |
| "grad_norm": 0.23897279798984528, | |
| "learning_rate": 1.1314544332441885e-06, | |
| "loss": 0.0416, | |
| "step": 486000 | |
| }, | |
| { | |
| "epoch": 49.8870868404845, | |
| "eval_cer": 0.01481042654028436, | |
| "eval_loss": 0.024302508682012558, | |
| "eval_runtime": 12.6346, | |
| "eval_samples_per_second": 7.123, | |
| "eval_steps_per_second": 0.079, | |
| "eval_wer": 0.04736371760500447, | |
| "step": 486000 | |
| }, | |
| { | |
| "epoch": 49.98973516731677, | |
| "grad_norm": 0.2672029733657837, | |
| "learning_rate": 1.0285949393128985e-07, | |
| "loss": 0.0416, | |
| "step": 487000 | |
| }, | |
| { | |
| "epoch": 49.98973516731677, | |
| "eval_cer": 0.01481042654028436, | |
| "eval_loss": 0.024304790422320366, | |
| "eval_runtime": 12.659, | |
| "eval_samples_per_second": 7.11, | |
| "eval_steps_per_second": 0.079, | |
| "eval_wer": 0.04736371760500447, | |
| "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 | |
| } | |