| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 49.99896587383661, | |
| "global_step": 32200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 0.000998, | |
| "loss": 2.9272, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "eval_cer": 0.23553399465944927, | |
| "eval_loss": 0.7603499889373779, | |
| "eval_runtime": 139.9494, | |
| "eval_samples_per_second": 27.46, | |
| "eval_steps_per_second": 3.437, | |
| "eval_wer": 0.7733832367653939, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 0.0009842586750788643, | |
| "loss": 0.6157, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "eval_cer": 0.12468736029385112, | |
| "eval_loss": 0.40030696988105774, | |
| "eval_runtime": 143.7328, | |
| "eval_samples_per_second": 26.737, | |
| "eval_steps_per_second": 3.346, | |
| "eval_wer": 0.48655846259041996, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "learning_rate": 0.0009684858044164038, | |
| "loss": 0.4452, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "eval_cer": 0.0910430989693462, | |
| "eval_loss": 0.2960306406021118, | |
| "eval_runtime": 141.1265, | |
| "eval_samples_per_second": 27.231, | |
| "eval_steps_per_second": 3.408, | |
| "eval_wer": 0.3688775326591572, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "learning_rate": 0.0009527129337539432, | |
| "loss": 0.3631, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "eval_cer": 0.07961287048560348, | |
| "eval_loss": 0.25798073410987854, | |
| "eval_runtime": 138.8783, | |
| "eval_samples_per_second": 27.672, | |
| "eval_steps_per_second": 3.463, | |
| "eval_wer": 0.3205455788678159, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.88, | |
| "learning_rate": 0.0009369400630914827, | |
| "loss": 0.3153, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.88, | |
| "eval_cer": 0.07467709000398728, | |
| "eval_loss": 0.24647507071495056, | |
| "eval_runtime": 136.7566, | |
| "eval_samples_per_second": 28.101, | |
| "eval_steps_per_second": 3.517, | |
| "eval_wer": 0.29772915392089827, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.66, | |
| "learning_rate": 0.0009211671924290221, | |
| "loss": 0.2795, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.66, | |
| "eval_cer": 0.06940299409156266, | |
| "eval_loss": 0.2273630052804947, | |
| "eval_runtime": 144.7547, | |
| "eval_samples_per_second": 26.548, | |
| "eval_steps_per_second": 3.323, | |
| "eval_wer": 0.27890740274228953, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 5.43, | |
| "learning_rate": 0.0009053943217665615, | |
| "loss": 0.2615, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 5.43, | |
| "eval_cer": 0.06749392844628638, | |
| "eval_loss": 0.22766220569610596, | |
| "eval_runtime": 139.1146, | |
| "eval_samples_per_second": 27.625, | |
| "eval_steps_per_second": 3.458, | |
| "eval_wer": 0.26850685572389965, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 6.21, | |
| "learning_rate": 0.000889621451104101, | |
| "loss": 0.2389, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 6.21, | |
| "eval_cer": 0.06267897490424465, | |
| "eval_loss": 0.21347470581531525, | |
| "eval_runtime": 143.4039, | |
| "eval_samples_per_second": 26.798, | |
| "eval_steps_per_second": 3.354, | |
| "eval_wer": 0.2517724115593623, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 6.99, | |
| "learning_rate": 0.0008738485804416404, | |
| "loss": 0.2229, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 6.99, | |
| "eval_cer": 0.06139216799777678, | |
| "eval_loss": 0.20536719262599945, | |
| "eval_runtime": 140.3579, | |
| "eval_samples_per_second": 27.38, | |
| "eval_steps_per_second": 3.427, | |
| "eval_wer": 0.24493468168568036, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 7.76, | |
| "learning_rate": 0.0008580757097791798, | |
| "loss": 0.2067, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 7.76, | |
| "eval_cer": 0.05967038410884091, | |
| "eval_loss": 0.20964160561561584, | |
| "eval_runtime": 144.9834, | |
| "eval_samples_per_second": 26.506, | |
| "eval_steps_per_second": 3.318, | |
| "eval_wer": 0.23780904739626443, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 8.54, | |
| "learning_rate": 0.0008423028391167192, | |
| "loss": 0.1977, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 8.54, | |
| "eval_cer": 0.06000265819267064, | |
| "eval_loss": 0.20422270894050598, | |
| "eval_runtime": 139.8836, | |
| "eval_samples_per_second": 27.473, | |
| "eval_steps_per_second": 3.439, | |
| "eval_wer": 0.23874473674739985, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 9.32, | |
| "learning_rate": 0.0008265299684542586, | |
| "loss": 0.1896, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 9.32, | |
| "eval_cer": 0.059458936964585624, | |
| "eval_loss": 0.2110028713941574, | |
| "eval_runtime": 139.3458, | |
| "eval_samples_per_second": 27.579, | |
| "eval_steps_per_second": 3.452, | |
| "eval_wer": 0.23827689207183214, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 10.09, | |
| "learning_rate": 0.000810757097791798, | |
| "loss": 0.1801, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 10.09, | |
| "eval_cer": 0.05477689305607578, | |
| "eval_loss": 0.19087238609790802, | |
| "eval_runtime": 138.0429, | |
| "eval_samples_per_second": 27.839, | |
| "eval_steps_per_second": 3.484, | |
| "eval_wer": 0.21650412063195018, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 10.87, | |
| "learning_rate": 0.0007949842271293376, | |
| "loss": 0.174, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 10.87, | |
| "eval_cer": 0.05590662494109687, | |
| "eval_loss": 0.18831488490104675, | |
| "eval_runtime": 140.5023, | |
| "eval_samples_per_second": 27.352, | |
| "eval_steps_per_second": 3.423, | |
| "eval_wer": 0.2206427466081261, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 11.65, | |
| "learning_rate": 0.000779211356466877, | |
| "loss": 0.1685, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 11.65, | |
| "eval_cer": 0.052807413941012286, | |
| "eval_loss": 0.1847681999206543, | |
| "eval_runtime": 140.5872, | |
| "eval_samples_per_second": 27.335, | |
| "eval_steps_per_second": 3.421, | |
| "eval_wer": 0.209702378810235, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 12.42, | |
| "learning_rate": 0.0007634384858044164, | |
| "loss": 0.1591, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 12.42, | |
| "eval_cer": 0.05136353201309764, | |
| "eval_loss": 0.18508584797382355, | |
| "eval_runtime": 138.1047, | |
| "eval_samples_per_second": 27.827, | |
| "eval_steps_per_second": 3.483, | |
| "eval_wer": 0.20394429049555549, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 13.2, | |
| "learning_rate": 0.0007476971608832807, | |
| "loss": 0.1537, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 13.2, | |
| "eval_cer": 0.05175017821973587, | |
| "eval_loss": 0.18811114132404327, | |
| "eval_runtime": 143.7666, | |
| "eval_samples_per_second": 26.731, | |
| "eval_steps_per_second": 3.346, | |
| "eval_wer": 0.20653543023716125, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 13.97, | |
| "learning_rate": 0.0007319242902208201, | |
| "loss": 0.1504, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 13.97, | |
| "eval_cer": 0.04994381547309788, | |
| "eval_loss": 0.18396763503551483, | |
| "eval_runtime": 144.751, | |
| "eval_samples_per_second": 26.549, | |
| "eval_steps_per_second": 3.323, | |
| "eval_wer": 0.1972145247777738, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 14.75, | |
| "learning_rate": 0.0007161514195583596, | |
| "loss": 0.145, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 14.75, | |
| "eval_cer": 0.05168976474994865, | |
| "eval_loss": 0.1845300793647766, | |
| "eval_runtime": 142.1949, | |
| "eval_samples_per_second": 27.026, | |
| "eval_steps_per_second": 3.383, | |
| "eval_wer": 0.20293662504048657, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 15.53, | |
| "learning_rate": 0.000700378548895899, | |
| "loss": 0.1417, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 15.53, | |
| "eval_cer": 0.050717107886374346, | |
| "eval_loss": 0.18841156363487244, | |
| "eval_runtime": 137.4986, | |
| "eval_samples_per_second": 27.949, | |
| "eval_steps_per_second": 3.498, | |
| "eval_wer": 0.20027350919494727, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 16.3, | |
| "learning_rate": 0.0006846056782334384, | |
| "loss": 0.1364, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 16.3, | |
| "eval_cer": 0.05172601283182098, | |
| "eval_loss": 0.20098073780536652, | |
| "eval_runtime": 138.6706, | |
| "eval_samples_per_second": 27.713, | |
| "eval_steps_per_second": 3.469, | |
| "eval_wer": 0.2036563860798215, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 17.08, | |
| "learning_rate": 0.0006688643533123028, | |
| "loss": 0.1331, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 17.08, | |
| "eval_cer": 0.04831265178884284, | |
| "eval_loss": 0.183824360370636, | |
| "eval_runtime": 137.3455, | |
| "eval_samples_per_second": 27.981, | |
| "eval_steps_per_second": 3.502, | |
| "eval_wer": 0.19232014971029618, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 17.86, | |
| "learning_rate": 0.0006530914826498424, | |
| "loss": 0.129, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 17.86, | |
| "eval_cer": 0.048886579751821466, | |
| "eval_loss": 0.18181012570858002, | |
| "eval_runtime": 140.6277, | |
| "eval_samples_per_second": 27.327, | |
| "eval_steps_per_second": 3.42, | |
| "eval_wer": 0.19221218555439595, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 18.63, | |
| "learning_rate": 0.0006373186119873818, | |
| "loss": 0.1198, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 18.63, | |
| "eval_cer": 0.04653649577709846, | |
| "eval_loss": 0.17603321373462677, | |
| "eval_runtime": 144.3916, | |
| "eval_samples_per_second": 26.615, | |
| "eval_steps_per_second": 3.331, | |
| "eval_wer": 0.1861302047720157, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 19.41, | |
| "learning_rate": 0.0006215457413249212, | |
| "loss": 0.1203, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 19.41, | |
| "eval_cer": 0.04645795826637507, | |
| "eval_loss": 0.168550506234169, | |
| "eval_runtime": 142.0056, | |
| "eval_samples_per_second": 27.062, | |
| "eval_steps_per_second": 3.387, | |
| "eval_wer": 0.18393493360204413, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 20.19, | |
| "learning_rate": 0.0006058044164037855, | |
| "loss": 0.1225, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 20.19, | |
| "eval_cer": 0.047938088276162055, | |
| "eval_loss": 0.18283818662166595, | |
| "eval_runtime": 142.8894, | |
| "eval_samples_per_second": 26.895, | |
| "eval_steps_per_second": 3.366, | |
| "eval_wer": 0.19199625724259545, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 20.96, | |
| "learning_rate": 0.0005900315457413249, | |
| "loss": 0.1145, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 20.96, | |
| "eval_cer": 0.04463347147880092, | |
| "eval_loss": 0.16731047630310059, | |
| "eval_runtime": 141.4047, | |
| "eval_samples_per_second": 27.177, | |
| "eval_steps_per_second": 3.402, | |
| "eval_wer": 0.17839277359916508, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 21.74, | |
| "learning_rate": 0.0005742586750788644, | |
| "loss": 0.1053, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 21.74, | |
| "eval_cer": 0.04560008699539649, | |
| "eval_loss": 0.1801901012659073, | |
| "eval_runtime": 142.8084, | |
| "eval_samples_per_second": 26.91, | |
| "eval_steps_per_second": 3.368, | |
| "eval_wer": 0.18098391334077085, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 22.51, | |
| "learning_rate": 0.0005584858044164039, | |
| "loss": 0.1071, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 22.51, | |
| "eval_cer": 0.044385776252673294, | |
| "eval_loss": 0.17691390216350555, | |
| "eval_runtime": 140.9354, | |
| "eval_samples_per_second": 27.268, | |
| "eval_steps_per_second": 3.413, | |
| "eval_wer": 0.17752906035196314, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 23.29, | |
| "learning_rate": 0.0005427444794952682, | |
| "loss": 0.1053, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 23.29, | |
| "eval_cer": 0.04566654181216244, | |
| "eval_loss": 0.19200138747692108, | |
| "eval_runtime": 142.1681, | |
| "eval_samples_per_second": 27.031, | |
| "eval_steps_per_second": 3.383, | |
| "eval_wer": 0.17832079749523158, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 24.07, | |
| "learning_rate": 0.0005270031545741325, | |
| "loss": 0.1024, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 24.07, | |
| "eval_cer": 0.04464555417275836, | |
| "eval_loss": 0.19041383266448975, | |
| "eval_runtime": 138.4185, | |
| "eval_samples_per_second": 27.764, | |
| "eval_steps_per_second": 3.475, | |
| "eval_wer": 0.17752906035196314, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 24.84, | |
| "learning_rate": 0.000511230283911672, | |
| "loss": 0.0987, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 24.84, | |
| "eval_cer": 0.044615347437864744, | |
| "eval_loss": 0.17931906878948212, | |
| "eval_runtime": 138.8631, | |
| "eval_samples_per_second": 27.675, | |
| "eval_steps_per_second": 3.464, | |
| "eval_wer": 0.17616151437722677, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 25.62, | |
| "learning_rate": 0.0004954574132492114, | |
| "loss": 0.0949, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 25.62, | |
| "eval_cer": 0.04433744547684352, | |
| "eval_loss": 0.1801224648952484, | |
| "eval_runtime": 140.9756, | |
| "eval_samples_per_second": 27.26, | |
| "eval_steps_per_second": 3.412, | |
| "eval_wer": 0.17655738294886097, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 26.4, | |
| "learning_rate": 0.00047968454258675084, | |
| "loss": 0.0942, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 26.4, | |
| "eval_cer": 0.04232567693292896, | |
| "eval_loss": 0.17306886613368988, | |
| "eval_runtime": 140.9004, | |
| "eval_samples_per_second": 27.275, | |
| "eval_steps_per_second": 3.414, | |
| "eval_wer": 0.1659409076186706, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 27.17, | |
| "learning_rate": 0.00046391167192429025, | |
| "loss": 0.0906, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 27.17, | |
| "eval_cer": 0.042374007708758744, | |
| "eval_loss": 0.17762278020381927, | |
| "eval_runtime": 142.288, | |
| "eval_samples_per_second": 27.009, | |
| "eval_steps_per_second": 3.38, | |
| "eval_wer": 0.16975564112714578, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 27.95, | |
| "learning_rate": 0.00044813880126182967, | |
| "loss": 0.0861, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 27.95, | |
| "eval_cer": 0.040622017084929254, | |
| "eval_loss": 0.17157311737537384, | |
| "eval_runtime": 139.3451, | |
| "eval_samples_per_second": 27.579, | |
| "eval_steps_per_second": 3.452, | |
| "eval_wer": 0.16000287904415733, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 28.73, | |
| "learning_rate": 0.00043239747634069405, | |
| "loss": 0.0851, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 28.73, | |
| "eval_cer": 0.041008663291567486, | |
| "eval_loss": 0.16616342961788177, | |
| "eval_runtime": 138.87, | |
| "eval_samples_per_second": 27.673, | |
| "eval_steps_per_second": 3.464, | |
| "eval_wer": 0.1629538993054306, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 29.5, | |
| "learning_rate": 0.00041662460567823347, | |
| "loss": 0.0844, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 29.5, | |
| "eval_cer": 0.039298962096589056, | |
| "eval_loss": 0.16713382303714752, | |
| "eval_runtime": 139.8577, | |
| "eval_samples_per_second": 27.478, | |
| "eval_steps_per_second": 3.439, | |
| "eval_wer": 0.1572317990427178, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 30.28, | |
| "learning_rate": 0.0004008517350157729, | |
| "loss": 0.0792, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 30.28, | |
| "eval_cer": 0.04074888537148243, | |
| "eval_loss": 0.17681923508644104, | |
| "eval_runtime": 140.8425, | |
| "eval_samples_per_second": 27.286, | |
| "eval_steps_per_second": 3.415, | |
| "eval_wer": 0.15993090294022386, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 31.06, | |
| "learning_rate": 0.0003850788643533123, | |
| "loss": 0.0798, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 31.06, | |
| "eval_cer": 0.039365416913355, | |
| "eval_loss": 0.17318624258041382, | |
| "eval_runtime": 138.8235, | |
| "eval_samples_per_second": 27.683, | |
| "eval_steps_per_second": 3.465, | |
| "eval_wer": 0.15579227696404793, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 31.83, | |
| "learning_rate": 0.0003693059936908517, | |
| "loss": 0.0779, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 31.83, | |
| "eval_cer": 0.03882773703224871, | |
| "eval_loss": 0.1694292426109314, | |
| "eval_runtime": 142.3851, | |
| "eval_samples_per_second": 26.99, | |
| "eval_steps_per_second": 3.378, | |
| "eval_wer": 0.15442473098931156, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 32.61, | |
| "learning_rate": 0.0003535331230283912, | |
| "loss": 0.0718, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 32.61, | |
| "eval_cer": 0.03985476601863151, | |
| "eval_loss": 0.17093372344970703, | |
| "eval_runtime": 144.4958, | |
| "eval_samples_per_second": 26.596, | |
| "eval_steps_per_second": 3.329, | |
| "eval_wer": 0.15784359592615252, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 33.38, | |
| "learning_rate": 0.0003377917981072555, | |
| "loss": 0.0732, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 33.38, | |
| "eval_cer": 0.03907543225837633, | |
| "eval_loss": 0.1696743220090866, | |
| "eval_runtime": 141.4742, | |
| "eval_samples_per_second": 27.164, | |
| "eval_steps_per_second": 3.4, | |
| "eval_wer": 0.15226544787130672, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 34.16, | |
| "learning_rate": 0.00032201892744479493, | |
| "loss": 0.0708, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 34.16, | |
| "eval_cer": 0.037462392615057453, | |
| "eval_loss": 0.16164687275886536, | |
| "eval_runtime": 142.4307, | |
| "eval_samples_per_second": 26.982, | |
| "eval_steps_per_second": 3.377, | |
| "eval_wer": 0.14744304890776264, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 34.94, | |
| "learning_rate": 0.0003062460567823344, | |
| "loss": 0.0678, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 34.94, | |
| "eval_cer": 0.037498640696929784, | |
| "eval_loss": 0.16977904736995697, | |
| "eval_runtime": 140.9926, | |
| "eval_samples_per_second": 27.257, | |
| "eval_steps_per_second": 3.412, | |
| "eval_wer": 0.14737107280382913, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 35.71, | |
| "learning_rate": 0.0002904731861198738, | |
| "loss": 0.0642, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 35.71, | |
| "eval_cer": 0.036900547346036275, | |
| "eval_loss": 0.16813962161540985, | |
| "eval_runtime": 139.7128, | |
| "eval_samples_per_second": 27.506, | |
| "eval_steps_per_second": 3.443, | |
| "eval_wer": 0.14593155072515926, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 36.49, | |
| "learning_rate": 0.00027470031545741324, | |
| "loss": 0.0661, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 36.49, | |
| "eval_cer": 0.03566207121539818, | |
| "eval_loss": 0.1611974984407425, | |
| "eval_runtime": 141.2001, | |
| "eval_samples_per_second": 27.217, | |
| "eval_steps_per_second": 3.407, | |
| "eval_wer": 0.14110915176161515, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 37.27, | |
| "learning_rate": 0.00025892744479495266, | |
| "loss": 0.0629, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 37.27, | |
| "eval_cer": 0.035529161581866295, | |
| "eval_loss": 0.1662248969078064, | |
| "eval_runtime": 141.4579, | |
| "eval_samples_per_second": 27.167, | |
| "eval_steps_per_second": 3.4, | |
| "eval_wer": 0.14143304422931588, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 38.04, | |
| "learning_rate": 0.00024315457413249213, | |
| "loss": 0.0587, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 38.04, | |
| "eval_cer": 0.03514251537522806, | |
| "eval_loss": 0.16591240465641022, | |
| "eval_runtime": 138.7975, | |
| "eval_samples_per_second": 27.688, | |
| "eval_steps_per_second": 3.465, | |
| "eval_wer": 0.14078525929391442, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 38.82, | |
| "learning_rate": 0.00022738170347003155, | |
| "loss": 0.0581, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 38.82, | |
| "eval_cer": 0.03515459806918551, | |
| "eval_loss": 0.1612222045660019, | |
| "eval_runtime": 140.4043, | |
| "eval_samples_per_second": 27.371, | |
| "eval_steps_per_second": 3.426, | |
| "eval_wer": 0.13823010760427537, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 39.6, | |
| "learning_rate": 0.000211608832807571, | |
| "loss": 0.0556, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 39.6, | |
| "eval_cer": 0.03452629798339838, | |
| "eval_loss": 0.16471649706363678, | |
| "eval_runtime": 139.8247, | |
| "eval_samples_per_second": 27.484, | |
| "eval_steps_per_second": 3.44, | |
| "eval_wer": 0.1376183107208407, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 40.37, | |
| "learning_rate": 0.00019586750788643534, | |
| "loss": 0.0543, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 40.37, | |
| "eval_cer": 0.03374092287616447, | |
| "eval_loss": 0.16579236090183258, | |
| "eval_runtime": 140.5138, | |
| "eval_samples_per_second": 27.35, | |
| "eval_steps_per_second": 3.423, | |
| "eval_wer": 0.13347968474466476, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 41.15, | |
| "learning_rate": 0.00018009463722397476, | |
| "loss": 0.052, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 41.15, | |
| "eval_cer": 0.03429068545122821, | |
| "eval_loss": 0.17156867682933807, | |
| "eval_runtime": 144.8911, | |
| "eval_samples_per_second": 26.523, | |
| "eval_steps_per_second": 3.32, | |
| "eval_wer": 0.1369345377334725, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 41.92, | |
| "learning_rate": 0.0001643217665615142, | |
| "loss": 0.0513, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 41.92, | |
| "eval_cer": 0.03304012662663267, | |
| "eval_loss": 0.15995080769062042, | |
| "eval_runtime": 141.0194, | |
| "eval_samples_per_second": 27.252, | |
| "eval_steps_per_second": 3.411, | |
| "eval_wer": 0.13168028214632743, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 42.7, | |
| "learning_rate": 0.00014854889589905362, | |
| "loss": 0.0491, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 42.7, | |
| "eval_cer": 0.03279243140050506, | |
| "eval_loss": 0.16710308194160461, | |
| "eval_runtime": 143.6022, | |
| "eval_samples_per_second": 26.761, | |
| "eval_steps_per_second": 3.35, | |
| "eval_wer": 0.13110447331485947, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 43.48, | |
| "learning_rate": 0.0001328391167192429, | |
| "loss": 0.0463, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 43.48, | |
| "eval_cer": 0.032375578458973216, | |
| "eval_loss": 0.16127389669418335, | |
| "eval_runtime": 141.8046, | |
| "eval_samples_per_second": 27.101, | |
| "eval_steps_per_second": 3.392, | |
| "eval_wer": 0.1289092021448879, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 44.25, | |
| "learning_rate": 0.00011706624605678232, | |
| "loss": 0.0468, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 44.25, | |
| "eval_cer": 0.031523748534973355, | |
| "eval_loss": 0.15992802381515503, | |
| "eval_runtime": 142.9591, | |
| "eval_samples_per_second": 26.882, | |
| "eval_steps_per_second": 3.365, | |
| "eval_wer": 0.12599416993558138, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 45.03, | |
| "learning_rate": 0.00010129337539432177, | |
| "loss": 0.0435, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 45.03, | |
| "eval_cer": 0.030816910938462838, | |
| "eval_loss": 0.15560874342918396, | |
| "eval_runtime": 140.0872, | |
| "eval_samples_per_second": 27.433, | |
| "eval_steps_per_second": 3.434, | |
| "eval_wer": 0.12318710188217512, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 45.81, | |
| "learning_rate": 8.555205047318612e-05, | |
| "loss": 0.043, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 45.81, | |
| "eval_cer": 0.03093773787803729, | |
| "eval_loss": 0.1587720513343811, | |
| "eval_runtime": 140.3288, | |
| "eval_samples_per_second": 27.386, | |
| "eval_steps_per_second": 3.428, | |
| "eval_wer": 0.12397883902544356, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 46.58, | |
| "learning_rate": 6.977917981072555e-05, | |
| "loss": 0.0421, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 46.58, | |
| "eval_cer": 0.03079274555054795, | |
| "eval_loss": 0.1567341834306717, | |
| "eval_runtime": 141.7614, | |
| "eval_samples_per_second": 27.109, | |
| "eval_steps_per_second": 3.393, | |
| "eval_wer": 0.12167560369957174, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 47.36, | |
| "learning_rate": 5.400630914826499e-05, | |
| "loss": 0.04, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 47.36, | |
| "eval_cer": 0.0301584041177821, | |
| "eval_loss": 0.1532924771308899, | |
| "eval_runtime": 143.1401, | |
| "eval_samples_per_second": 26.848, | |
| "eval_steps_per_second": 3.36, | |
| "eval_wer": 0.11984021304926765, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 48.14, | |
| "learning_rate": 3.823343848580442e-05, | |
| "loss": 0.0389, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 48.14, | |
| "eval_cer": 0.0296992617473992, | |
| "eval_loss": 0.15820840001106262, | |
| "eval_runtime": 144.2309, | |
| "eval_samples_per_second": 26.645, | |
| "eval_steps_per_second": 3.335, | |
| "eval_wer": 0.11847266707453126, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 48.91, | |
| "learning_rate": 2.2460567823343848e-05, | |
| "loss": 0.0387, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 48.91, | |
| "eval_cer": 0.029675096359484312, | |
| "eval_loss": 0.1575988531112671, | |
| "eval_runtime": 140.1473, | |
| "eval_samples_per_second": 27.421, | |
| "eval_steps_per_second": 3.432, | |
| "eval_wer": 0.11872458343829849, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 49.69, | |
| "learning_rate": 6.687697160883281e-06, | |
| "loss": 0.0376, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 49.69, | |
| "eval_cer": 0.02951802133803753, | |
| "eval_loss": 0.15600277483463287, | |
| "eval_runtime": 140.2646, | |
| "eval_samples_per_second": 27.398, | |
| "eval_steps_per_second": 3.429, | |
| "eval_wer": 0.11822075071076403, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "step": 32200, | |
| "total_flos": 1.8729227608976848e+20, | |
| "train_loss": 0.17004985039278586, | |
| "train_runtime": 77839.5185, | |
| "train_samples_per_second": 19.876, | |
| "train_steps_per_second": 0.414 | |
| } | |
| ], | |
| "max_steps": 32200, | |
| "num_train_epochs": 50, | |
| "total_flos": 1.8729227608976848e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |