| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 99.9728, | |
| "global_step": 46800, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 2.976e-05, | |
| "loss": 9.3762, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "learning_rate": 2.9678617710583153e-05, | |
| "loss": 2.6416, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "eval_cer": 0.15406803531769678, | |
| "eval_loss": 0.8599101305007935, | |
| "eval_runtime": 83.463, | |
| "eval_samples_per_second": 17.972, | |
| "eval_steps_per_second": 2.252, | |
| "eval_wer": 0.644871073981846, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 2.9354643628509722e-05, | |
| "loss": 0.506, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "learning_rate": 2.9030669546436284e-05, | |
| "loss": 0.2633, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "eval_cer": 0.033489518444287956, | |
| "eval_loss": 0.18973715603351593, | |
| "eval_runtime": 81.5499, | |
| "eval_samples_per_second": 18.394, | |
| "eval_steps_per_second": 2.305, | |
| "eval_wer": 0.14306369989557394, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 5.33, | |
| "learning_rate": 2.8706695464362852e-05, | |
| "loss": 0.2016, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "learning_rate": 2.8383369330453565e-05, | |
| "loss": 0.1739, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "eval_cer": 0.028912301608796924, | |
| "eval_loss": 0.17320764064788818, | |
| "eval_runtime": 78.6521, | |
| "eval_samples_per_second": 19.071, | |
| "eval_steps_per_second": 2.39, | |
| "eval_wer": 0.11454735320106033, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 7.47, | |
| "learning_rate": 2.805939524838013e-05, | |
| "loss": 0.1493, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 8.53, | |
| "learning_rate": 2.7735421166306696e-05, | |
| "loss": 0.1378, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 8.53, | |
| "eval_cer": 0.027585179567737393, | |
| "eval_loss": 0.17293226718902588, | |
| "eval_runtime": 72.8342, | |
| "eval_samples_per_second": 20.595, | |
| "eval_steps_per_second": 2.581, | |
| "eval_wer": 0.10659490722146357, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "learning_rate": 2.741144708423326e-05, | |
| "loss": 0.127, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 10.67, | |
| "learning_rate": 2.708747300215983e-05, | |
| "loss": 0.1172, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 10.67, | |
| "eval_cer": 0.02684036617734684, | |
| "eval_loss": 0.17730334401130676, | |
| "eval_runtime": 69.8987, | |
| "eval_samples_per_second": 21.46, | |
| "eval_steps_per_second": 2.69, | |
| "eval_wer": 0.10185557072857257, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 11.73, | |
| "learning_rate": 2.676349892008639e-05, | |
| "loss": 0.1112, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "learning_rate": 2.643952483801296e-05, | |
| "loss": 0.1049, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "eval_cer": 0.02552678619793077, | |
| "eval_loss": 0.1701209545135498, | |
| "eval_runtime": 70.2839, | |
| "eval_samples_per_second": 21.342, | |
| "eval_steps_per_second": 2.675, | |
| "eval_wer": 0.09374246927464053, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 13.87, | |
| "learning_rate": 2.6115550755939525e-05, | |
| "loss": 0.1026, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 14.93, | |
| "learning_rate": 2.579157667386609e-05, | |
| "loss": 0.0951, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 14.93, | |
| "eval_cer": 0.025296571149991873, | |
| "eval_loss": 0.17183555662631989, | |
| "eval_runtime": 70.8748, | |
| "eval_samples_per_second": 21.164, | |
| "eval_steps_per_second": 2.653, | |
| "eval_wer": 0.09326050285163467, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "learning_rate": 2.546760259179266e-05, | |
| "loss": 0.0894, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 17.07, | |
| "learning_rate": 2.514362850971922e-05, | |
| "loss": 0.0851, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 17.07, | |
| "eval_cer": 0.02386111261578463, | |
| "eval_loss": 0.17866037786006927, | |
| "eval_runtime": 70.4791, | |
| "eval_samples_per_second": 21.283, | |
| "eval_steps_per_second": 2.667, | |
| "eval_wer": 0.08338019118001445, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 18.13, | |
| "learning_rate": 2.481965442764579e-05, | |
| "loss": 0.0841, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 19.2, | |
| "learning_rate": 2.4495680345572355e-05, | |
| "loss": 0.0809, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 19.2, | |
| "eval_cer": 0.023549645197984942, | |
| "eval_loss": 0.18024054169654846, | |
| "eval_runtime": 70.9657, | |
| "eval_samples_per_second": 21.137, | |
| "eval_steps_per_second": 2.649, | |
| "eval_wer": 0.08354084665434974, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 20.27, | |
| "learning_rate": 2.417170626349892e-05, | |
| "loss": 0.0768, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 21.33, | |
| "learning_rate": 2.384773218142549e-05, | |
| "loss": 0.0756, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 21.33, | |
| "eval_cer": 0.02386111261578463, | |
| "eval_loss": 0.17842911183834076, | |
| "eval_runtime": 71.3112, | |
| "eval_samples_per_second": 21.035, | |
| "eval_steps_per_second": 2.636, | |
| "eval_wer": 0.08554904008354085, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 22.4, | |
| "learning_rate": 2.352375809935205e-05, | |
| "loss": 0.0726, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 23.47, | |
| "learning_rate": 2.319978401727862e-05, | |
| "loss": 0.0708, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 23.47, | |
| "eval_cer": 0.023522561074698013, | |
| "eval_loss": 0.1748334914445877, | |
| "eval_runtime": 70.5277, | |
| "eval_samples_per_second": 21.268, | |
| "eval_steps_per_second": 2.666, | |
| "eval_wer": 0.08241625833400273, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 24.53, | |
| "learning_rate": 2.2875809935205184e-05, | |
| "loss": 0.0723, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 25.6, | |
| "learning_rate": 2.255183585313175e-05, | |
| "loss": 0.0657, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 25.6, | |
| "eval_cer": 0.022764205622663996, | |
| "eval_loss": 0.18295687437057495, | |
| "eval_runtime": 70.62, | |
| "eval_samples_per_second": 21.24, | |
| "eval_steps_per_second": 2.662, | |
| "eval_wer": 0.07960478753313519, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 26.67, | |
| "learning_rate": 2.2227861771058315e-05, | |
| "loss": 0.0673, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 27.73, | |
| "learning_rate": 2.1903887688984883e-05, | |
| "loss": 0.0605, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 27.73, | |
| "eval_cer": 0.022967336547315965, | |
| "eval_loss": 0.18958806991577148, | |
| "eval_runtime": 70.8663, | |
| "eval_samples_per_second": 21.167, | |
| "eval_steps_per_second": 2.653, | |
| "eval_wer": 0.07976544300747047, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 28.8, | |
| "learning_rate": 2.157991360691145e-05, | |
| "loss": 0.0629, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 29.87, | |
| "learning_rate": 2.1257235421166306e-05, | |
| "loss": 0.0583, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 29.87, | |
| "eval_cer": 0.02242565408157738, | |
| "eval_loss": 0.18885250389575958, | |
| "eval_runtime": 71.1848, | |
| "eval_samples_per_second": 21.072, | |
| "eval_steps_per_second": 2.641, | |
| "eval_wer": 0.07775724957827938, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 30.93, | |
| "learning_rate": 2.0933261339092874e-05, | |
| "loss": 0.0611, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "learning_rate": 2.060928725701944e-05, | |
| "loss": 0.0608, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "eval_cer": 0.022263149341855804, | |
| "eval_loss": 0.1849208027124405, | |
| "eval_runtime": 72.432, | |
| "eval_samples_per_second": 20.709, | |
| "eval_steps_per_second": 2.596, | |
| "eval_wer": 0.07574905614908828, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 33.07, | |
| "learning_rate": 2.0285313174946005e-05, | |
| "loss": 0.0574, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 34.13, | |
| "learning_rate": 1.9961987041036718e-05, | |
| "loss": 0.0556, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 34.13, | |
| "eval_cer": 0.0223037755267862, | |
| "eval_loss": 0.18723702430725098, | |
| "eval_runtime": 71.3886, | |
| "eval_samples_per_second": 21.012, | |
| "eval_steps_per_second": 2.633, | |
| "eval_wer": 0.07671298899510001, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 35.2, | |
| "learning_rate": 1.9638012958963286e-05, | |
| "loss": 0.0534, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 36.27, | |
| "learning_rate": 1.9314038876889848e-05, | |
| "loss": 0.0534, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 36.27, | |
| "eval_cer": 0.022141270787064623, | |
| "eval_loss": 0.1893489509820938, | |
| "eval_runtime": 71.3518, | |
| "eval_samples_per_second": 21.023, | |
| "eval_steps_per_second": 2.635, | |
| "eval_wer": 0.07510643425174712, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 37.33, | |
| "learning_rate": 1.8990064794816417e-05, | |
| "loss": 0.0528, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 38.4, | |
| "learning_rate": 1.8666090712742982e-05, | |
| "loss": 0.0523, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 38.4, | |
| "eval_cer": 0.021789177184334544, | |
| "eval_loss": 0.19247621297836304, | |
| "eval_runtime": 71.3463, | |
| "eval_samples_per_second": 21.024, | |
| "eval_steps_per_second": 2.635, | |
| "eval_wer": 0.07293758534822074, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 39.47, | |
| "learning_rate": 1.8342116630669547e-05, | |
| "loss": 0.0507, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 40.53, | |
| "learning_rate": 1.801879049676026e-05, | |
| "loss": 0.0494, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 40.53, | |
| "eval_cer": 0.0220735604788473, | |
| "eval_loss": 0.1956777721643448, | |
| "eval_runtime": 71.5278, | |
| "eval_samples_per_second": 20.971, | |
| "eval_steps_per_second": 2.628, | |
| "eval_wer": 0.07454414009157362, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 41.6, | |
| "learning_rate": 1.7694816414686825e-05, | |
| "loss": 0.0504, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 42.67, | |
| "learning_rate": 1.7370842332613394e-05, | |
| "loss": 0.0475, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 42.67, | |
| "eval_cer": 0.021748550999404148, | |
| "eval_loss": 0.19612860679626465, | |
| "eval_runtime": 71.56, | |
| "eval_samples_per_second": 20.961, | |
| "eval_steps_per_second": 2.627, | |
| "eval_wer": 0.07398184593140011, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 43.73, | |
| "learning_rate": 1.7046868250539956e-05, | |
| "loss": 0.0489, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 44.8, | |
| "learning_rate": 1.6722894168466524e-05, | |
| "loss": 0.048, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 44.8, | |
| "eval_cer": 0.021355831211743677, | |
| "eval_loss": 0.19570261240005493, | |
| "eval_runtime": 71.295, | |
| "eval_samples_per_second": 21.039, | |
| "eval_steps_per_second": 2.637, | |
| "eval_wer": 0.0714113583420355, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 45.87, | |
| "learning_rate": 1.639892008639309e-05, | |
| "loss": 0.0453, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 46.93, | |
| "learning_rate": 1.6074946004319655e-05, | |
| "loss": 0.0459, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 46.93, | |
| "eval_cer": 0.021450625643247928, | |
| "eval_loss": 0.19677624106407166, | |
| "eval_runtime": 71.8431, | |
| "eval_samples_per_second": 20.879, | |
| "eval_steps_per_second": 2.617, | |
| "eval_wer": 0.07165234155353843, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 48.0, | |
| "learning_rate": 1.5750971922246223e-05, | |
| "loss": 0.045, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 49.07, | |
| "learning_rate": 1.5426997840172785e-05, | |
| "loss": 0.0435, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 49.07, | |
| "eval_cer": 0.02172146687611722, | |
| "eval_loss": 0.20079585909843445, | |
| "eval_runtime": 74.2582, | |
| "eval_samples_per_second": 20.2, | |
| "eval_steps_per_second": 2.532, | |
| "eval_wer": 0.07173266929070608, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 50.13, | |
| "learning_rate": 1.5103023758099354e-05, | |
| "loss": 0.043, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 51.2, | |
| "learning_rate": 1.4779049676025917e-05, | |
| "loss": 0.0428, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 51.2, | |
| "eval_cer": 0.021233952656952493, | |
| "eval_loss": 0.19910749793052673, | |
| "eval_runtime": 73.3587, | |
| "eval_samples_per_second": 20.447, | |
| "eval_steps_per_second": 2.563, | |
| "eval_wer": 0.06964414812434734, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 52.27, | |
| "learning_rate": 1.4455075593952484e-05, | |
| "loss": 0.0436, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 53.33, | |
| "learning_rate": 1.4131101511879051e-05, | |
| "loss": 0.0418, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 53.33, | |
| "eval_cer": 0.02154542007475218, | |
| "eval_loss": 0.2034166008234024, | |
| "eval_runtime": 75.9701, | |
| "eval_samples_per_second": 19.745, | |
| "eval_steps_per_second": 2.475, | |
| "eval_wer": 0.0714113583420355, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 54.4, | |
| "learning_rate": 1.3807127429805616e-05, | |
| "loss": 0.0415, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 55.47, | |
| "learning_rate": 1.3483153347732181e-05, | |
| "loss": 0.0404, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 55.47, | |
| "eval_cer": 0.02099019554737013, | |
| "eval_loss": 0.20137159526348114, | |
| "eval_runtime": 89.8064, | |
| "eval_samples_per_second": 16.703, | |
| "eval_steps_per_second": 2.093, | |
| "eval_wer": 0.06843923206683268, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 56.53, | |
| "learning_rate": 1.3159827213822896e-05, | |
| "loss": 0.0415, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 57.6, | |
| "learning_rate": 1.2836501079913607e-05, | |
| "loss": 0.0394, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 57.6, | |
| "eval_cer": 0.021003737609013595, | |
| "eval_loss": 0.20499302446842194, | |
| "eval_runtime": 84.8526, | |
| "eval_samples_per_second": 17.678, | |
| "eval_steps_per_second": 2.216, | |
| "eval_wer": 0.0681179211181621, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 58.67, | |
| "learning_rate": 1.2512526997840172e-05, | |
| "loss": 0.0396, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 59.73, | |
| "learning_rate": 1.218855291576674e-05, | |
| "loss": 0.0399, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 59.73, | |
| "eval_cer": 0.021112074102161312, | |
| "eval_loss": 0.2038762867450714, | |
| "eval_runtime": 88.8478, | |
| "eval_samples_per_second": 16.883, | |
| "eval_steps_per_second": 2.116, | |
| "eval_wer": 0.07004578681018556, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 60.8, | |
| "learning_rate": 1.1864578833693305e-05, | |
| "loss": 0.0367, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 61.87, | |
| "learning_rate": 1.1541252699784017e-05, | |
| "loss": 0.0389, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 61.87, | |
| "eval_cer": 0.021355831211743677, | |
| "eval_loss": 0.20906661450862885, | |
| "eval_runtime": 92.2839, | |
| "eval_samples_per_second": 16.254, | |
| "eval_steps_per_second": 2.037, | |
| "eval_wer": 0.0694031649128444, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 62.93, | |
| "learning_rate": 1.1217278617710583e-05, | |
| "loss": 0.0381, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 64.0, | |
| "learning_rate": 1.089330453563715e-05, | |
| "loss": 0.038, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 64.0, | |
| "eval_cer": 0.02097665348572667, | |
| "eval_loss": 0.21002130210399628, | |
| "eval_runtime": 89.2275, | |
| "eval_samples_per_second": 16.811, | |
| "eval_steps_per_second": 2.107, | |
| "eval_wer": 0.07020644228452085, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 65.07, | |
| "learning_rate": 1.0569978401727863e-05, | |
| "loss": 0.0367, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 66.13, | |
| "learning_rate": 1.0246004319654428e-05, | |
| "loss": 0.0361, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 66.13, | |
| "eval_cer": 0.021450625643247928, | |
| "eval_loss": 0.21194988489151, | |
| "eval_runtime": 81.4435, | |
| "eval_samples_per_second": 18.418, | |
| "eval_steps_per_second": 2.308, | |
| "eval_wer": 0.07028677002168848, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 67.2, | |
| "learning_rate": 9.922030237580995e-06, | |
| "loss": 0.0346, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 68.27, | |
| "learning_rate": 9.59805615550756e-06, | |
| "loss": 0.0359, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 68.27, | |
| "eval_cer": 0.02134228915010021, | |
| "eval_loss": 0.21081893146038055, | |
| "eval_runtime": 80.4987, | |
| "eval_samples_per_second": 18.634, | |
| "eval_steps_per_second": 2.335, | |
| "eval_wer": 0.0714113583420355, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 69.33, | |
| "learning_rate": 9.274082073434125e-06, | |
| "loss": 0.0363, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 70.4, | |
| "learning_rate": 8.95010799136069e-06, | |
| "loss": 0.0354, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 70.4, | |
| "eval_cer": 0.02112561616380478, | |
| "eval_loss": 0.2120192050933838, | |
| "eval_runtime": 87.4388, | |
| "eval_samples_per_second": 17.155, | |
| "eval_steps_per_second": 2.15, | |
| "eval_wer": 0.06988513133585027, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 71.47, | |
| "learning_rate": 8.626133909287257e-06, | |
| "loss": 0.0349, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 72.53, | |
| "learning_rate": 8.302159827213824e-06, | |
| "loss": 0.0364, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 72.53, | |
| "eval_cer": 0.02107144791723092, | |
| "eval_loss": 0.21277020871639252, | |
| "eval_runtime": 86.1965, | |
| "eval_samples_per_second": 17.402, | |
| "eval_steps_per_second": 2.181, | |
| "eval_wer": 0.0688408707526709, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 73.6, | |
| "learning_rate": 7.97818574514039e-06, | |
| "loss": 0.0332, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 74.67, | |
| "learning_rate": 7.654211663066955e-06, | |
| "loss": 0.0361, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 74.67, | |
| "eval_cer": 0.02124749471859596, | |
| "eval_loss": 0.2133743017911911, | |
| "eval_runtime": 84.4863, | |
| "eval_samples_per_second": 17.754, | |
| "eval_steps_per_second": 2.225, | |
| "eval_wer": 0.0694031649128444, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 75.73, | |
| "learning_rate": 7.330885529157668e-06, | |
| "loss": 0.0349, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 76.8, | |
| "learning_rate": 7.0069114470842335e-06, | |
| "loss": 0.0332, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 76.8, | |
| "eval_cer": 0.02104436379394399, | |
| "eval_loss": 0.2175784856081009, | |
| "eval_runtime": 74.8124, | |
| "eval_samples_per_second": 20.05, | |
| "eval_steps_per_second": 2.513, | |
| "eval_wer": 0.06980480359868263, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 77.87, | |
| "learning_rate": 6.6829373650108e-06, | |
| "loss": 0.032, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 78.93, | |
| "learning_rate": 6.358963282937365e-06, | |
| "loss": 0.0341, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 78.93, | |
| "eval_cer": 0.020827690807648555, | |
| "eval_loss": 0.21700014173984528, | |
| "eval_runtime": 79.1989, | |
| "eval_samples_per_second": 18.94, | |
| "eval_steps_per_second": 2.374, | |
| "eval_wer": 0.0688408707526709, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 80.0, | |
| "learning_rate": 6.034989200863931e-06, | |
| "loss": 0.0323, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 81.07, | |
| "learning_rate": 5.711015118790497e-06, | |
| "loss": 0.032, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 81.07, | |
| "eval_cer": 0.020908943177509343, | |
| "eval_loss": 0.215650275349617, | |
| "eval_runtime": 74.4056, | |
| "eval_samples_per_second": 20.16, | |
| "eval_steps_per_second": 2.527, | |
| "eval_wer": 0.06859988754116797, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 82.26, | |
| "learning_rate": 5.387041036717062e-06, | |
| "loss": 0.0321, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 83.33, | |
| "learning_rate": 5.063066954643629e-06, | |
| "loss": 0.0318, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 83.33, | |
| "eval_cer": 0.020908943177509343, | |
| "eval_loss": 0.2165554016828537, | |
| "eval_runtime": 69.4748, | |
| "eval_samples_per_second": 21.591, | |
| "eval_steps_per_second": 2.706, | |
| "eval_wer": 0.06851955980400032, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 84.4, | |
| "learning_rate": 4.739092872570194e-06, | |
| "loss": 0.0338, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 85.47, | |
| "learning_rate": 4.415766738660907e-06, | |
| "loss": 0.0325, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 85.47, | |
| "eval_cer": 0.020881859054222414, | |
| "eval_loss": 0.21718913316726685, | |
| "eval_runtime": 70.6439, | |
| "eval_samples_per_second": 21.233, | |
| "eval_steps_per_second": 2.661, | |
| "eval_wer": 0.06868021527833561, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 86.53, | |
| "learning_rate": 4.09244060475162e-06, | |
| "loss": 0.0318, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 87.6, | |
| "learning_rate": 3.7684665226781857e-06, | |
| "loss": 0.0316, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 87.6, | |
| "eval_cer": 0.020800606684361626, | |
| "eval_loss": 0.21813979744911194, | |
| "eval_runtime": 81.1292, | |
| "eval_samples_per_second": 18.489, | |
| "eval_steps_per_second": 2.317, | |
| "eval_wer": 0.06779661016949153, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 88.67, | |
| "learning_rate": 3.4444924406047518e-06, | |
| "loss": 0.0325, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 89.73, | |
| "learning_rate": 3.1205183585313174e-06, | |
| "loss": 0.0302, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 89.73, | |
| "eval_cer": 0.02084123286929202, | |
| "eval_loss": 0.21708616614341736, | |
| "eval_runtime": 81.2548, | |
| "eval_samples_per_second": 18.46, | |
| "eval_steps_per_second": 2.314, | |
| "eval_wer": 0.06787693790665916, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 90.8, | |
| "learning_rate": 2.7971922246220303e-06, | |
| "loss": 0.031, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 91.87, | |
| "learning_rate": 2.473218142548596e-06, | |
| "loss": 0.0318, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 91.87, | |
| "eval_cer": 0.02112561616380478, | |
| "eval_loss": 0.21793220937252045, | |
| "eval_runtime": 81.745, | |
| "eval_samples_per_second": 18.35, | |
| "eval_steps_per_second": 2.3, | |
| "eval_wer": 0.07020644228452085, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 92.93, | |
| "learning_rate": 2.149244060475162e-06, | |
| "loss": 0.0309, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 94.0, | |
| "learning_rate": 1.8252699784017279e-06, | |
| "loss": 0.0314, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 94.0, | |
| "eval_cer": 0.020800606684361626, | |
| "eval_loss": 0.21862204372882843, | |
| "eval_runtime": 82.362, | |
| "eval_samples_per_second": 18.212, | |
| "eval_steps_per_second": 2.283, | |
| "eval_wer": 0.06900152622700619, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 95.07, | |
| "learning_rate": 1.5012958963282937e-06, | |
| "loss": 0.0299, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 96.13, | |
| "learning_rate": 1.1779697624190064e-06, | |
| "loss": 0.0309, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 96.13, | |
| "eval_cer": 0.02097665348572667, | |
| "eval_loss": 0.219330295920372, | |
| "eval_runtime": 80.0094, | |
| "eval_samples_per_second": 18.748, | |
| "eval_steps_per_second": 2.35, | |
| "eval_wer": 0.06964414812434734, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 97.2, | |
| "learning_rate": 8.539956803455724e-07, | |
| "loss": 0.0304, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 98.27, | |
| "learning_rate": 5.300215982721382e-07, | |
| "loss": 0.031, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 98.27, | |
| "eval_cer": 0.0207735225610747, | |
| "eval_loss": 0.219051793217659, | |
| "eval_runtime": 82.8771, | |
| "eval_samples_per_second": 18.099, | |
| "eval_steps_per_second": 2.268, | |
| "eval_wer": 0.06859988754116797, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 99.33, | |
| "learning_rate": 2.0604751619870412e-07, | |
| "loss": 0.0309, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 99.97, | |
| "step": 46800, | |
| "total_flos": 2.6879054505409987e+20, | |
| "train_loss": 0.005243639966361543, | |
| "train_runtime": 17869.2377, | |
| "train_samples_per_second": 83.943, | |
| "train_steps_per_second": 2.619 | |
| } | |
| ], | |
| "max_steps": 46800, | |
| "num_train_epochs": 100, | |
| "total_flos": 2.6879054505409987e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |