| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 99.9584, | |
| "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": 2.9880000000000002e-05, | |
| "loss": 7.2457, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "learning_rate": 2.967732181425486e-05, | |
| "loss": 2.3129, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "eval_cer": 0.05796002383402849, | |
| "eval_loss": 0.5042470097541809, | |
| "eval_runtime": 80.7316, | |
| "eval_samples_per_second": 18.58, | |
| "eval_steps_per_second": 2.329, | |
| "eval_wer": 0.270302835569122, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 2.9353347732181423e-05, | |
| "loss": 0.389, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "learning_rate": 2.9029373650107992e-05, | |
| "loss": 0.2251, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "eval_cer": 0.02945398407453551, | |
| "eval_loss": 0.17824731767177582, | |
| "eval_runtime": 80.3146, | |
| "eval_samples_per_second": 18.677, | |
| "eval_steps_per_second": 2.341, | |
| "eval_wer": 0.11984898385412483, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 5.33, | |
| "learning_rate": 2.870539956803456e-05, | |
| "loss": 0.1727, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "learning_rate": 2.8381425485961122e-05, | |
| "loss": 0.1462, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "eval_cer": 0.02651535669790369, | |
| "eval_loss": 0.1635013222694397, | |
| "eval_runtime": 72.2551, | |
| "eval_samples_per_second": 20.76, | |
| "eval_steps_per_second": 2.602, | |
| "eval_wer": 0.10185557072857257, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 7.47, | |
| "learning_rate": 2.805745140388769e-05, | |
| "loss": 0.128, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 8.53, | |
| "learning_rate": 2.7733477321814256e-05, | |
| "loss": 0.1162, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 8.53, | |
| "eval_cer": 0.02478197280754022, | |
| "eval_loss": 0.1619153469800949, | |
| "eval_runtime": 81.2175, | |
| "eval_samples_per_second": 18.469, | |
| "eval_steps_per_second": 2.315, | |
| "eval_wer": 0.09309984737729939, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "learning_rate": 2.740950323974082e-05, | |
| "loss": 0.1095, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 10.67, | |
| "learning_rate": 2.7085529157667386e-05, | |
| "loss": 0.0988, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 10.67, | |
| "eval_cer": 0.024890309300687936, | |
| "eval_loss": 0.1653970628976822, | |
| "eval_runtime": 72.7324, | |
| "eval_samples_per_second": 20.624, | |
| "eval_steps_per_second": 2.585, | |
| "eval_wer": 0.09398345248614347, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 11.73, | |
| "learning_rate": 2.676155507559395e-05, | |
| "loss": 0.0949, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "learning_rate": 2.643758099352052e-05, | |
| "loss": 0.0904, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "eval_cer": 0.024186122095227778, | |
| "eval_loss": 0.17023487389087677, | |
| "eval_runtime": 80.5666, | |
| "eval_samples_per_second": 18.618, | |
| "eval_steps_per_second": 2.333, | |
| "eval_wer": 0.08450477950036148, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 13.87, | |
| "learning_rate": 2.6113606911447086e-05, | |
| "loss": 0.0859, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 14.93, | |
| "learning_rate": 2.578963282937365e-05, | |
| "loss": 0.0813, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 14.93, | |
| "eval_cer": 0.02392882292400195, | |
| "eval_loss": 0.16578005254268646, | |
| "eval_runtime": 72.3723, | |
| "eval_samples_per_second": 20.726, | |
| "eval_steps_per_second": 2.598, | |
| "eval_wer": 0.08458510723752911, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 16.02, | |
| "learning_rate": 2.5466306695464364e-05, | |
| "loss": 0.0763, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 17.09, | |
| "learning_rate": 2.514233261339093e-05, | |
| "loss": 0.074, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 17.09, | |
| "eval_cer": 0.02395590704728888, | |
| "eval_loss": 0.1762518584728241, | |
| "eval_runtime": 72.6367, | |
| "eval_samples_per_second": 20.651, | |
| "eval_steps_per_second": 2.588, | |
| "eval_wer": 0.07928347658446462, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 18.16, | |
| "learning_rate": 2.4818358531317497e-05, | |
| "loss": 0.0714, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 19.22, | |
| "learning_rate": 2.4495032397408207e-05, | |
| "loss": 0.0692, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 19.22, | |
| "eval_cer": 0.0242538324034451, | |
| "eval_loss": 0.17680838704109192, | |
| "eval_runtime": 68.8274, | |
| "eval_samples_per_second": 21.794, | |
| "eval_steps_per_second": 2.731, | |
| "eval_wer": 0.0834605189171821, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 20.3, | |
| "learning_rate": 2.4171058315334776e-05, | |
| "loss": 0.0658, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 21.36, | |
| "learning_rate": 2.3847084233261337e-05, | |
| "loss": 0.0652, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 21.36, | |
| "eval_cer": 0.02365798169113266, | |
| "eval_loss": 0.18117107450962067, | |
| "eval_runtime": 72.0293, | |
| "eval_samples_per_second": 20.825, | |
| "eval_steps_per_second": 2.61, | |
| "eval_wer": 0.07968511527030284, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 22.43, | |
| "learning_rate": 2.3523110151187906e-05, | |
| "loss": 0.0625, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 23.5, | |
| "learning_rate": 2.319913606911447e-05, | |
| "loss": 0.0593, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 23.5, | |
| "eval_cer": 0.022127728725421156, | |
| "eval_loss": 0.18102437257766724, | |
| "eval_runtime": 69.723, | |
| "eval_samples_per_second": 21.514, | |
| "eval_steps_per_second": 2.696, | |
| "eval_wer": 0.07502610651457949, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 24.56, | |
| "learning_rate": 2.2875809935205184e-05, | |
| "loss": 0.0604, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 25.63, | |
| "learning_rate": 2.255183585313175e-05, | |
| "loss": 0.0547, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 25.63, | |
| "eval_cer": 0.023265261903472185, | |
| "eval_loss": 0.18346643447875977, | |
| "eval_runtime": 70.9539, | |
| "eval_samples_per_second": 21.14, | |
| "eval_steps_per_second": 2.65, | |
| "eval_wer": 0.07944413205879991, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 26.7, | |
| "learning_rate": 2.2227861771058315e-05, | |
| "loss": 0.0565, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 27.76, | |
| "learning_rate": 2.1903887688984883e-05, | |
| "loss": 0.0514, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 27.76, | |
| "eval_cer": 0.022385027896646984, | |
| "eval_loss": 0.18281148374080658, | |
| "eval_runtime": 70.1644, | |
| "eval_samples_per_second": 21.378, | |
| "eval_steps_per_second": 2.679, | |
| "eval_wer": 0.07607036709775886, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 28.83, | |
| "learning_rate": 2.157991360691145e-05, | |
| "loss": 0.053, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 29.9, | |
| "learning_rate": 2.1255939524838014e-05, | |
| "loss": 0.0488, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 29.9, | |
| "eval_cer": 0.02242565408157738, | |
| "eval_loss": 0.18435174226760864, | |
| "eval_runtime": 69.6075, | |
| "eval_samples_per_second": 21.549, | |
| "eval_steps_per_second": 2.701, | |
| "eval_wer": 0.07663266125793236, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 30.96, | |
| "learning_rate": 2.093196544276458e-05, | |
| "loss": 0.0514, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 32.03, | |
| "learning_rate": 2.0607991360691144e-05, | |
| "loss": 0.0478, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 32.03, | |
| "eval_cer": 0.022588158821298953, | |
| "eval_loss": 0.1909506469964981, | |
| "eval_runtime": 72.4928, | |
| "eval_samples_per_second": 20.692, | |
| "eval_steps_per_second": 2.593, | |
| "eval_wer": 0.0768736444694353, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 33.1, | |
| "learning_rate": 2.0284017278617713e-05, | |
| "loss": 0.0489, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 34.16, | |
| "learning_rate": 1.9960043196544274e-05, | |
| "loss": 0.0459, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 34.16, | |
| "eval_cer": 0.02386111261578463, | |
| "eval_loss": 0.19651177525520325, | |
| "eval_runtime": 69.572, | |
| "eval_samples_per_second": 21.56, | |
| "eval_steps_per_second": 2.702, | |
| "eval_wer": 0.08313920796851153, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 35.23, | |
| "learning_rate": 1.9636069114470843e-05, | |
| "loss": 0.0481, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 36.3, | |
| "learning_rate": 1.9312095032397408e-05, | |
| "loss": 0.0429, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 36.3, | |
| "eval_cer": 0.021951681924056117, | |
| "eval_loss": 0.20002886652946472, | |
| "eval_runtime": 69.2113, | |
| "eval_samples_per_second": 21.673, | |
| "eval_steps_per_second": 2.716, | |
| "eval_wer": 0.07599003936059122, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 37.36, | |
| "learning_rate": 1.898876889848812e-05, | |
| "loss": 0.0461, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 38.43, | |
| "learning_rate": 1.8664794816414686e-05, | |
| "loss": 0.0443, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 38.43, | |
| "eval_cer": 0.02277774768430746, | |
| "eval_loss": 0.20385122299194336, | |
| "eval_runtime": 72.6401, | |
| "eval_samples_per_second": 20.65, | |
| "eval_steps_per_second": 2.588, | |
| "eval_wer": 0.0774359386296088, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 39.5, | |
| "learning_rate": 1.834082073434125e-05, | |
| "loss": 0.0411, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 40.56, | |
| "learning_rate": 1.8017494600431965e-05, | |
| "loss": 0.0398, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 40.56, | |
| "eval_cer": 0.021856887492551866, | |
| "eval_loss": 0.19806204736232758, | |
| "eval_runtime": 72.4106, | |
| "eval_samples_per_second": 20.715, | |
| "eval_steps_per_second": 2.596, | |
| "eval_wer": 0.07550807293758535, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 41.63, | |
| "learning_rate": 1.7693520518358533e-05, | |
| "loss": 0.0431, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 42.7, | |
| "learning_rate": 1.7369546436285095e-05, | |
| "loss": 0.0408, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 42.7, | |
| "eval_cer": 0.02387465467742809, | |
| "eval_loss": 0.2053176462650299, | |
| "eval_runtime": 69.6266, | |
| "eval_samples_per_second": 21.543, | |
| "eval_steps_per_second": 2.7, | |
| "eval_wer": 0.0775965941039441, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 43.76, | |
| "learning_rate": 1.7045572354211664e-05, | |
| "loss": 0.0399, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 44.83, | |
| "learning_rate": 1.6721598272138232e-05, | |
| "loss": 0.0406, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 44.83, | |
| "eval_cer": 0.022060018417203835, | |
| "eval_loss": 0.2049773633480072, | |
| "eval_runtime": 70.2646, | |
| "eval_samples_per_second": 21.348, | |
| "eval_steps_per_second": 2.676, | |
| "eval_wer": 0.07398184593140011, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 45.9, | |
| "learning_rate": 1.6397624190064794e-05, | |
| "loss": 0.0381, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 46.96, | |
| "learning_rate": 1.6073650107991363e-05, | |
| "loss": 0.0383, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 46.96, | |
| "eval_cer": 0.022357943773360055, | |
| "eval_loss": 0.21280354261398315, | |
| "eval_runtime": 72.639, | |
| "eval_samples_per_second": 20.65, | |
| "eval_steps_per_second": 2.588, | |
| "eval_wer": 0.07325889629689132, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 48.03, | |
| "learning_rate": 1.5749676025917928e-05, | |
| "loss": 0.0365, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 49.1, | |
| "learning_rate": 1.5425701943844493e-05, | |
| "loss": 0.0379, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 49.1, | |
| "eval_cer": 0.021978766047343046, | |
| "eval_loss": 0.21096549928188324, | |
| "eval_runtime": 73.1028, | |
| "eval_samples_per_second": 20.519, | |
| "eval_steps_per_second": 2.572, | |
| "eval_wer": 0.07309824082255603, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 50.16, | |
| "learning_rate": 1.5102375809935208e-05, | |
| "loss": 0.0361, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 51.23, | |
| "learning_rate": 1.4778401727861771e-05, | |
| "loss": 0.0369, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 51.23, | |
| "eval_cer": 0.021951681924056117, | |
| "eval_loss": 0.2144922912120819, | |
| "eval_runtime": 69.8484, | |
| "eval_samples_per_second": 21.475, | |
| "eval_steps_per_second": 2.692, | |
| "eval_wer": 0.07446381235440598, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 52.3, | |
| "learning_rate": 1.4455075593952484e-05, | |
| "loss": 0.0334, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 53.36, | |
| "learning_rate": 1.4131101511879051e-05, | |
| "loss": 0.0341, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 53.36, | |
| "eval_cer": 0.022208981095281945, | |
| "eval_loss": 0.2145662158727646, | |
| "eval_runtime": 69.6873, | |
| "eval_samples_per_second": 21.525, | |
| "eval_steps_per_second": 2.698, | |
| "eval_wer": 0.07245561892521488, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 54.43, | |
| "learning_rate": 1.3807127429805616e-05, | |
| "loss": 0.0356, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 55.5, | |
| "learning_rate": 1.3483801295896327e-05, | |
| "loss": 0.0322, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 55.5, | |
| "eval_cer": 0.021626672444612968, | |
| "eval_loss": 0.21297892928123474, | |
| "eval_runtime": 73.9282, | |
| "eval_samples_per_second": 20.29, | |
| "eval_steps_per_second": 2.543, | |
| "eval_wer": 0.07100971965619729, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 56.56, | |
| "learning_rate": 1.3159827213822896e-05, | |
| "loss": 0.0327, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 57.63, | |
| "learning_rate": 1.2835853131749461e-05, | |
| "loss": 0.0316, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 57.63, | |
| "eval_cer": 0.02224960728021234, | |
| "eval_loss": 0.2133886069059372, | |
| "eval_runtime": 69.9039, | |
| "eval_samples_per_second": 21.458, | |
| "eval_steps_per_second": 2.689, | |
| "eval_wer": 0.0715720138163708, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 58.7, | |
| "learning_rate": 1.2511879049676026e-05, | |
| "loss": 0.0319, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 59.76, | |
| "learning_rate": 1.218855291576674e-05, | |
| "loss": 0.0324, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 59.76, | |
| "eval_cer": 0.022236065218568874, | |
| "eval_loss": 0.21722277998924255, | |
| "eval_runtime": 69.753, | |
| "eval_samples_per_second": 21.504, | |
| "eval_steps_per_second": 2.695, | |
| "eval_wer": 0.07309824082255603, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 60.83, | |
| "learning_rate": 1.1864578833693305e-05, | |
| "loss": 0.0296, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 61.9, | |
| "learning_rate": 1.1540604751619871e-05, | |
| "loss": 0.0315, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 61.9, | |
| "eval_cer": 0.02280483180759439, | |
| "eval_loss": 0.22069784998893738, | |
| "eval_runtime": 79.4911, | |
| "eval_samples_per_second": 18.87, | |
| "eval_steps_per_second": 2.365, | |
| "eval_wer": 0.07454414009157362, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 62.96, | |
| "learning_rate": 1.1216630669546437e-05, | |
| "loss": 0.0315, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 64.03, | |
| "learning_rate": 1.0892656587473002e-05, | |
| "loss": 0.0294, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 64.03, | |
| "eval_cer": 0.021843345430908403, | |
| "eval_loss": 0.2183454930782318, | |
| "eval_runtime": 77.818, | |
| "eval_samples_per_second": 19.276, | |
| "eval_steps_per_second": 2.416, | |
| "eval_wer": 0.07165234155353843, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 65.1, | |
| "learning_rate": 1.0569330453563715e-05, | |
| "loss": 0.0301, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 66.16, | |
| "learning_rate": 1.024535637149028e-05, | |
| "loss": 0.028, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 66.16, | |
| "eval_cer": 0.021355831211743677, | |
| "eval_loss": 0.2184668779373169, | |
| "eval_runtime": 80.0286, | |
| "eval_samples_per_second": 18.743, | |
| "eval_steps_per_second": 2.349, | |
| "eval_wer": 0.06956382038717969, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 67.23, | |
| "learning_rate": 9.921382289416847e-06, | |
| "loss": 0.0283, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 68.3, | |
| "learning_rate": 9.597408207343414e-06, | |
| "loss": 0.0263, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 68.3, | |
| "eval_cer": 0.021504793889821787, | |
| "eval_loss": 0.21668484807014465, | |
| "eval_runtime": 70.5125, | |
| "eval_samples_per_second": 21.273, | |
| "eval_steps_per_second": 2.666, | |
| "eval_wer": 0.06964414812434734, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 69.36, | |
| "learning_rate": 9.273434125269979e-06, | |
| "loss": 0.0286, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 70.43, | |
| "learning_rate": 8.949460043196544e-06, | |
| "loss": 0.0299, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 70.43, | |
| "eval_cer": 0.021707924814473756, | |
| "eval_loss": 0.22012607753276825, | |
| "eval_runtime": 72.8041, | |
| "eval_samples_per_second": 20.603, | |
| "eval_steps_per_second": 2.582, | |
| "eval_wer": 0.07092939191902964, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 71.5, | |
| "learning_rate": 8.62548596112311e-06, | |
| "loss": 0.0267, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 72.56, | |
| "learning_rate": 8.301511879049676e-06, | |
| "loss": 0.0273, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 72.56, | |
| "eval_cer": 0.02219543903363848, | |
| "eval_loss": 0.21641910076141357, | |
| "eval_runtime": 72.2901, | |
| "eval_samples_per_second": 20.75, | |
| "eval_steps_per_second": 2.601, | |
| "eval_wer": 0.07237529118804724, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 73.63, | |
| "learning_rate": 7.977537796976242e-06, | |
| "loss": 0.0267, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 74.7, | |
| "learning_rate": 7.653563714902809e-06, | |
| "loss": 0.0269, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 74.7, | |
| "eval_cer": 0.021951681924056117, | |
| "eval_loss": 0.22398144006729126, | |
| "eval_runtime": 69.5423, | |
| "eval_samples_per_second": 21.57, | |
| "eval_steps_per_second": 2.703, | |
| "eval_wer": 0.06932283717567676, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 75.85, | |
| "learning_rate": 7.329589632829374e-06, | |
| "loss": 0.0265, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 76.92, | |
| "learning_rate": 7.00561555075594e-06, | |
| "loss": 0.0264, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 76.92, | |
| "eval_cer": 0.021789177184334544, | |
| "eval_loss": 0.22204121947288513, | |
| "eval_runtime": 71.976, | |
| "eval_samples_per_second": 20.84, | |
| "eval_steps_per_second": 2.612, | |
| "eval_wer": 0.07036709775885613, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 77.99, | |
| "learning_rate": 6.682289416846653e-06, | |
| "loss": 0.0257, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 79.05, | |
| "learning_rate": 6.358963282937365e-06, | |
| "loss": 0.0257, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 79.05, | |
| "eval_cer": 0.02172146687611722, | |
| "eval_loss": 0.22285164892673492, | |
| "eval_runtime": 68.7898, | |
| "eval_samples_per_second": 21.806, | |
| "eval_steps_per_second": 2.733, | |
| "eval_wer": 0.0688408707526709, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 80.12, | |
| "learning_rate": 6.034989200863931e-06, | |
| "loss": 0.0257, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 81.19, | |
| "learning_rate": 5.711015118790497e-06, | |
| "loss": 0.0251, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 81.19, | |
| "eval_cer": 0.02154542007475218, | |
| "eval_loss": 0.22632832825183868, | |
| "eval_runtime": 71.6223, | |
| "eval_samples_per_second": 20.943, | |
| "eval_steps_per_second": 2.625, | |
| "eval_wer": 0.0694031649128444, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 82.25, | |
| "learning_rate": 5.387041036717062e-06, | |
| "loss": 0.0235, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 83.32, | |
| "learning_rate": 5.063066954643629e-06, | |
| "loss": 0.0245, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 83.32, | |
| "eval_cer": 0.02101727967065706, | |
| "eval_loss": 0.22526498138904572, | |
| "eval_runtime": 70.7229, | |
| "eval_samples_per_second": 21.21, | |
| "eval_steps_per_second": 2.658, | |
| "eval_wer": 0.06731464374648566, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 84.39, | |
| "learning_rate": 4.739740820734342e-06, | |
| "loss": 0.0252, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 85.45, | |
| "learning_rate": 4.415766738660907e-06, | |
| "loss": 0.0243, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 85.45, | |
| "eval_cer": 0.02151833595146525, | |
| "eval_loss": 0.2263830602169037, | |
| "eval_runtime": 69.4576, | |
| "eval_samples_per_second": 21.596, | |
| "eval_steps_per_second": 2.707, | |
| "eval_wer": 0.06916218170134147, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 86.52, | |
| "learning_rate": 4.09244060475162e-06, | |
| "loss": 0.0228, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 87.59, | |
| "learning_rate": 3.7684665226781857e-06, | |
| "loss": 0.0236, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 87.59, | |
| "eval_cer": 0.021667298629543363, | |
| "eval_loss": 0.22610026597976685, | |
| "eval_runtime": 69.0446, | |
| "eval_samples_per_second": 21.725, | |
| "eval_steps_per_second": 2.723, | |
| "eval_wer": 0.06892119848983855, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 88.65, | |
| "learning_rate": 3.4444924406047518e-06, | |
| "loss": 0.0244, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 89.72, | |
| "learning_rate": 3.1205183585313174e-06, | |
| "loss": 0.0225, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 89.72, | |
| "eval_cer": 0.021179784410378637, | |
| "eval_loss": 0.22654421627521515, | |
| "eval_runtime": 72.2716, | |
| "eval_samples_per_second": 20.755, | |
| "eval_steps_per_second": 2.601, | |
| "eval_wer": 0.06803759338099445, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 90.79, | |
| "learning_rate": 2.7965442764578835e-06, | |
| "loss": 0.0216, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 91.85, | |
| "learning_rate": 2.472570194384449e-06, | |
| "loss": 0.023, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 91.85, | |
| "eval_cer": 0.021030821732300524, | |
| "eval_loss": 0.22652284801006317, | |
| "eval_runtime": 72.2083, | |
| "eval_samples_per_second": 20.773, | |
| "eval_steps_per_second": 2.604, | |
| "eval_wer": 0.06739497148365331, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 92.92, | |
| "learning_rate": 2.149244060475162e-06, | |
| "loss": 0.0228, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 93.99, | |
| "learning_rate": 1.825917926565875e-06, | |
| "loss": 0.0217, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 93.99, | |
| "eval_cer": 0.020936027300796273, | |
| "eval_loss": 0.22653113305568695, | |
| "eval_runtime": 69.3882, | |
| "eval_samples_per_second": 21.618, | |
| "eval_steps_per_second": 2.709, | |
| "eval_wer": 0.06771628243232389, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 95.05, | |
| "learning_rate": 1.5019438444924408e-06, | |
| "loss": 0.0203, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 96.12, | |
| "learning_rate": 1.1779697624190064e-06, | |
| "loss": 0.022, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 96.12, | |
| "eval_cer": 0.021057905855587453, | |
| "eval_loss": 0.22544603049755096, | |
| "eval_runtime": 72.3315, | |
| "eval_samples_per_second": 20.738, | |
| "eval_steps_per_second": 2.599, | |
| "eval_wer": 0.06851955980400032, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 97.19, | |
| "learning_rate": 8.539956803455724e-07, | |
| "loss": 0.0215, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 98.25, | |
| "learning_rate": 5.300215982721382e-07, | |
| "loss": 0.0219, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 98.25, | |
| "eval_cer": 0.020759980499431233, | |
| "eval_loss": 0.22618666291236877, | |
| "eval_runtime": 73.3776, | |
| "eval_samples_per_second": 20.442, | |
| "eval_steps_per_second": 2.562, | |
| "eval_wer": 0.06715398827215037, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 99.32, | |
| "learning_rate": 2.0604751619870412e-07, | |
| "loss": 0.0204, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 99.96, | |
| "step": 46800, | |
| "total_flos": 2.6878408769720543e+20, | |
| "train_loss": 0.005932311532843826, | |
| "train_runtime": 23991.1474, | |
| "train_samples_per_second": 62.523, | |
| "train_steps_per_second": 1.951 | |
| } | |
| ], | |
| "max_steps": 46800, | |
| "num_train_epochs": 100, | |
| "total_flos": 2.6878408769720543e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |