| { | |
| "best_metric": 49.03147699757869, | |
| "best_model_checkpoint": "./checkpoint-400", | |
| "epoch": 11.894117647058824, | |
| "global_step": 500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 4.3134, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 6e-06, | |
| "loss": 2.9329, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 2.0584, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 9.829787234042554e-06, | |
| "loss": 1.566, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 9.617021276595745e-06, | |
| "loss": 1.3777, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "learning_rate": 9.404255319148937e-06, | |
| "loss": 1.1469, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "learning_rate": 9.191489361702128e-06, | |
| "loss": 1.0638, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 8.97872340425532e-06, | |
| "loss": 0.9974, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 8.765957446808512e-06, | |
| "loss": 0.9615, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "learning_rate": 8.553191489361703e-06, | |
| "loss": 0.8262, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "eval_cer": 149.33157314260887, | |
| "eval_loss": 0.8188337683677673, | |
| "eval_runtime": 459.9395, | |
| "eval_samples_per_second": 1.113, | |
| "eval_steps_per_second": 0.139, | |
| "eval_wer": 146.85230024213075, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 8.340425531914894e-06, | |
| "loss": 0.7986, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "learning_rate": 8.127659574468085e-06, | |
| "loss": 0.7707, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "learning_rate": 7.914893617021278e-06, | |
| "loss": 0.7623, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "learning_rate": 7.702127659574469e-06, | |
| "loss": 0.6451, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "learning_rate": 7.48936170212766e-06, | |
| "loss": 0.6205, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "learning_rate": 7.2765957446808524e-06, | |
| "loss": 0.6097, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "learning_rate": 7.0638297872340434e-06, | |
| "loss": 0.6217, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 4.28, | |
| "learning_rate": 6.8510638297872344e-06, | |
| "loss": 0.4584, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 4.52, | |
| "learning_rate": 6.6382978723404254e-06, | |
| "loss": 0.4929, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 4.75, | |
| "learning_rate": 6.425531914893618e-06, | |
| "loss": 0.4843, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 4.75, | |
| "eval_cer": 22.099862972494236, | |
| "eval_loss": 0.6699215173721313, | |
| "eval_runtime": 1062.3991, | |
| "eval_samples_per_second": 0.482, | |
| "eval_steps_per_second": 0.06, | |
| "eval_wer": 52.66343825665859, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 4.99, | |
| "learning_rate": 6.212765957446809e-06, | |
| "loss": 0.4678, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 5.24, | |
| "learning_rate": 6e-06, | |
| "loss": 0.4014, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 5.47, | |
| "learning_rate": 5.787234042553191e-06, | |
| "loss": 0.3926, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 5.71, | |
| "learning_rate": 5.574468085106384e-06, | |
| "loss": 0.3806, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 5.94, | |
| "learning_rate": 5.361702127659575e-06, | |
| "loss": 0.3855, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 6.19, | |
| "learning_rate": 5.148936170212766e-06, | |
| "loss": 0.3478, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 6.42, | |
| "learning_rate": 4.936170212765958e-06, | |
| "loss": 0.3078, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 6.66, | |
| "learning_rate": 4.7234042553191496e-06, | |
| "loss": 0.3147, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 6.89, | |
| "learning_rate": 4.5106382978723406e-06, | |
| "loss": 0.3213, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 7.14, | |
| "learning_rate": 4.297872340425532e-06, | |
| "loss": 0.287, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 7.14, | |
| "eval_cer": 20.677784833394604, | |
| "eval_loss": 0.6913720369338989, | |
| "eval_runtime": 1003.603, | |
| "eval_samples_per_second": 0.51, | |
| "eval_steps_per_second": 0.064, | |
| "eval_wer": 50.13619854721549, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 7.38, | |
| "learning_rate": 4.085106382978723e-06, | |
| "loss": 0.2602, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 7.61, | |
| "learning_rate": 3.872340425531915e-06, | |
| "loss": 0.2612, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 7.85, | |
| "learning_rate": 3.6595744680851063e-06, | |
| "loss": 0.2618, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 8.09, | |
| "learning_rate": 3.446808510638298e-06, | |
| "loss": 0.2494, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "learning_rate": 3.23404255319149e-06, | |
| "loss": 0.2123, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 8.56, | |
| "learning_rate": 3.021276595744681e-06, | |
| "loss": 0.2173, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "learning_rate": 2.808510638297873e-06, | |
| "loss": 0.2093, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 9.05, | |
| "learning_rate": 2.595744680851064e-06, | |
| "loss": 0.2209, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 9.28, | |
| "learning_rate": 2.3829787234042557e-06, | |
| "loss": 0.1742, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 9.52, | |
| "learning_rate": 2.170212765957447e-06, | |
| "loss": 0.1828, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 9.52, | |
| "eval_cer": 20.460546104742487, | |
| "eval_loss": 0.7369570136070251, | |
| "eval_runtime": 1004.0546, | |
| "eval_samples_per_second": 0.51, | |
| "eval_steps_per_second": 0.064, | |
| "eval_wer": 49.03147699757869, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 9.75, | |
| "learning_rate": 1.9574468085106385e-06, | |
| "loss": 0.1902, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 9.99, | |
| "learning_rate": 1.74468085106383e-06, | |
| "loss": 0.1809, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 10.24, | |
| "learning_rate": 1.5319148936170214e-06, | |
| "loss": 0.1705, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 10.47, | |
| "learning_rate": 1.3191489361702128e-06, | |
| "loss": 0.1593, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 10.71, | |
| "learning_rate": 1.1063829787234042e-06, | |
| "loss": 0.1596, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 10.94, | |
| "learning_rate": 8.936170212765959e-07, | |
| "loss": 0.161, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 11.19, | |
| "learning_rate": 6.808510638297873e-07, | |
| "loss": 0.1594, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 11.42, | |
| "learning_rate": 4.6808510638297873e-07, | |
| "loss": 0.1436, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 11.66, | |
| "learning_rate": 2.553191489361702e-07, | |
| "loss": 0.1517, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 11.89, | |
| "learning_rate": 4.2553191489361707e-08, | |
| "loss": 0.1493, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 11.89, | |
| "eval_cer": 20.8064570034424, | |
| "eval_loss": 0.7532366514205933, | |
| "eval_runtime": 925.6199, | |
| "eval_samples_per_second": 0.553, | |
| "eval_steps_per_second": 0.069, | |
| "eval_wer": 49.583837772397096, | |
| "step": 500 | |
| } | |
| ], | |
| "max_steps": 500, | |
| "num_train_epochs": 12, | |
| "total_flos": 9.30774491652096e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |