| { | |
| "best_metric": 0.18266724050045013, | |
| "best_model_checkpoint": "./checkpoint-3200", | |
| "epoch": 56.14035087719298, | |
| "global_step": 3200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 3.51, | |
| "learning_rate": 4.980500000000001e-05, | |
| "loss": 4.0311, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "eval_cer": 0.23737986848760748, | |
| "eval_loss": 0.794314444065094, | |
| "eval_runtime": 15.7, | |
| "eval_samples_per_second": 21.338, | |
| "eval_steps_per_second": 0.382, | |
| "eval_wer": 0.8981264637002342, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 7.02, | |
| "learning_rate": 8e-05, | |
| "loss": 1.4388, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 7.02, | |
| "eval_cer": 0.06575619625695499, | |
| "eval_loss": 0.2545942962169647, | |
| "eval_runtime": 14.891, | |
| "eval_samples_per_second": 22.497, | |
| "eval_steps_per_second": 0.403, | |
| "eval_wer": 0.38212334113973456, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 10.53, | |
| "learning_rate": 8e-05, | |
| "loss": 1.0949, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 10.53, | |
| "eval_cer": 0.057309054122407686, | |
| "eval_loss": 0.22006691992282867, | |
| "eval_runtime": 15.2331, | |
| "eval_samples_per_second": 21.992, | |
| "eval_steps_per_second": 0.394, | |
| "eval_wer": 0.32162373145979706, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 14.04, | |
| "learning_rate": 8e-05, | |
| "loss": 1.0279, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 14.04, | |
| "eval_cer": 0.0582701062215478, | |
| "eval_loss": 0.22504645586013794, | |
| "eval_runtime": 15.1767, | |
| "eval_samples_per_second": 22.073, | |
| "eval_steps_per_second": 0.395, | |
| "eval_wer": 0.32708821233411395, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 17.54, | |
| "learning_rate": 8e-05, | |
| "loss": 0.9923, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 17.54, | |
| "eval_cer": 0.0543247344461305, | |
| "eval_loss": 0.2073642760515213, | |
| "eval_runtime": 15.1305, | |
| "eval_samples_per_second": 22.141, | |
| "eval_steps_per_second": 0.397, | |
| "eval_wer": 0.3110850897736144, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 21.05, | |
| "learning_rate": 8e-05, | |
| "loss": 0.972, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 21.05, | |
| "eval_cer": 0.05361659079413252, | |
| "eval_loss": 0.21649114787578583, | |
| "eval_runtime": 14.9827, | |
| "eval_samples_per_second": 22.359, | |
| "eval_steps_per_second": 0.4, | |
| "eval_wer": 0.29547228727556596, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 24.56, | |
| "learning_rate": 8e-05, | |
| "loss": 0.9587, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 24.56, | |
| "eval_cer": 0.05351542741527567, | |
| "eval_loss": 0.2064175009727478, | |
| "eval_runtime": 15.1921, | |
| "eval_samples_per_second": 22.051, | |
| "eval_steps_per_second": 0.395, | |
| "eval_wer": 0.3017174082747853, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 28.07, | |
| "learning_rate": 8e-05, | |
| "loss": 0.9421, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 28.07, | |
| "eval_cer": 0.051947395042994435, | |
| "eval_loss": 0.2061864584684372, | |
| "eval_runtime": 15.0418, | |
| "eval_samples_per_second": 22.271, | |
| "eval_steps_per_second": 0.399, | |
| "eval_wer": 0.28844652615144417, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 31.58, | |
| "learning_rate": 7.059500000000001e-05, | |
| "loss": 0.9189, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 31.58, | |
| "eval_cer": 0.05073343449671219, | |
| "eval_loss": 0.2014162391424179, | |
| "eval_runtime": 15.125, | |
| "eval_samples_per_second": 22.149, | |
| "eval_steps_per_second": 0.397, | |
| "eval_wer": 0.2822014051522248, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 35.09, | |
| "learning_rate": 6.109500000000001e-05, | |
| "loss": 0.8919, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 35.09, | |
| "eval_cer": 0.04881133029843197, | |
| "eval_loss": 0.19518214464187622, | |
| "eval_runtime": 15.0854, | |
| "eval_samples_per_second": 22.207, | |
| "eval_steps_per_second": 0.398, | |
| "eval_wer": 0.2689305230288837, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 38.6, | |
| "learning_rate": 5.1594999999999996e-05, | |
| "loss": 0.8615, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 38.6, | |
| "eval_cer": 0.04795144157814871, | |
| "eval_loss": 0.20196911692619324, | |
| "eval_runtime": 15.1604, | |
| "eval_samples_per_second": 22.097, | |
| "eval_steps_per_second": 0.396, | |
| "eval_wer": 0.2685402029664325, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 42.11, | |
| "learning_rate": 4.2095e-05, | |
| "loss": 0.834, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 42.11, | |
| "eval_cer": 0.04668689934243804, | |
| "eval_loss": 0.2001034915447235, | |
| "eval_runtime": 15.091, | |
| "eval_samples_per_second": 22.199, | |
| "eval_steps_per_second": 0.398, | |
| "eval_wer": 0.2654176424668228, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 45.61, | |
| "learning_rate": 3.2595e-05, | |
| "loss": 0.8056, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 45.61, | |
| "eval_cer": 0.04481537683358624, | |
| "eval_loss": 0.1934908777475357, | |
| "eval_runtime": 15.4173, | |
| "eval_samples_per_second": 21.729, | |
| "eval_steps_per_second": 0.389, | |
| "eval_wer": 0.2498048399687744, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 49.12, | |
| "learning_rate": 2.3095e-05, | |
| "loss": 0.7888, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 49.12, | |
| "eval_cer": 0.04461305007587253, | |
| "eval_loss": 0.18915079534053802, | |
| "eval_runtime": 15.29, | |
| "eval_samples_per_second": 21.91, | |
| "eval_steps_per_second": 0.392, | |
| "eval_wer": 0.24512099921935987, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 52.63, | |
| "learning_rate": 1.3595000000000008e-05, | |
| "loss": 0.761, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 52.63, | |
| "eval_cer": 0.044056651492159836, | |
| "eval_loss": 0.18836112320423126, | |
| "eval_runtime": 16.204, | |
| "eval_samples_per_second": 20.674, | |
| "eval_steps_per_second": 0.37, | |
| "eval_wer": 0.24316939890710382, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 56.14, | |
| "learning_rate": 4.095000000000005e-06, | |
| "loss": 0.742, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 56.14, | |
| "eval_cer": 0.04274152756702074, | |
| "eval_loss": 0.18266724050045013, | |
| "eval_runtime": 15.3463, | |
| "eval_samples_per_second": 21.829, | |
| "eval_steps_per_second": 0.391, | |
| "eval_wer": 0.2388758782201405, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 56.14, | |
| "step": 3200, | |
| "total_flos": 2.6268075931237872e+20, | |
| "train_loss": 1.1288447761535645, | |
| "train_runtime": 34425.8492, | |
| "train_samples_per_second": 11.898, | |
| "train_steps_per_second": 0.093 | |
| } | |
| ], | |
| "max_steps": 3200, | |
| "num_train_epochs": 57, | |
| "total_flos": 2.6268075931237872e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |