| { | |
| "best_metric": 0.10217397660017014, | |
| "best_model_checkpoint": "/workspace/output/ka/wav2vec2-xls-r-1b-ka-2/checkpoint-4000", | |
| "epoch": 64.512, | |
| "global_step": 4000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.61, | |
| "learning_rate": 1.7678499999999997e-05, | |
| "loss": 5.1431, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "learning_rate": 3.500349999999999e-05, | |
| "loss": 2.6801, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 4.83, | |
| "learning_rate": 5.232849999999999e-05, | |
| "loss": 1.4529, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 6.45, | |
| "learning_rate": 6.965349999999999e-05, | |
| "loss": 1.2839, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 6.45, | |
| "eval_cer": 0.05568531627678561, | |
| "eval_loss": 0.22294116020202637, | |
| "eval_runtime": 52.1848, | |
| "eval_samples_per_second": 25.774, | |
| "eval_steps_per_second": 0.422, | |
| "eval_wer": 0.3608917324802559, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 8.06, | |
| "learning_rate": 7e-05, | |
| "loss": 1.2486, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 9.67, | |
| "learning_rate": 7e-05, | |
| "loss": 1.139, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 11.29, | |
| "learning_rate": 7e-05, | |
| "loss": 1.0481, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 12.9, | |
| "learning_rate": 7e-05, | |
| "loss": 0.9775, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 12.9, | |
| "eval_cer": 0.03166960125528526, | |
| "eval_loss": 0.127132385969162, | |
| "eval_runtime": 50.7158, | |
| "eval_samples_per_second": 26.52, | |
| "eval_steps_per_second": 0.434, | |
| "eval_wer": 0.2202339298210537, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 14.51, | |
| "learning_rate": 7e-05, | |
| "loss": 0.9449, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 16.13, | |
| "learning_rate": 7e-05, | |
| "loss": 0.9374, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 17.74, | |
| "learning_rate": 7e-05, | |
| "loss": 0.9197, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 19.35, | |
| "learning_rate": 7e-05, | |
| "loss": 0.9045, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 19.35, | |
| "eval_cer": 0.02939379783678896, | |
| "eval_loss": 0.12676405906677246, | |
| "eval_runtime": 50.6864, | |
| "eval_samples_per_second": 26.536, | |
| "eval_steps_per_second": 0.434, | |
| "eval_wer": 0.20303908827351794, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 20.96, | |
| "learning_rate": 7e-05, | |
| "loss": 0.8889, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 22.58, | |
| "learning_rate": 7e-05, | |
| "loss": 0.8873, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 24.19, | |
| "learning_rate": 7e-05, | |
| "loss": 0.8828, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 25.8, | |
| "learning_rate": 7e-05, | |
| "loss": 0.8652, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 25.8, | |
| "eval_cer": 0.028746990549426858, | |
| "eval_loss": 0.12111534923315048, | |
| "eval_runtime": 50.502, | |
| "eval_samples_per_second": 26.633, | |
| "eval_steps_per_second": 0.436, | |
| "eval_wer": 0.19404178746376088, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 27.42, | |
| "learning_rate": 7e-05, | |
| "loss": 0.8775, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 29.03, | |
| "learning_rate": 7e-05, | |
| "loss": 0.8587, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 30.64, | |
| "learning_rate": 7e-05, | |
| "loss": 0.8518, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 32.26, | |
| "learning_rate": 7e-05, | |
| "loss": 0.8505, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 32.26, | |
| "eval_cer": 0.027573155101991927, | |
| "eval_loss": 0.11918943375349045, | |
| "eval_runtime": 50.5112, | |
| "eval_samples_per_second": 26.628, | |
| "eval_steps_per_second": 0.436, | |
| "eval_wer": 0.19124262721183644, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 33.86, | |
| "learning_rate": 6.67415e-05, | |
| "loss": 0.8436, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 35.48, | |
| "learning_rate": 6.34165e-05, | |
| "loss": 0.8399, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 37.1, | |
| "learning_rate": 6.009149999999999e-05, | |
| "loss": 0.8258, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 38.7, | |
| "learning_rate": 5.679975e-05, | |
| "loss": 0.8168, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 38.7, | |
| "eval_cer": 0.025956136883586667, | |
| "eval_loss": 0.10860513150691986, | |
| "eval_runtime": 51.1919, | |
| "eval_samples_per_second": 26.274, | |
| "eval_steps_per_second": 0.43, | |
| "eval_wer": 0.17634709587123862, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 40.32, | |
| "learning_rate": 5.3474749999999995e-05, | |
| "loss": 0.8068, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 41.93, | |
| "learning_rate": 5.0149749999999995e-05, | |
| "loss": 0.7971, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 43.54, | |
| "learning_rate": 4.6824749999999996e-05, | |
| "loss": 0.7893, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 45.16, | |
| "learning_rate": 4.3499749999999996e-05, | |
| "loss": 0.7737, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 45.16, | |
| "eval_cer": 0.025620755327176687, | |
| "eval_loss": 0.10982357710599899, | |
| "eval_runtime": 50.6716, | |
| "eval_samples_per_second": 26.543, | |
| "eval_steps_per_second": 0.434, | |
| "eval_wer": 0.17534739578126562, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 46.77, | |
| "learning_rate": 4.017475e-05, | |
| "loss": 0.7748, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 48.38, | |
| "learning_rate": 3.684974999999999e-05, | |
| "loss": 0.762, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 49.99, | |
| "learning_rate": 3.352475e-05, | |
| "loss": 0.753, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 51.61, | |
| "learning_rate": 3.0199749999999995e-05, | |
| "loss": 0.744, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 51.61, | |
| "eval_cer": 0.02391989171966893, | |
| "eval_loss": 0.10541926324367523, | |
| "eval_runtime": 50.5525, | |
| "eval_samples_per_second": 26.606, | |
| "eval_steps_per_second": 0.435, | |
| "eval_wer": 0.16455063480955714, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 53.22, | |
| "learning_rate": 2.6874749999999995e-05, | |
| "loss": 0.7233, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 54.83, | |
| "learning_rate": 2.3549749999999996e-05, | |
| "loss": 0.725, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 56.45, | |
| "learning_rate": 2.0224749999999996e-05, | |
| "loss": 0.7158, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 58.06, | |
| "learning_rate": 1.6899749999999997e-05, | |
| "loss": 0.7114, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 58.06, | |
| "eval_cer": 0.022829901661336496, | |
| "eval_loss": 0.10338227450847626, | |
| "eval_runtime": 51.0232, | |
| "eval_samples_per_second": 26.361, | |
| "eval_steps_per_second": 0.431, | |
| "eval_wer": 0.15725282415275418, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 59.67, | |
| "learning_rate": 1.3574749999999999e-05, | |
| "loss": 0.696, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 61.29, | |
| "learning_rate": 1.0249750000000001e-05, | |
| "loss": 0.7046, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 62.9, | |
| "learning_rate": 6.924750000000003e-06, | |
| "loss": 0.6915, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 64.51, | |
| "learning_rate": 3.5997500000000035e-06, | |
| "loss": 0.6773, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 64.51, | |
| "eval_cer": 0.02207529315941404, | |
| "eval_loss": 0.10217397660017014, | |
| "eval_runtime": 50.9761, | |
| "eval_samples_per_second": 26.385, | |
| "eval_steps_per_second": 0.432, | |
| "eval_wer": 0.15265420373887834, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 64.51, | |
| "step": 4000, | |
| "total_flos": 3.3884437885491655e+20, | |
| "train_loss": 1.020361618041992, | |
| "train_runtime": 40092.8966, | |
| "train_samples_per_second": 12.77, | |
| "train_steps_per_second": 0.1 | |
| } | |
| ], | |
| "max_steps": 4000, | |
| "num_train_epochs": 65, | |
| "total_flos": 3.3884437885491655e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |