| { | |
| "best_metric": 0.18776850201669637, | |
| "best_model_checkpoint": "/workspace/cv-corpus-8.0-2022-01-19/output/checkpoint-13000", | |
| "epoch": 8.317338451695457, | |
| "global_step": 13000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 9.92e-05, | |
| "loss": 6.7916, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 0.00019920000000000002, | |
| "loss": 1.354, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "eval_loss": 0.4108898937702179, | |
| "eval_runtime": 71.0896, | |
| "eval_samples_per_second": 146.125, | |
| "eval_steps_per_second": 1.829, | |
| "eval_wer": 0.4493387111903199, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 0.00019321941216678059, | |
| "loss": 0.7084, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 0.00018638414217361587, | |
| "loss": 0.5886, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "eval_loss": 0.2797781527042389, | |
| "eval_runtime": 71.0704, | |
| "eval_samples_per_second": 146.165, | |
| "eval_steps_per_second": 1.829, | |
| "eval_wer": 0.3099334021198762, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 0.00017954887218045112, | |
| "loss": 0.5386, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 0.0001727136021872864, | |
| "loss": 0.4977, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "eval_loss": 0.23867885768413544, | |
| "eval_runtime": 71.0833, | |
| "eval_samples_per_second": 146.138, | |
| "eval_steps_per_second": 1.829, | |
| "eval_wer": 0.2673295188068662, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 0.00016587833219412168, | |
| "loss": 0.4531, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 0.00015905673274094328, | |
| "loss": 0.4253, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "eval_loss": 0.22657370567321777, | |
| "eval_runtime": 71.5875, | |
| "eval_samples_per_second": 145.109, | |
| "eval_steps_per_second": 1.816, | |
| "eval_wer": 0.2523215458212175, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "learning_rate": 0.00015222146274777856, | |
| "loss": 0.413, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 0.0001453861927546138, | |
| "loss": 0.3942, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "eval_loss": 0.21706202626228333, | |
| "eval_runtime": 72.0993, | |
| "eval_samples_per_second": 144.079, | |
| "eval_steps_per_second": 1.803, | |
| "eval_wer": 0.2437294812869337, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "learning_rate": 0.00013855092276144907, | |
| "loss": 0.3741, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "learning_rate": 0.00013171565276828435, | |
| "loss": 0.3619, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "eval_loss": 0.20762862265110016, | |
| "eval_runtime": 72.0546, | |
| "eval_samples_per_second": 144.168, | |
| "eval_steps_per_second": 1.804, | |
| "eval_wer": 0.22530719444705, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "learning_rate": 0.00012489405331510595, | |
| "loss": 0.3435, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "learning_rate": 0.00011805878332194122, | |
| "loss": 0.3245, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "eval_loss": 0.2087564468383789, | |
| "eval_runtime": 71.9965, | |
| "eval_samples_per_second": 144.285, | |
| "eval_steps_per_second": 1.806, | |
| "eval_wer": 0.21862864646843636, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "learning_rate": 0.0001112235133287765, | |
| "loss": 0.3135, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 5.12, | |
| "learning_rate": 0.0001044019138755981, | |
| "loss": 0.308, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 5.12, | |
| "eval_loss": 0.2086208015680313, | |
| "eval_runtime": 68.9232, | |
| "eval_samples_per_second": 150.718, | |
| "eval_steps_per_second": 1.886, | |
| "eval_wer": 0.22063596285526685, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 5.44, | |
| "learning_rate": 9.756664388243337e-05, | |
| "loss": 0.292, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 5.76, | |
| "learning_rate": 9.073137388926864e-05, | |
| "loss": 0.2881, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 5.76, | |
| "eval_loss": 0.20888157188892365, | |
| "eval_runtime": 70.2391, | |
| "eval_samples_per_second": 147.895, | |
| "eval_steps_per_second": 1.851, | |
| "eval_wer": 0.21048682112372197, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 6.08, | |
| "learning_rate": 8.389610389610389e-05, | |
| "loss": 0.2717, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "learning_rate": 7.706083390293917e-05, | |
| "loss": 0.2557, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "eval_loss": 0.20148096978664398, | |
| "eval_runtime": 70.1726, | |
| "eval_samples_per_second": 148.035, | |
| "eval_steps_per_second": 1.853, | |
| "eval_wer": 0.20035643935840916, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 6.72, | |
| "learning_rate": 7.022556390977444e-05, | |
| "loss": 0.2536, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 7.04, | |
| "learning_rate": 6.33902939166097e-05, | |
| "loss": 0.248, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 7.04, | |
| "eval_loss": 0.2043762356042862, | |
| "eval_runtime": 70.2184, | |
| "eval_samples_per_second": 147.938, | |
| "eval_steps_per_second": 1.851, | |
| "eval_wer": 0.19529124847575274, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 7.36, | |
| "learning_rate": 5.655502392344498e-05, | |
| "loss": 0.2308, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 7.68, | |
| "learning_rate": 4.971975393028025e-05, | |
| "loss": 0.2251, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 7.68, | |
| "eval_loss": 0.20575200021266937, | |
| "eval_runtime": 70.8946, | |
| "eval_samples_per_second": 146.527, | |
| "eval_steps_per_second": 1.834, | |
| "eval_wer": 0.19315261232529782, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 4.291182501708818e-05, | |
| "loss": 0.2207, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 8.32, | |
| "learning_rate": 3.6076555023923446e-05, | |
| "loss": 0.2052, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 8.32, | |
| "eval_loss": 0.21170856058597565, | |
| "eval_runtime": 70.7957, | |
| "eval_samples_per_second": 146.732, | |
| "eval_steps_per_second": 1.836, | |
| "eval_wer": 0.18776850201669637, | |
| "step": 13000 | |
| } | |
| ], | |
| "max_steps": 15630, | |
| "num_train_epochs": 10, | |
| "total_flos": 8.268631785694088e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |