| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9568302779420463, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 4.945002956830278e-05, | |
| "loss": 6.8243, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.885866351271437e-05, | |
| "loss": 4.1701, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.8267297457125965e-05, | |
| "loss": 3.6081, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.7675931401537554e-05, | |
| "loss": 3.3137, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.7084565345949144e-05, | |
| "loss": 3.0057, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.6493199290360734e-05, | |
| "loss": 2.7865, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 4.590183323477233e-05, | |
| "loss": 2.4937, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 4.531046717918392e-05, | |
| "loss": 2.2845, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 4.471910112359551e-05, | |
| "loss": 1.8935, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 4.41277350680071e-05, | |
| "loss": 1.6164, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "eval_cer": 0.5792523669604962, | |
| "eval_loss": 1.4108693599700928, | |
| "eval_runtime": 78.7069, | |
| "eval_samples_per_second": 19.096, | |
| "eval_steps_per_second": 2.389, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 4.353636901241869e-05, | |
| "loss": 1.3669, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 4.294500295683028e-05, | |
| "loss": 1.1989, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 4.235363690124187e-05, | |
| "loss": 1.0658, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.176227084565346e-05, | |
| "loss": 0.8932, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 4.117090479006505e-05, | |
| "loss": 0.8178, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 4.057953873447664e-05, | |
| "loss": 0.6796, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 3.9988172678888235e-05, | |
| "loss": 0.6462, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 3.9396806623299825e-05, | |
| "loss": 0.4301, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 3.8805440567711414e-05, | |
| "loss": 0.4144, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 3.8214074512123004e-05, | |
| "loss": 0.3434, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "eval_cer": 0.2175971269996735, | |
| "eval_loss": 0.3876412510871887, | |
| "eval_runtime": 79.3343, | |
| "eval_samples_per_second": 18.945, | |
| "eval_steps_per_second": 2.37, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "learning_rate": 3.7622708456534594e-05, | |
| "loss": 0.3263, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 3.703134240094619e-05, | |
| "loss": 0.2942, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 3.643997634535778e-05, | |
| "loss": 0.2743, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "learning_rate": 3.584861028976937e-05, | |
| "loss": 0.2679, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 3.525724423418096e-05, | |
| "loss": 0.2553, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "learning_rate": 3.466587817859255e-05, | |
| "loss": 0.2329, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 3.4074512123004146e-05, | |
| "loss": 0.171, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "learning_rate": 3.3483146067415736e-05, | |
| "loss": 0.2264, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 3.2891780011827325e-05, | |
| "loss": 0.2091, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "learning_rate": 3.2300413956238915e-05, | |
| "loss": 0.1679, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "eval_cer": 0.11859288279464578, | |
| "eval_loss": 0.22618062794208527, | |
| "eval_runtime": 78.6862, | |
| "eval_samples_per_second": 19.101, | |
| "eval_steps_per_second": 2.389, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 3.17090479006505e-05, | |
| "loss": 0.1629, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 3.1117681845062095e-05, | |
| "loss": 0.1669, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "learning_rate": 3.0526315789473684e-05, | |
| "loss": 0.1582, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "learning_rate": 2.9934949733885274e-05, | |
| "loss": 0.1066, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "learning_rate": 2.9343583678296864e-05, | |
| "loss": 0.08, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "learning_rate": 2.875221762270846e-05, | |
| "loss": 0.0651, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "learning_rate": 2.816085156712005e-05, | |
| "loss": 0.063, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "learning_rate": 2.756948551153164e-05, | |
| "loss": 0.045, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "learning_rate": 2.697811945594323e-05, | |
| "loss": 0.0691, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "learning_rate": 2.638675340035482e-05, | |
| "loss": 0.0816, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "eval_cer": 0.06341821743388834, | |
| "eval_loss": 0.12735812366008759, | |
| "eval_runtime": 79.0538, | |
| "eval_samples_per_second": 19.012, | |
| "eval_steps_per_second": 2.378, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "learning_rate": 2.5795387344766413e-05, | |
| "loss": 0.0732, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "learning_rate": 2.5204021289178002e-05, | |
| "loss": 0.0486, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "learning_rate": 2.4612655233589592e-05, | |
| "loss": 0.0618, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "learning_rate": 2.4021289178001182e-05, | |
| "loss": 0.0494, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "learning_rate": 2.3429923122412775e-05, | |
| "loss": 0.0557, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "learning_rate": 2.2838557066824365e-05, | |
| "loss": 0.045, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "learning_rate": 2.2247191011235958e-05, | |
| "loss": 0.0346, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "learning_rate": 2.1655824955647548e-05, | |
| "loss": 0.0437, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "learning_rate": 2.1064458900059137e-05, | |
| "loss": 0.0366, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "learning_rate": 2.0473092844470727e-05, | |
| "loss": 0.0421, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "eval_cer": 0.0381162259222984, | |
| "eval_loss": 0.08168739080429077, | |
| "eval_runtime": 78.8974, | |
| "eval_samples_per_second": 19.05, | |
| "eval_steps_per_second": 2.383, | |
| "step": 5000 | |
| } | |
| ], | |
| "max_steps": 8455, | |
| "num_train_epochs": 5, | |
| "total_flos": 7.242528907589714e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |