| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.19019281123542, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 0.0003, | |
| "loss": 2235.1048, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 0.0003, | |
| "loss": 1615.3083, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 0.0003, | |
| "loss": 1223.9066, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 0.0003, | |
| "loss": 602.0252, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 0.0003, | |
| "loss": 478.0811, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "eval_cer": 0.2157549269468942, | |
| "eval_loss": 376.9426574707031, | |
| "eval_runtime": 1031.6167, | |
| "eval_samples_per_second": 7.186, | |
| "eval_steps_per_second": 0.719, | |
| "eval_wer": 0.7605279103234114, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 0.0003, | |
| "loss": 421.4082, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 0.0003, | |
| "loss": 394.7663, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 0.0003, | |
| "loss": 379.9422, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 0.0003, | |
| "loss": 373.7412, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 0.0003, | |
| "loss": 349.4335, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "eval_cer": 0.15168241477586192, | |
| "eval_loss": 274.2783508300781, | |
| "eval_runtime": 1026.2842, | |
| "eval_samples_per_second": 7.223, | |
| "eval_steps_per_second": 0.723, | |
| "eval_wer": 0.5870256759827311, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 0.0003, | |
| "loss": 346.411, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 0.0003, | |
| "loss": 335.0877, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 0.0003, | |
| "loss": 330.8003, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 0.0003, | |
| "loss": 324.7547, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 0.0003, | |
| "loss": 308.4796, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "eval_cer": 0.13990026347021764, | |
| "eval_loss": 249.27769470214844, | |
| "eval_runtime": 1029.4748, | |
| "eval_samples_per_second": 7.201, | |
| "eval_steps_per_second": 0.721, | |
| "eval_wer": 0.5321707187760357, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 0.0003, | |
| "loss": 311.2614, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 0.0003, | |
| "loss": 299.9645, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 0.0003, | |
| "loss": 299.9626, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 0.0003, | |
| "loss": 291.1237, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 0.0003, | |
| "loss": 280.2542, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "eval_cer": 0.12837916028623791, | |
| "eval_loss": 232.43905639648438, | |
| "eval_runtime": 1027.1097, | |
| "eval_samples_per_second": 7.217, | |
| "eval_steps_per_second": 0.722, | |
| "eval_wer": 0.48547678557903506, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 0.0003, | |
| "loss": 297.7207, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 0.0003, | |
| "loss": 278.3572, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 0.0003, | |
| "loss": 282.0629, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 0.0003, | |
| "loss": 277.2473, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 0.0003, | |
| "loss": 270.2169, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "eval_cer": 0.12593015213453937, | |
| "eval_loss": 211.3223114013672, | |
| "eval_runtime": 1032.1072, | |
| "eval_samples_per_second": 7.182, | |
| "eval_steps_per_second": 0.719, | |
| "eval_wer": 0.46879497083996063, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 0.0003, | |
| "loss": 272.5899, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 0.0003, | |
| "loss": 262.8966, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 0.0003, | |
| "loss": 267.9274, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 0.0003, | |
| "loss": 254.0362, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 0.0003, | |
| "loss": 261.8612, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "eval_cer": 0.11589729236582261, | |
| "eval_loss": 196.6202850341797, | |
| "eval_runtime": 1032.1872, | |
| "eval_samples_per_second": 7.182, | |
| "eval_steps_per_second": 0.719, | |
| "eval_wer": 0.4461864727713398, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 0.0003, | |
| "loss": 263.8872, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 0.0003, | |
| "loss": 263.278, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 0.0003, | |
| "loss": 252.0182, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 0.0003, | |
| "loss": 256.6052, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 0.0003, | |
| "loss": 251.7534, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "eval_cer": 0.12125281568657002, | |
| "eval_loss": 186.7011260986328, | |
| "eval_runtime": 1035.9213, | |
| "eval_samples_per_second": 7.156, | |
| "eval_steps_per_second": 0.716, | |
| "eval_wer": 0.4463568885859274, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 0.0003, | |
| "loss": 230.9344, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 0.0003, | |
| "loss": 246.4894, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 0.0003, | |
| "loss": 239.637, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 0.0003, | |
| "loss": 251.8844, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 0.0003, | |
| "loss": 242.5334, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "eval_cer": 0.10615777533175987, | |
| "eval_loss": 173.9308319091797, | |
| "eval_runtime": 1037.0086, | |
| "eval_samples_per_second": 7.148, | |
| "eval_steps_per_second": 0.716, | |
| "eval_wer": 0.4061387563432553, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 0.0003, | |
| "loss": 237.7298, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 0.0003, | |
| "loss": 238.2786, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 0.0003, | |
| "loss": 209.7386, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "learning_rate": 0.0003, | |
| "loss": 201.507, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 0.0003, | |
| "loss": 217.7602, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "eval_cer": 0.1168149976182723, | |
| "eval_loss": 161.6786651611328, | |
| "eval_runtime": 1036.8937, | |
| "eval_samples_per_second": 7.149, | |
| "eval_steps_per_second": 0.716, | |
| "eval_wer": 0.4201885935014769, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 0.0003, | |
| "loss": 209.7251, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 0.0003, | |
| "loss": 212.4263, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 0.0003, | |
| "loss": 220.1039, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 0.0003, | |
| "loss": 218.4896, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 0.0, | |
| "loss": 217.312, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "eval_cer": 0.11287236361581252, | |
| "eval_loss": 172.40834045410156, | |
| "eval_runtime": 1040.3699, | |
| "eval_samples_per_second": 7.125, | |
| "eval_steps_per_second": 0.713, | |
| "eval_wer": 0.41411042944785276, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "step": 5000, | |
| "total_flos": 8.046046597054969e+18, | |
| "train_loss": 372.1765015625, | |
| "train_runtime": 24790.5372, | |
| "train_samples_per_second": 2.017, | |
| "train_steps_per_second": 0.202 | |
| } | |
| ], | |
| "max_steps": 5000, | |
| "num_train_epochs": 2, | |
| "total_flos": 8.046046597054969e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |