| { | |
| "best_metric": 0.5082262754440308, | |
| "best_model_checkpoint": "./checkpoint-1174", | |
| "epoch": 18.998296422487222, | |
| "global_step": 5576, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "eval_cer": 0.10684727658186562, | |
| "eval_loss": 0.6773447394371033, | |
| "eval_runtime": 207.0582, | |
| "eval_samples_per_second": 11.026, | |
| "eval_steps_per_second": 0.691, | |
| "eval_wer": 0.26575449135102064, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 3.660068259385666e-05, | |
| "loss": 0.9117, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_cer": 0.08164138943248532, | |
| "eval_loss": 0.536730945110321, | |
| "eval_runtime": 206.1216, | |
| "eval_samples_per_second": 11.076, | |
| "eval_steps_per_second": 0.694, | |
| "eval_wer": 0.21280382668669004, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_cer": 0.07224396607958251, | |
| "eval_loss": 0.5175191164016724, | |
| "eval_runtime": 202.5864, | |
| "eval_samples_per_second": 11.269, | |
| "eval_steps_per_second": 0.706, | |
| "eval_wer": 0.19884309472161968, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "learning_rate": 3.3187713310580205e-05, | |
| "loss": 0.2625, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_cer": 0.07057240704500979, | |
| "eval_loss": 0.5082262754440308, | |
| "eval_runtime": 199.851, | |
| "eval_samples_per_second": 11.424, | |
| "eval_steps_per_second": 0.716, | |
| "eval_wer": 0.1921130207464264, | |
| "step": 1174 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_cer": 0.0687275766470972, | |
| "eval_loss": 0.5106597542762756, | |
| "eval_runtime": 201.3886, | |
| "eval_samples_per_second": 11.336, | |
| "eval_steps_per_second": 0.71, | |
| "eval_wer": 0.1898325824573113, | |
| "step": 1467 | |
| }, | |
| { | |
| "epoch": 5.11, | |
| "learning_rate": 2.9774744027303753e-05, | |
| "loss": 0.112, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_cer": 0.06810583822570124, | |
| "eval_loss": 0.5201366543769836, | |
| "eval_runtime": 202.2293, | |
| "eval_samples_per_second": 11.289, | |
| "eval_steps_per_second": 0.707, | |
| "eval_wer": 0.18510484454085321, | |
| "step": 1761 | |
| }, | |
| { | |
| "epoch": 6.81, | |
| "learning_rate": 2.6361774744027308e-05, | |
| "loss": 0.0483, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_cer": 0.064925799086758, | |
| "eval_loss": 0.5361577272415161, | |
| "eval_runtime": 201.8712, | |
| "eval_samples_per_second": 11.309, | |
| "eval_steps_per_second": 0.708, | |
| "eval_wer": 0.18154513599199065, | |
| "step": 2054 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_cer": 0.06522137964774952, | |
| "eval_loss": 0.5386066436767578, | |
| "eval_runtime": 201.6773, | |
| "eval_samples_per_second": 11.32, | |
| "eval_steps_per_second": 0.709, | |
| "eval_wer": 0.18009900439401524, | |
| "step": 2348 | |
| }, | |
| { | |
| "epoch": 8.52, | |
| "learning_rate": 2.2948805460750853e-05, | |
| "loss": 0.022, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_cer": 0.0632644324853229, | |
| "eval_loss": 0.5573871731758118, | |
| "eval_runtime": 201.4546, | |
| "eval_samples_per_second": 11.333, | |
| "eval_steps_per_second": 0.71, | |
| "eval_wer": 0.17509316424717727, | |
| "step": 2641 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_cer": 0.06333577951728636, | |
| "eval_loss": 0.5539766550064087, | |
| "eval_runtime": 201.1768, | |
| "eval_samples_per_second": 11.348, | |
| "eval_steps_per_second": 0.711, | |
| "eval_wer": 0.1772623616441404, | |
| "step": 2935 | |
| }, | |
| { | |
| "epoch": 10.22, | |
| "learning_rate": 1.9535836177474404e-05, | |
| "loss": 0.0116, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_cer": 0.06017612524461839, | |
| "eval_loss": 0.5528882741928101, | |
| "eval_runtime": 201.6037, | |
| "eval_samples_per_second": 11.324, | |
| "eval_steps_per_second": 0.709, | |
| "eval_wer": 0.17019856499249125, | |
| "step": 3228 | |
| }, | |
| { | |
| "epoch": 11.93, | |
| "learning_rate": 1.6122866894197952e-05, | |
| "loss": 0.005, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_cer": 0.0598601598173516, | |
| "eval_loss": 0.5630314946174622, | |
| "eval_runtime": 201.2994, | |
| "eval_samples_per_second": 11.341, | |
| "eval_steps_per_second": 0.71, | |
| "eval_wer": 0.16864119250236387, | |
| "step": 3522 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_cer": 0.06034939660795825, | |
| "eval_loss": 0.5675850510597229, | |
| "eval_runtime": 201.2983, | |
| "eval_samples_per_second": 11.341, | |
| "eval_steps_per_second": 0.71, | |
| "eval_wer": 0.1693086378552756, | |
| "step": 3815 | |
| }, | |
| { | |
| "epoch": 13.63, | |
| "learning_rate": 1.2709897610921502e-05, | |
| "loss": 0.0035, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_cer": 0.05873899217221135, | |
| "eval_loss": 0.5630576610565186, | |
| "eval_runtime": 201.6395, | |
| "eval_samples_per_second": 11.322, | |
| "eval_steps_per_second": 0.709, | |
| "eval_wer": 0.1666388564436287, | |
| "step": 4109 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_cer": 0.058494373776908026, | |
| "eval_loss": 0.5673080682754517, | |
| "eval_runtime": 203.0728, | |
| "eval_samples_per_second": 11.242, | |
| "eval_steps_per_second": 0.704, | |
| "eval_wer": 0.16613827242894488, | |
| "step": 4402 | |
| }, | |
| { | |
| "epoch": 15.33, | |
| "learning_rate": 9.296928327645052e-06, | |
| "loss": 0.0018, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_cer": 0.0591466894977169, | |
| "eval_loss": 0.5768471956253052, | |
| "eval_runtime": 202.2731, | |
| "eval_samples_per_second": 11.287, | |
| "eval_steps_per_second": 0.707, | |
| "eval_wer": 0.16852995161021192, | |
| "step": 4696 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_cer": 0.05765859425962166, | |
| "eval_loss": 0.5726827383041382, | |
| "eval_runtime": 202.414, | |
| "eval_samples_per_second": 11.279, | |
| "eval_steps_per_second": 0.706, | |
| "eval_wer": 0.16636075421324878, | |
| "step": 4989 | |
| }, | |
| { | |
| "epoch": 17.04, | |
| "learning_rate": 5.883959044368602e-06, | |
| "loss": 0.0012, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_cer": 0.05809686888454012, | |
| "eval_loss": 0.5776260495185852, | |
| "eval_runtime": 202.9462, | |
| "eval_samples_per_second": 11.249, | |
| "eval_steps_per_second": 0.705, | |
| "eval_wer": 0.16647199510540076, | |
| "step": 5283 | |
| }, | |
| { | |
| "epoch": 18.74, | |
| "learning_rate": 2.4709897610921506e-06, | |
| "loss": 0.0008, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_cer": 0.05747513046314416, | |
| "eval_loss": 0.5768410563468933, | |
| "eval_runtime": 201.776, | |
| "eval_samples_per_second": 11.315, | |
| "eval_steps_per_second": 0.709, | |
| "eval_wer": 0.1651371043995773, | |
| "step": 5576 | |
| } | |
| ], | |
| "max_steps": 5860, | |
| "num_train_epochs": 20, | |
| "total_flos": 9.587705362442952e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |