| { | |
| "best_metric": 0.25546682656055525, | |
| "best_model_checkpoint": "./jako-xlsr/checkpoint-11000", | |
| "epoch": 12.564249000571102, | |
| "eval_steps": 1000, | |
| "global_step": 11000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 0.00012563451776649744, | |
| "loss": 24.8869, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 0.0002512690355329949, | |
| "loss": 4.8854, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 0.00029761998271934643, | |
| "loss": 3.5667, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "eval_cer": 0.5188284138707948, | |
| "eval_loss": 2.2322683334350586, | |
| "eval_runtime": 333.3102, | |
| "eval_samples_per_second": 10.522, | |
| "eval_steps_per_second": 0.66, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 0.00029373183567669466, | |
| "loss": 2.1771, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "learning_rate": 0.0002898436886340429, | |
| "loss": 1.8324, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "learning_rate": 0.00028595554159139105, | |
| "loss": 1.5569, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "eval_cer": 0.3527113824461948, | |
| "eval_loss": 1.3106200695037842, | |
| "eval_runtime": 333.5876, | |
| "eval_samples_per_second": 10.513, | |
| "eval_steps_per_second": 0.659, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 0.00028206739454873927, | |
| "loss": 1.4453, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "learning_rate": 0.0002781792475060875, | |
| "loss": 1.38, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "learning_rate": 0.0002742911004634357, | |
| "loss": 1.2238, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "eval_cer": 0.3098580033250661, | |
| "eval_loss": 1.1108620166778564, | |
| "eval_runtime": 331.1347, | |
| "eval_samples_per_second": 10.591, | |
| "eval_steps_per_second": 0.664, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "learning_rate": 0.0002704029534207839, | |
| "loss": 1.203, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 4.15, | |
| "learning_rate": 0.0002665148063781321, | |
| "loss": 1.151, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 4.52, | |
| "learning_rate": 0.0002626266593354803, | |
| "loss": 1.0593, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 4.57, | |
| "eval_cer": 0.28911722220708075, | |
| "eval_loss": 1.0389584302902222, | |
| "eval_runtime": 331.5451, | |
| "eval_samples_per_second": 10.578, | |
| "eval_steps_per_second": 0.664, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "learning_rate": 0.0002587385122928285, | |
| "loss": 1.0674, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 5.28, | |
| "learning_rate": 0.0002548503652501767, | |
| "loss": 0.974, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 5.65, | |
| "learning_rate": 0.00025096221820752494, | |
| "loss": 0.9658, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 5.71, | |
| "eval_cer": 0.29180634669719185, | |
| "eval_loss": 0.973089337348938, | |
| "eval_runtime": 331.0672, | |
| "eval_samples_per_second": 10.593, | |
| "eval_steps_per_second": 0.665, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 6.03, | |
| "learning_rate": 0.0002470740711648731, | |
| "loss": 0.9596, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 6.41, | |
| "learning_rate": 0.00024318592412222133, | |
| "loss": 0.8626, | |
| "step": 5610 | |
| }, | |
| { | |
| "epoch": 6.78, | |
| "learning_rate": 0.00023929777707956955, | |
| "loss": 0.8796, | |
| "step": 5940 | |
| }, | |
| { | |
| "epoch": 6.85, | |
| "eval_cer": 0.26956656037357024, | |
| "eval_loss": 0.9479135870933533, | |
| "eval_runtime": 330.5292, | |
| "eval_samples_per_second": 10.61, | |
| "eval_steps_per_second": 0.666, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 7.16, | |
| "learning_rate": 0.00023540963003691772, | |
| "loss": 0.8311, | |
| "step": 6270 | |
| }, | |
| { | |
| "epoch": 7.54, | |
| "learning_rate": 0.00023152148299426594, | |
| "loss": 0.7889, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 7.92, | |
| "learning_rate": 0.00022763333595161414, | |
| "loss": 0.8022, | |
| "step": 6930 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_cer": 0.2710383109391041, | |
| "eval_loss": 0.9330962300300598, | |
| "eval_runtime": 330.2975, | |
| "eval_samples_per_second": 10.618, | |
| "eval_steps_per_second": 0.666, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 8.29, | |
| "learning_rate": 0.00022374518890896236, | |
| "loss": 0.7418, | |
| "step": 7260 | |
| }, | |
| { | |
| "epoch": 8.67, | |
| "learning_rate": 0.00021985704186631055, | |
| "loss": 0.739, | |
| "step": 7590 | |
| }, | |
| { | |
| "epoch": 9.05, | |
| "learning_rate": 0.00021596889482365878, | |
| "loss": 0.7392, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 9.14, | |
| "eval_cer": 0.2745814141524261, | |
| "eval_loss": 0.925165593624115, | |
| "eval_runtime": 338.4979, | |
| "eval_samples_per_second": 10.36, | |
| "eval_steps_per_second": 0.65, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 9.42, | |
| "learning_rate": 0.000212080747781007, | |
| "loss": 0.6765, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "learning_rate": 0.00020819260073835517, | |
| "loss": 0.6778, | |
| "step": 8580 | |
| }, | |
| { | |
| "epoch": 10.18, | |
| "learning_rate": 0.0002043044536957034, | |
| "loss": 0.6694, | |
| "step": 8910 | |
| }, | |
| { | |
| "epoch": 10.28, | |
| "eval_cer": 0.2590008448938432, | |
| "eval_loss": 0.9317852854728699, | |
| "eval_runtime": 336.5027, | |
| "eval_samples_per_second": 10.422, | |
| "eval_steps_per_second": 0.654, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 10.55, | |
| "learning_rate": 0.00020041630665305158, | |
| "loss": 0.621, | |
| "step": 9240 | |
| }, | |
| { | |
| "epoch": 10.93, | |
| "learning_rate": 0.0001965281596103998, | |
| "loss": 0.6494, | |
| "step": 9570 | |
| }, | |
| { | |
| "epoch": 11.31, | |
| "learning_rate": 0.000192640012567748, | |
| "loss": 0.5977, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 11.42, | |
| "eval_cer": 0.2674225286855087, | |
| "eval_loss": 0.9348525404930115, | |
| "eval_runtime": 338.5078, | |
| "eval_samples_per_second": 10.36, | |
| "eval_steps_per_second": 0.65, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 11.68, | |
| "learning_rate": 0.00018875186552509622, | |
| "loss": 0.5786, | |
| "step": 10230 | |
| }, | |
| { | |
| "epoch": 12.06, | |
| "learning_rate": 0.0001848637184824444, | |
| "loss": 0.582, | |
| "step": 10560 | |
| }, | |
| { | |
| "epoch": 12.44, | |
| "learning_rate": 0.00018097557143979261, | |
| "loss": 0.5484, | |
| "step": 10890 | |
| }, | |
| { | |
| "epoch": 12.56, | |
| "eval_cer": 0.25546682656055525, | |
| "eval_loss": 0.9408797025680542, | |
| "eval_runtime": 334.3327, | |
| "eval_samples_per_second": 10.49, | |
| "eval_steps_per_second": 0.658, | |
| "step": 11000 | |
| } | |
| ], | |
| "logging_steps": 330, | |
| "max_steps": 26250, | |
| "num_train_epochs": 30, | |
| "save_steps": 1000, | |
| "total_flos": 5.272851812517832e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |