| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 500.0, | |
| "global_step": 6000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 8.33, | |
| "learning_rate": 5.82e-05, | |
| "loss": 11.9523, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 16.67, | |
| "learning_rate": 0.0001182, | |
| "loss": 3.7399, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "learning_rate": 0.00017819999999999997, | |
| "loss": 3.074, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 33.33, | |
| "learning_rate": 0.0002382, | |
| "loss": 2.8995, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 41.67, | |
| "learning_rate": 0.0002982, | |
| "loss": 2.2627, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 41.67, | |
| "eval_loss": 1.4175914525985718, | |
| "eval_runtime": 5.6701, | |
| "eval_samples_per_second": 25.044, | |
| "eval_steps_per_second": 0.882, | |
| "eval_wer": 0.8282476024411508, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 0.0002947090909090909, | |
| "loss": 1.5352, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 58.33, | |
| "learning_rate": 0.00028925454545454543, | |
| "loss": 1.2177, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 66.67, | |
| "learning_rate": 0.00028379999999999996, | |
| "loss": 0.9821, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 75.0, | |
| "learning_rate": 0.0002783454545454545, | |
| "loss": 0.8097, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 83.33, | |
| "learning_rate": 0.0002728909090909091, | |
| "loss": 0.6692, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 83.33, | |
| "eval_loss": 1.3905961513519287, | |
| "eval_runtime": 5.5949, | |
| "eval_samples_per_second": 25.38, | |
| "eval_steps_per_second": 0.894, | |
| "eval_wer": 0.7375762859633828, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 91.67, | |
| "learning_rate": 0.0002674363636363636, | |
| "loss": 0.5587, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "learning_rate": 0.0002619818181818182, | |
| "loss": 0.4692, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 108.33, | |
| "learning_rate": 0.0002565272727272727, | |
| "loss": 0.3949, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 116.67, | |
| "learning_rate": 0.00025107272727272725, | |
| "loss": 0.3484, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 125.0, | |
| "learning_rate": 0.0002456181818181818, | |
| "loss": 0.2874, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 125.0, | |
| "eval_loss": 1.791098952293396, | |
| "eval_runtime": 5.4161, | |
| "eval_samples_per_second": 26.218, | |
| "eval_steps_per_second": 0.923, | |
| "eval_wer": 0.7733217088055798, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 133.33, | |
| "learning_rate": 0.00024016363636363634, | |
| "loss": 0.2444, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 141.67, | |
| "learning_rate": 0.0002347090909090909, | |
| "loss": 0.2249, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 150.0, | |
| "learning_rate": 0.00022925454545454543, | |
| "loss": 0.1996, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 158.33, | |
| "learning_rate": 0.0002238, | |
| "loss": 0.1676, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 166.67, | |
| "learning_rate": 0.00021834545454545452, | |
| "loss": 0.1504, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 166.67, | |
| "eval_loss": 1.761537790298462, | |
| "eval_runtime": 5.7426, | |
| "eval_samples_per_second": 24.728, | |
| "eval_steps_per_second": 0.871, | |
| "eval_wer": 0.8169136878814298, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 175.0, | |
| "learning_rate": 0.00021289090909090905, | |
| "loss": 0.1356, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 183.33, | |
| "learning_rate": 0.0002074363636363636, | |
| "loss": 0.1207, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 191.67, | |
| "learning_rate": 0.00020198181818181814, | |
| "loss": 0.119, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 200.0, | |
| "learning_rate": 0.0001965272727272727, | |
| "loss": 0.1085, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 208.33, | |
| "learning_rate": 0.00019107272727272723, | |
| "loss": 0.102, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 208.33, | |
| "eval_loss": 1.9240303039550781, | |
| "eval_runtime": 5.3949, | |
| "eval_samples_per_second": 26.321, | |
| "eval_steps_per_second": 0.927, | |
| "eval_wer": 0.8456843940714909, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 216.67, | |
| "learning_rate": 0.0001856181818181818, | |
| "loss": 0.0899, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 225.0, | |
| "learning_rate": 0.00018016363636363632, | |
| "loss": 0.0838, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 233.33, | |
| "learning_rate": 0.0001747090909090909, | |
| "loss": 0.0825, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 241.67, | |
| "learning_rate": 0.00016925454545454547, | |
| "loss": 0.0699, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 250.0, | |
| "learning_rate": 0.0001638, | |
| "loss": 0.071, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 250.0, | |
| "eval_loss": 1.9720656871795654, | |
| "eval_runtime": 5.5146, | |
| "eval_samples_per_second": 25.75, | |
| "eval_steps_per_second": 0.907, | |
| "eval_wer": 0.7741935483870968, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 258.33, | |
| "learning_rate": 0.00015834545454545456, | |
| "loss": 0.0616, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 266.67, | |
| "learning_rate": 0.00015289090909090909, | |
| "loss": 0.0564, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 275.0, | |
| "learning_rate": 0.00014743636363636362, | |
| "loss": 0.0538, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 283.33, | |
| "learning_rate": 0.00014198181818181818, | |
| "loss": 0.047, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 291.67, | |
| "learning_rate": 0.0001365272727272727, | |
| "loss": 0.046, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 291.67, | |
| "eval_loss": 2.0821120738983154, | |
| "eval_runtime": 5.458, | |
| "eval_samples_per_second": 26.017, | |
| "eval_steps_per_second": 0.916, | |
| "eval_wer": 0.8326068003487358, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 300.0, | |
| "learning_rate": 0.00013107272727272727, | |
| "loss": 0.0408, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 308.33, | |
| "learning_rate": 0.0001256181818181818, | |
| "loss": 0.0401, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 316.67, | |
| "learning_rate": 0.00012016363636363635, | |
| "loss": 0.037, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 325.0, | |
| "learning_rate": 0.0001147090909090909, | |
| "loss": 0.0362, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 333.33, | |
| "learning_rate": 0.00010925454545454544, | |
| "loss": 0.0357, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 333.33, | |
| "eval_loss": 2.052865505218506, | |
| "eval_runtime": 5.4898, | |
| "eval_samples_per_second": 25.866, | |
| "eval_steps_per_second": 0.911, | |
| "eval_wer": 0.8308631211857018, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 341.67, | |
| "learning_rate": 0.00010379999999999999, | |
| "loss": 0.0336, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 350.0, | |
| "learning_rate": 9.834545454545455e-05, | |
| "loss": 0.0283, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 358.33, | |
| "learning_rate": 9.289090909090909e-05, | |
| "loss": 0.0274, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 366.67, | |
| "learning_rate": 8.743636363636364e-05, | |
| "loss": 0.0256, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 375.0, | |
| "learning_rate": 8.198181818181818e-05, | |
| "loss": 0.0241, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 375.0, | |
| "eval_loss": 2.0744874477386475, | |
| "eval_runtime": 5.4725, | |
| "eval_samples_per_second": 25.948, | |
| "eval_steps_per_second": 0.914, | |
| "eval_wer": 0.8029642545771578, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 383.33, | |
| "learning_rate": 7.652727272727273e-05, | |
| "loss": 0.0266, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 391.67, | |
| "learning_rate": 7.107272727272727e-05, | |
| "loss": 0.0243, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 400.0, | |
| "learning_rate": 6.561818181818182e-05, | |
| "loss": 0.0191, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 408.33, | |
| "learning_rate": 6.016363636363636e-05, | |
| "loss": 0.0193, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 416.67, | |
| "learning_rate": 5.4709090909090905e-05, | |
| "loss": 0.0165, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 416.67, | |
| "eval_loss": 2.2303550243377686, | |
| "eval_runtime": 5.5806, | |
| "eval_samples_per_second": 25.445, | |
| "eval_steps_per_second": 0.896, | |
| "eval_wer": 0.8238884045335658, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 425.0, | |
| "learning_rate": 4.925454545454545e-05, | |
| "loss": 0.0195, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 433.33, | |
| "learning_rate": 4.3799999999999994e-05, | |
| "loss": 0.0159, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 441.67, | |
| "learning_rate": 3.84e-05, | |
| "loss": 0.0146, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 450.0, | |
| "learning_rate": 3.294545454545454e-05, | |
| "loss": 0.0145, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 458.33, | |
| "learning_rate": 2.749090909090909e-05, | |
| "loss": 0.0114, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 458.33, | |
| "eval_loss": 2.157787561416626, | |
| "eval_runtime": 5.9361, | |
| "eval_samples_per_second": 23.921, | |
| "eval_steps_per_second": 0.842, | |
| "eval_wer": 0.8430688753269399, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 466.67, | |
| "learning_rate": 2.2036363636363632e-05, | |
| "loss": 0.0132, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 475.0, | |
| "learning_rate": 1.658181818181818e-05, | |
| "loss": 0.0142, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 483.33, | |
| "learning_rate": 1.1127272727272727e-05, | |
| "loss": 0.0132, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 491.67, | |
| "learning_rate": 5.672727272727272e-06, | |
| "loss": 0.0128, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 500.0, | |
| "learning_rate": 2.1818181818181815e-07, | |
| "loss": 0.012, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 500.0, | |
| "eval_loss": 2.165587902069092, | |
| "eval_runtime": 5.7782, | |
| "eval_samples_per_second": 24.575, | |
| "eval_steps_per_second": 0.865, | |
| "eval_wer": 0.8352223190932868, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 500.0, | |
| "step": 6000, | |
| "total_flos": 3.5552286921447256e+19, | |
| "train_loss": 0.5668576599160831, | |
| "train_runtime": 11637.7564, | |
| "train_samples_per_second": 16.24, | |
| "train_steps_per_second": 0.516 | |
| } | |
| ], | |
| "max_steps": 6000, | |
| "num_train_epochs": 500, | |
| "total_flos": 3.5552286921447256e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |