| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 6.191950464396285, | |
| "eval_steps": 100, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.30959752321981426, | |
| "eval_loss": 3.562331438064575, | |
| "eval_runtime": 151.987, | |
| "eval_samples_per_second": 37.214, | |
| "eval_steps_per_second": 4.652, | |
| "eval_wer": 1.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6191950464396285, | |
| "eval_loss": 3.1175434589385986, | |
| "eval_runtime": 149.1497, | |
| "eval_samples_per_second": 37.922, | |
| "eval_steps_per_second": 4.74, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.9287925696594427, | |
| "eval_loss": 3.293273687362671, | |
| "eval_runtime": 149.8183, | |
| "eval_samples_per_second": 37.752, | |
| "eval_steps_per_second": 4.719, | |
| "eval_wer": 1.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.238390092879257, | |
| "eval_loss": 0.9764726161956787, | |
| "eval_runtime": 149.5566, | |
| "eval_samples_per_second": 37.818, | |
| "eval_steps_per_second": 4.727, | |
| "eval_wer": 0.6866203399078814, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.5479876160990713, | |
| "grad_norm": 0.48750218749046326, | |
| "learning_rate": 0.00029519999999999997, | |
| "loss": 3.7398, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.5479876160990713, | |
| "eval_loss": 0.6787042617797852, | |
| "eval_runtime": 150.878, | |
| "eval_samples_per_second": 37.487, | |
| "eval_steps_per_second": 4.686, | |
| "eval_wer": 0.492176341255958, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.8575851393188856, | |
| "eval_loss": 0.6015087366104126, | |
| "eval_runtime": 150.6376, | |
| "eval_samples_per_second": 37.547, | |
| "eval_steps_per_second": 4.693, | |
| "eval_wer": 0.43465840702283703, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.1671826625387, | |
| "eval_loss": 0.5698955059051514, | |
| "eval_runtime": 150.4703, | |
| "eval_samples_per_second": 37.589, | |
| "eval_steps_per_second": 4.699, | |
| "eval_wer": 0.42732422846688384, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.476780185758514, | |
| "eval_loss": 0.5419376492500305, | |
| "eval_runtime": 150.5345, | |
| "eval_samples_per_second": 37.573, | |
| "eval_steps_per_second": 4.697, | |
| "eval_wer": 0.39578886552936077, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.7863777089783284, | |
| "eval_loss": 0.49708282947540283, | |
| "eval_runtime": 150.3649, | |
| "eval_samples_per_second": 37.615, | |
| "eval_steps_per_second": 4.702, | |
| "eval_wer": 0.3729839033236507, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.0959752321981426, | |
| "grad_norm": 0.4975910186767578, | |
| "learning_rate": 0.0002024, | |
| "loss": 0.5228, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.0959752321981426, | |
| "eval_loss": 0.49601131677627563, | |
| "eval_runtime": 150.6812, | |
| "eval_samples_per_second": 37.536, | |
| "eval_steps_per_second": 4.692, | |
| "eval_wer": 0.354303413522492, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.405572755417957, | |
| "eval_loss": 0.6976510286331177, | |
| "eval_runtime": 149.5565, | |
| "eval_samples_per_second": 37.818, | |
| "eval_steps_per_second": 4.727, | |
| "eval_wer": 0.4655036831378087, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.715170278637771, | |
| "eval_loss": 0.799861490726471, | |
| "eval_runtime": 151.2369, | |
| "eval_samples_per_second": 37.398, | |
| "eval_steps_per_second": 4.675, | |
| "eval_wer": 0.5316396783874436, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.024767801857585, | |
| "eval_loss": 1.0526387691497803, | |
| "eval_runtime": 150.3745, | |
| "eval_samples_per_second": 37.613, | |
| "eval_steps_per_second": 4.702, | |
| "eval_wer": 0.6214633050344242, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.3343653250774, | |
| "eval_loss": 1.1285134553909302, | |
| "eval_runtime": 150.4438, | |
| "eval_samples_per_second": 37.595, | |
| "eval_steps_per_second": 4.699, | |
| "eval_wer": 0.7571857296464509, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.643962848297214, | |
| "grad_norm": 3.2222559452056885, | |
| "learning_rate": 0.00010359999999999998, | |
| "loss": 0.9047, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.643962848297214, | |
| "eval_loss": 1.1592659950256348, | |
| "eval_runtime": 150.3382, | |
| "eval_samples_per_second": 37.622, | |
| "eval_steps_per_second": 4.703, | |
| "eval_wer": 0.7047070340710307, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.953560371517028, | |
| "eval_loss": 2.0400760173797607, | |
| "eval_runtime": 150.5385, | |
| "eval_samples_per_second": 37.572, | |
| "eval_steps_per_second": 4.696, | |
| "eval_wer": 0.966747444271477, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 5.2631578947368425, | |
| "eval_loss": 1.6264142990112305, | |
| "eval_runtime": 151.1274, | |
| "eval_samples_per_second": 37.425, | |
| "eval_steps_per_second": 4.678, | |
| "eval_wer": 0.8680489801158704, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 5.572755417956657, | |
| "eval_loss": 1.5916314125061035, | |
| "eval_runtime": 151.1725, | |
| "eval_samples_per_second": 37.414, | |
| "eval_steps_per_second": 4.677, | |
| "eval_wer": 0.8627369164353004, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 5.882352941176471, | |
| "eval_loss": 1.5764083862304688, | |
| "eval_runtime": 150.5972, | |
| "eval_samples_per_second": 37.557, | |
| "eval_steps_per_second": 4.695, | |
| "eval_wer": 0.8720129672128517, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 6.191950464396285, | |
| "grad_norm": 0.7321383953094482, | |
| "learning_rate": 4.399999999999999e-06, | |
| "loss": 1.6715, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 6.191950464396285, | |
| "eval_loss": 1.5562535524368286, | |
| "eval_runtime": 151.1236, | |
| "eval_samples_per_second": 37.426, | |
| "eval_steps_per_second": 4.678, | |
| "eval_wer": 0.8702315802988236, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 6.191950464396285, | |
| "step": 2000, | |
| "total_flos": 3.641759707440775e+19, | |
| "train_loss": 1.7097147216796875, | |
| "train_runtime": 12239.4176, | |
| "train_samples_per_second": 20.916, | |
| "train_steps_per_second": 0.163 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "save_steps": 400, | |
| "total_flos": 3.641759707440775e+19, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |