| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.536067892503536, | |
| "eval_steps": 100, | |
| "global_step": 2500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.14144271570014144, | |
| "eval_loss": 3.7463672161102295, | |
| "eval_runtime": 153.5435, | |
| "eval_samples_per_second": 36.836, | |
| "eval_steps_per_second": 4.605, | |
| "eval_wer": 1.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2828854314002829, | |
| "eval_loss": 2.9399216175079346, | |
| "eval_runtime": 150.7703, | |
| "eval_samples_per_second": 37.514, | |
| "eval_steps_per_second": 4.689, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4243281471004243, | |
| "eval_loss": 2.5961458683013916, | |
| "eval_runtime": 151.4236, | |
| "eval_samples_per_second": 37.352, | |
| "eval_steps_per_second": 4.669, | |
| "eval_wer": 0.9991041576682503, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5657708628005658, | |
| "eval_loss": 1.1618728637695312, | |
| "eval_runtime": 152.5862, | |
| "eval_samples_per_second": 37.068, | |
| "eval_steps_per_second": 4.633, | |
| "eval_wer": 0.7905328662155461, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7072135785007072, | |
| "grad_norm": 1.7047498226165771, | |
| "learning_rate": 0.00029699999999999996, | |
| "loss": 3.5448, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7072135785007072, | |
| "eval_loss": 0.946560800075531, | |
| "eval_runtime": 153.1427, | |
| "eval_samples_per_second": 36.933, | |
| "eval_steps_per_second": 4.617, | |
| "eval_wer": 0.6897506038937147, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8486562942008486, | |
| "eval_loss": 0.7894724607467651, | |
| "eval_runtime": 151.7162, | |
| "eval_samples_per_second": 37.28, | |
| "eval_steps_per_second": 4.66, | |
| "eval_wer": 0.6110604533602086, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9900990099009901, | |
| "eval_loss": 0.6820164918899536, | |
| "eval_runtime": 152.0086, | |
| "eval_samples_per_second": 37.208, | |
| "eval_steps_per_second": 4.651, | |
| "eval_wer": 0.5378893314776599, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.1315417256011315, | |
| "eval_loss": 0.6039016842842102, | |
| "eval_runtime": 152.002, | |
| "eval_samples_per_second": 37.21, | |
| "eval_steps_per_second": 4.651, | |
| "eval_wer": 0.47239685815296506, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.272984441301273, | |
| "eval_loss": 0.5631398558616638, | |
| "eval_runtime": 153.0321, | |
| "eval_samples_per_second": 36.96, | |
| "eval_steps_per_second": 4.62, | |
| "eval_wer": 0.46745372814384667, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.4144271570014144, | |
| "grad_norm": 0.8790757656097412, | |
| "learning_rate": 0.00022574999999999996, | |
| "loss": 0.7808, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.4144271570014144, | |
| "eval_loss": 0.5279428958892822, | |
| "eval_runtime": 152.9768, | |
| "eval_samples_per_second": 36.973, | |
| "eval_steps_per_second": 4.622, | |
| "eval_wer": 0.4291084769080642, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5558698727015559, | |
| "eval_loss": 0.5024306178092957, | |
| "eval_runtime": 152.8927, | |
| "eval_samples_per_second": 36.993, | |
| "eval_steps_per_second": 4.624, | |
| "eval_wer": 0.39940170529986724, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.6973125884016973, | |
| "eval_loss": 0.4894837439060211, | |
| "eval_runtime": 153.3103, | |
| "eval_samples_per_second": 36.892, | |
| "eval_steps_per_second": 4.612, | |
| "eval_wer": 0.3894514565436483, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.8387553041018387, | |
| "eval_loss": 0.4595918357372284, | |
| "eval_runtime": 153.0651, | |
| "eval_samples_per_second": 36.952, | |
| "eval_steps_per_second": 4.619, | |
| "eval_wer": 0.3695829534002016, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.9801980198019802, | |
| "eval_loss": 0.44729524850845337, | |
| "eval_runtime": 154.1998, | |
| "eval_samples_per_second": 36.68, | |
| "eval_steps_per_second": 4.585, | |
| "eval_wer": 0.3610884484330758, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.1216407355021216, | |
| "grad_norm": 0.8205087184906006, | |
| "learning_rate": 0.0001512, | |
| "loss": 0.6005, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.1216407355021216, | |
| "eval_loss": 0.43324384093284607, | |
| "eval_runtime": 150.9169, | |
| "eval_samples_per_second": 37.478, | |
| "eval_steps_per_second": 4.685, | |
| "eval_wer": 0.3474268528738942, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.263083451202263, | |
| "eval_loss": 0.4268616735935211, | |
| "eval_runtime": 152.3911, | |
| "eval_samples_per_second": 37.115, | |
| "eval_steps_per_second": 4.639, | |
| "eval_wer": 0.3418118411159636, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.4045261669024045, | |
| "eval_loss": 0.4155045449733734, | |
| "eval_runtime": 153.4832, | |
| "eval_samples_per_second": 36.851, | |
| "eval_steps_per_second": 4.606, | |
| "eval_wer": 0.33606885188206875, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.545968882602546, | |
| "eval_loss": 0.4121190905570984, | |
| "eval_runtime": 153.7529, | |
| "eval_samples_per_second": 36.786, | |
| "eval_steps_per_second": 4.598, | |
| "eval_wer": 0.32143142806865993, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.6874115983026874, | |
| "eval_loss": 0.4145391285419464, | |
| "eval_runtime": 159.3319, | |
| "eval_samples_per_second": 35.498, | |
| "eval_steps_per_second": 4.437, | |
| "eval_wer": 0.3366447505239078, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.828854314002829, | |
| "grad_norm": 0.8615767359733582, | |
| "learning_rate": 7.664999999999999e-05, | |
| "loss": 0.4666, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.828854314002829, | |
| "eval_loss": 0.39387884736061096, | |
| "eval_runtime": 153.1343, | |
| "eval_samples_per_second": 36.935, | |
| "eval_steps_per_second": 4.617, | |
| "eval_wer": 0.3114171905744589, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.9702970297029703, | |
| "eval_loss": 0.38894009590148926, | |
| "eval_runtime": 152.8789, | |
| "eval_samples_per_second": 36.997, | |
| "eval_steps_per_second": 4.625, | |
| "eval_wer": 0.30807377901489336, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.1117397454031117, | |
| "eval_loss": 0.3909347653388977, | |
| "eval_runtime": 154.4919, | |
| "eval_samples_per_second": 36.61, | |
| "eval_steps_per_second": 4.576, | |
| "eval_wer": 0.30644206619634945, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.253182461103253, | |
| "eval_loss": 0.3874327838420868, | |
| "eval_runtime": 153.3025, | |
| "eval_samples_per_second": 36.894, | |
| "eval_steps_per_second": 4.612, | |
| "eval_wer": 0.3015469277407176, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.3946251768033946, | |
| "eval_loss": 0.386868953704834, | |
| "eval_runtime": 153.0443, | |
| "eval_samples_per_second": 36.957, | |
| "eval_steps_per_second": 4.62, | |
| "eval_wer": 0.29833149365711636, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.536067892503536, | |
| "grad_norm": 0.6678842902183533, | |
| "learning_rate": 1.9499999999999995e-06, | |
| "loss": 0.3805, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.536067892503536, | |
| "eval_loss": 0.3846580684185028, | |
| "eval_runtime": 154.6331, | |
| "eval_samples_per_second": 36.577, | |
| "eval_steps_per_second": 4.572, | |
| "eval_wer": 0.29666778646958136, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.536067892503536, | |
| "step": 2500, | |
| "total_flos": 9.55129266706546e+18, | |
| "train_loss": 1.1546670959472656, | |
| "train_runtime": 6557.446, | |
| "train_samples_per_second": 12.2, | |
| "train_steps_per_second": 0.381 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 400, | |
| "total_flos": 9.55129266706546e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |