| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.6775106082036775, | |
| "eval_steps": 100, | |
| "global_step": 2600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.14144271570014144, | |
| "eval_loss": 3.3999907970428467, | |
| "eval_runtime": 162.0918, | |
| "eval_samples_per_second": 34.894, | |
| "eval_steps_per_second": 4.362, | |
| "eval_wer": 1.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2828854314002829, | |
| "eval_loss": 2.9512617588043213, | |
| "eval_runtime": 159.7591, | |
| "eval_samples_per_second": 35.403, | |
| "eval_steps_per_second": 4.425, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4243281471004243, | |
| "eval_loss": 1.9453805685043335, | |
| "eval_runtime": 160.8922, | |
| "eval_samples_per_second": 35.154, | |
| "eval_steps_per_second": 4.394, | |
| "eval_wer": 0.9325480252282904, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5657708628005658, | |
| "eval_loss": 1.1412426233291626, | |
| "eval_runtime": 161.9235, | |
| "eval_samples_per_second": 34.93, | |
| "eval_steps_per_second": 4.366, | |
| "eval_wer": 0.773025629503619, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7072135785007072, | |
| "grad_norm": 2.79933500289917, | |
| "learning_rate": 0.00029699999999999996, | |
| "loss": 3.2243, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7072135785007072, | |
| "eval_loss": 0.9250076413154602, | |
| "eval_runtime": 160.568, | |
| "eval_samples_per_second": 35.225, | |
| "eval_steps_per_second": 4.403, | |
| "eval_wer": 0.67890099661376, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8486562942008486, | |
| "eval_loss": 0.8018025159835815, | |
| "eval_runtime": 160.6715, | |
| "eval_samples_per_second": 35.202, | |
| "eval_steps_per_second": 4.4, | |
| "eval_wer": 0.5982892266213028, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9900990099009901, | |
| "eval_loss": 0.7181887030601501, | |
| "eval_runtime": 160.5511, | |
| "eval_samples_per_second": 35.229, | |
| "eval_steps_per_second": 4.404, | |
| "eval_wer": 0.5408675835727239, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.1315417256011315, | |
| "eval_loss": 0.6198393106460571, | |
| "eval_runtime": 161.1454, | |
| "eval_samples_per_second": 35.099, | |
| "eval_steps_per_second": 4.387, | |
| "eval_wer": 0.4769944311598273, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.272984441301273, | |
| "eval_loss": 0.6102356910705566, | |
| "eval_runtime": 160.7517, | |
| "eval_samples_per_second": 35.185, | |
| "eval_steps_per_second": 4.398, | |
| "eval_wer": 0.47124905714881804, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.4144271570014144, | |
| "grad_norm": 1.0946769714355469, | |
| "learning_rate": 0.00022928571428571426, | |
| "loss": 0.7983, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.4144271570014144, | |
| "eval_loss": 0.5605342984199524, | |
| "eval_runtime": 160.7985, | |
| "eval_samples_per_second": 35.174, | |
| "eval_steps_per_second": 4.397, | |
| "eval_wer": 0.4425863812167996, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5558698727015559, | |
| "eval_loss": 0.5336170196533203, | |
| "eval_runtime": 160.9878, | |
| "eval_samples_per_second": 35.133, | |
| "eval_steps_per_second": 4.392, | |
| "eval_wer": 0.4162186451830335, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.6973125884016973, | |
| "eval_loss": 0.5258753299713135, | |
| "eval_runtime": 161.623, | |
| "eval_samples_per_second": 34.995, | |
| "eval_steps_per_second": 4.374, | |
| "eval_wer": 0.41161271685577183, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.8387553041018387, | |
| "eval_loss": 0.4960057735443115, | |
| "eval_runtime": 161.2649, | |
| "eval_samples_per_second": 35.073, | |
| "eval_steps_per_second": 4.384, | |
| "eval_wer": 0.3872349986358749, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.9801980198019802, | |
| "eval_loss": 0.48569169640541077, | |
| "eval_runtime": 161.7331, | |
| "eval_samples_per_second": 34.971, | |
| "eval_steps_per_second": 4.371, | |
| "eval_wer": 0.38676959124392163, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.1216407355021216, | |
| "grad_norm": 1.4740066528320312, | |
| "learning_rate": 0.0001582857142857143, | |
| "loss": 0.6274, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.1216407355021216, | |
| "eval_loss": 0.46894121170043945, | |
| "eval_runtime": 161.6698, | |
| "eval_samples_per_second": 34.985, | |
| "eval_steps_per_second": 4.373, | |
| "eval_wer": 0.36555343358315545, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.263083451202263, | |
| "eval_loss": 0.4679875373840332, | |
| "eval_runtime": 161.7805, | |
| "eval_samples_per_second": 34.961, | |
| "eval_steps_per_second": 4.37, | |
| "eval_wer": 0.3562131886825761, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.4045261669024045, | |
| "eval_loss": 0.4536493718624115, | |
| "eval_runtime": 162.2799, | |
| "eval_samples_per_second": 34.853, | |
| "eval_steps_per_second": 4.357, | |
| "eval_wer": 0.3535330840461556, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.545968882602546, | |
| "eval_loss": 0.44862428307533264, | |
| "eval_runtime": 161.7828, | |
| "eval_samples_per_second": 34.96, | |
| "eval_steps_per_second": 4.37, | |
| "eval_wer": 0.3500826499333986, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.6874115983026874, | |
| "eval_loss": 0.43955981731414795, | |
| "eval_runtime": 162.5762, | |
| "eval_samples_per_second": 34.79, | |
| "eval_steps_per_second": 4.349, | |
| "eval_wer": 0.35051596026383786, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.828854314002829, | |
| "grad_norm": 0.927208662033081, | |
| "learning_rate": 8.728571428571428e-05, | |
| "loss": 0.4939, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.828854314002829, | |
| "eval_loss": 0.4298574924468994, | |
| "eval_runtime": 162.5099, | |
| "eval_samples_per_second": 34.804, | |
| "eval_steps_per_second": 4.351, | |
| "eval_wer": 0.3340020221148754, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.9702970297029703, | |
| "eval_loss": 0.4291674494743347, | |
| "eval_runtime": 162.0566, | |
| "eval_samples_per_second": 34.901, | |
| "eval_steps_per_second": 4.363, | |
| "eval_wer": 0.33114538364012774, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.1117397454031117, | |
| "eval_loss": 0.42761147022247314, | |
| "eval_runtime": 162.7807, | |
| "eval_samples_per_second": 34.746, | |
| "eval_steps_per_second": 4.343, | |
| "eval_wer": 0.3270690568278474, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.253182461103253, | |
| "eval_loss": 0.4232628345489502, | |
| "eval_runtime": 161.9234, | |
| "eval_samples_per_second": 34.93, | |
| "eval_steps_per_second": 4.366, | |
| "eval_wer": 0.3260419508593988, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.3946251768033946, | |
| "eval_loss": 0.4192351996898651, | |
| "eval_runtime": 162.4093, | |
| "eval_samples_per_second": 34.826, | |
| "eval_steps_per_second": 4.353, | |
| "eval_wer": 0.32230264319301566, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.536067892503536, | |
| "grad_norm": 0.7725916504859924, | |
| "learning_rate": 1.614285714285714e-05, | |
| "loss": 0.4072, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.536067892503536, | |
| "eval_loss": 0.4178767800331116, | |
| "eval_runtime": 162.2884, | |
| "eval_samples_per_second": 34.852, | |
| "eval_steps_per_second": 4.356, | |
| "eval_wer": 0.31946205324902505, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.6775106082036775, | |
| "eval_loss": 0.4158582091331482, | |
| "eval_runtime": 162.2061, | |
| "eval_samples_per_second": 34.869, | |
| "eval_steps_per_second": 4.359, | |
| "eval_wer": 0.3171350162892587, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 3.6775106082036775, | |
| "step": 2600, | |
| "total_flos": 9.931497341662648e+18, | |
| "train_loss": 1.0819015429570125, | |
| "train_runtime": 7138.5232, | |
| "train_samples_per_second": 11.655, | |
| "train_steps_per_second": 0.364 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2600, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 400, | |
| "total_flos": 9.931497341662648e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |