| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.013162988772745, | |
| "eval_steps": 100, | |
| "global_step": 2600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07742934572202866, | |
| "eval_loss": 3.5345706939697266, | |
| "eval_runtime": 152.1587, | |
| "eval_samples_per_second": 37.172, | |
| "eval_steps_per_second": 4.646, | |
| "eval_wer": 1.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1548586914440573, | |
| "eval_loss": 2.982929229736328, | |
| "eval_runtime": 150.2931, | |
| "eval_samples_per_second": 37.633, | |
| "eval_steps_per_second": 4.704, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.23228803716608595, | |
| "eval_loss": 2.770493268966675, | |
| "eval_runtime": 150.0834, | |
| "eval_samples_per_second": 37.686, | |
| "eval_steps_per_second": 4.711, | |
| "eval_wer": 1.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3097173828881146, | |
| "eval_loss": 1.3696156740188599, | |
| "eval_runtime": 150.2358, | |
| "eval_samples_per_second": 37.647, | |
| "eval_steps_per_second": 4.706, | |
| "eval_wer": 0.8535090112500201, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.38714672861014326, | |
| "grad_norm": 3.0665018558502197, | |
| "learning_rate": 0.00029699999999999996, | |
| "loss": 3.7305, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.38714672861014326, | |
| "eval_loss": 1.0936249494552612, | |
| "eval_runtime": 153.8749, | |
| "eval_samples_per_second": 36.757, | |
| "eval_steps_per_second": 4.595, | |
| "eval_wer": 0.7465295052237968, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4645760743321719, | |
| "eval_loss": 0.8456823825836182, | |
| "eval_runtime": 153.7133, | |
| "eval_samples_per_second": 36.796, | |
| "eval_steps_per_second": 4.599, | |
| "eval_wer": 0.6413313861116015, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5420054200542005, | |
| "eval_loss": 0.7860042452812195, | |
| "eval_runtime": 152.7402, | |
| "eval_samples_per_second": 37.03, | |
| "eval_steps_per_second": 4.629, | |
| "eval_wer": 0.5835566753863684, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.6194347657762292, | |
| "eval_loss": 0.7366299629211426, | |
| "eval_runtime": 151.7301, | |
| "eval_samples_per_second": 37.277, | |
| "eval_steps_per_second": 4.66, | |
| "eval_wer": 0.563736739901462, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.6968641114982579, | |
| "eval_loss": 0.7318999171257019, | |
| "eval_runtime": 152.5414, | |
| "eval_samples_per_second": 37.078, | |
| "eval_steps_per_second": 4.635, | |
| "eval_wer": 0.5493572563431818, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.7742934572202865, | |
| "grad_norm": 2.5090723037719727, | |
| "learning_rate": 0.00022928571428571426, | |
| "loss": 0.7504, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7742934572202865, | |
| "eval_loss": 0.6438552737236023, | |
| "eval_runtime": 151.1771, | |
| "eval_samples_per_second": 37.413, | |
| "eval_steps_per_second": 4.677, | |
| "eval_wer": 0.5104074721959204, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.8517228029423152, | |
| "eval_loss": 0.6213911175727844, | |
| "eval_runtime": 151.6033, | |
| "eval_samples_per_second": 37.308, | |
| "eval_steps_per_second": 4.663, | |
| "eval_wer": 0.4759191795991077, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.9291521486643438, | |
| "eval_loss": 0.5957211852073669, | |
| "eval_runtime": 152.1457, | |
| "eval_samples_per_second": 37.175, | |
| "eval_steps_per_second": 4.647, | |
| "eval_wer": 0.4628396270321452, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.0065814943863725, | |
| "eval_loss": 0.5717456340789795, | |
| "eval_runtime": 152.0809, | |
| "eval_samples_per_second": 37.191, | |
| "eval_steps_per_second": 4.649, | |
| "eval_wer": 0.43531639678387446, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.084010840108401, | |
| "eval_loss": 0.549981951713562, | |
| "eval_runtime": 151.522, | |
| "eval_samples_per_second": 37.328, | |
| "eval_steps_per_second": 4.666, | |
| "eval_wer": 0.4192197204345942, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.1614401858304297, | |
| "grad_norm": 0.4617447853088379, | |
| "learning_rate": 0.00015799999999999996, | |
| "loss": 0.5571, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.1614401858304297, | |
| "eval_loss": 0.5342110395431519, | |
| "eval_runtime": 152.7132, | |
| "eval_samples_per_second": 37.037, | |
| "eval_steps_per_second": 4.63, | |
| "eval_wer": 0.4073277591436504, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.2388695315524583, | |
| "eval_loss": 0.5206533670425415, | |
| "eval_runtime": 151.157, | |
| "eval_samples_per_second": 37.418, | |
| "eval_steps_per_second": 4.677, | |
| "eval_wer": 0.4023687631397346, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.316298877274487, | |
| "eval_loss": 0.5142083168029785, | |
| "eval_runtime": 151.9132, | |
| "eval_samples_per_second": 37.232, | |
| "eval_steps_per_second": 4.654, | |
| "eval_wer": 0.3968641170900804, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.3937282229965158, | |
| "eval_loss": 0.5083270072937012, | |
| "eval_runtime": 152.7068, | |
| "eval_samples_per_second": 37.038, | |
| "eval_steps_per_second": 4.63, | |
| "eval_wer": 0.39583701112163183, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.4711575687185443, | |
| "eval_loss": 0.4886321723461151, | |
| "eval_runtime": 152.3343, | |
| "eval_samples_per_second": 37.129, | |
| "eval_steps_per_second": 4.641, | |
| "eval_wer": 0.3825488276548282, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.5485869144405728, | |
| "grad_norm": 0.441532164812088, | |
| "learning_rate": 8.685714285714285e-05, | |
| "loss": 0.4603, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.5485869144405728, | |
| "eval_loss": 0.4732557237148285, | |
| "eval_runtime": 155.1846, | |
| "eval_samples_per_second": 36.447, | |
| "eval_steps_per_second": 4.556, | |
| "eval_wer": 0.3743480284379965, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.6260162601626016, | |
| "eval_loss": 0.4615860879421234, | |
| "eval_runtime": 155.1621, | |
| "eval_samples_per_second": 36.452, | |
| "eval_steps_per_second": 4.557, | |
| "eval_wer": 0.3618622715090433, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.7034456058846303, | |
| "eval_loss": 0.4535791277885437, | |
| "eval_runtime": 152.4739, | |
| "eval_samples_per_second": 37.095, | |
| "eval_steps_per_second": 4.637, | |
| "eval_wer": 0.3627288921699218, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.7808749516066589, | |
| "eval_loss": 0.44882732629776, | |
| "eval_runtime": 153.2132, | |
| "eval_samples_per_second": 36.916, | |
| "eval_steps_per_second": 4.614, | |
| "eval_wer": 0.3487185248190528, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.8583042973286876, | |
| "eval_loss": 0.4429319202899933, | |
| "eval_runtime": 154.0199, | |
| "eval_samples_per_second": 36.723, | |
| "eval_steps_per_second": 4.59, | |
| "eval_wer": 0.34810868065028644, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.9357336430507162, | |
| "grad_norm": 0.48590919375419617, | |
| "learning_rate": 1.557142857142857e-05, | |
| "loss": 0.4163, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.9357336430507162, | |
| "eval_loss": 0.4377308487892151, | |
| "eval_runtime": 154.6322, | |
| "eval_samples_per_second": 36.577, | |
| "eval_steps_per_second": 4.572, | |
| "eval_wer": 0.3418978992473239, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.013162988772745, | |
| "eval_loss": 0.4348967373371124, | |
| "eval_runtime": 154.2249, | |
| "eval_samples_per_second": 36.674, | |
| "eval_steps_per_second": 4.584, | |
| "eval_wer": 0.3390733578340903, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.013162988772745, | |
| "step": 2600, | |
| "total_flos": 1.1633941226063049e+19, | |
| "train_loss": 1.1519982103201059, | |
| "train_runtime": 7259.2244, | |
| "train_samples_per_second": 11.461, | |
| "train_steps_per_second": 0.358 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2600, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 400, | |
| "total_flos": 1.1633941226063049e+19, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |