| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 8.936550491510276, | |
| "eval_steps": 200, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 2.9598, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 2.3436, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "eval_loss": 1.8791261911392212, | |
| "eval_runtime": 1327.8534, | |
| "eval_samples_per_second": 0.753, | |
| "eval_steps_per_second": 0.047, | |
| "eval_wer": 0.8870904221802143, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 6e-06, | |
| "loss": 1.5788, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.1682, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "eval_loss": 1.0307379961013794, | |
| "eval_runtime": 1238.5009, | |
| "eval_samples_per_second": 0.807, | |
| "eval_steps_per_second": 0.051, | |
| "eval_wer": 0.5047652804032766, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9354, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 9.987820251299121e-06, | |
| "loss": 0.7321, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "eval_loss": 0.6299881935119629, | |
| "eval_runtime": 830.7332, | |
| "eval_samples_per_second": 1.204, | |
| "eval_steps_per_second": 0.076, | |
| "eval_wer": 0.36645400126023947, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 9.951340343707852e-06, | |
| "loss": 0.5397, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 9.890738003669029e-06, | |
| "loss": 0.4564, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "eval_loss": 0.438092440366745, | |
| "eval_runtime": 723.2192, | |
| "eval_samples_per_second": 1.383, | |
| "eval_steps_per_second": 0.087, | |
| "eval_wer": 0.35148865784499056, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "learning_rate": 9.806308479691595e-06, | |
| "loss": 0.4291, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.4095, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "eval_loss": 0.40272918343544006, | |
| "eval_runtime": 636.5063, | |
| "eval_samples_per_second": 1.571, | |
| "eval_steps_per_second": 0.099, | |
| "eval_wer": 0.33297889098928796, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "learning_rate": 9.567727288213005e-06, | |
| "loss": 0.3992, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 9.414737964294636e-06, | |
| "loss": 0.3813, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "eval_loss": 0.3847169280052185, | |
| "eval_runtime": 616.7138, | |
| "eval_samples_per_second": 1.621, | |
| "eval_steps_per_second": 0.102, | |
| "eval_wer": 0.3359719596723377, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 9.24024048078213e-06, | |
| "loss": 0.3711, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.3667, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "eval_loss": 0.37336310744285583, | |
| "eval_runtime": 614.3217, | |
| "eval_samples_per_second": 1.628, | |
| "eval_steps_per_second": 0.103, | |
| "eval_wer": 0.33916194076874606, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.3654, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "learning_rate": 8.596699001693257e-06, | |
| "loss": 0.3583, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "eval_loss": 0.3648846447467804, | |
| "eval_runtime": 617.5084, | |
| "eval_samples_per_second": 1.619, | |
| "eval_steps_per_second": 0.102, | |
| "eval_wer": 0.34904694391934465, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "learning_rate": 8.345653031794292e-06, | |
| "loss": 0.353, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "learning_rate": 8.078307376628292e-06, | |
| "loss": 0.3454, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "eval_loss": 0.35879915952682495, | |
| "eval_runtime": 620.7465, | |
| "eval_samples_per_second": 1.611, | |
| "eval_steps_per_second": 0.101, | |
| "eval_wer": 0.35715973534971646, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "learning_rate": 7.795964517353734e-06, | |
| "loss": 0.3399, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.3422, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "eval_loss": 0.3536700904369354, | |
| "eval_runtime": 625.9316, | |
| "eval_samples_per_second": 1.598, | |
| "eval_steps_per_second": 0.101, | |
| "eval_wer": 0.3704710144927536, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "learning_rate": 7.191855733945388e-06, | |
| "loss": 0.3355, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "learning_rate": 6.873032967079562e-06, | |
| "loss": 0.3371, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "eval_loss": 0.3503468930721283, | |
| "eval_runtime": 634.2679, | |
| "eval_samples_per_second": 1.577, | |
| "eval_steps_per_second": 0.099, | |
| "eval_wer": 0.3811436672967864, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.3311, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "learning_rate": 6.209609477998339e-06, | |
| "loss": 0.3291, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "eval_loss": 0.347513347864151, | |
| "eval_runtime": 615.3576, | |
| "eval_samples_per_second": 1.625, | |
| "eval_steps_per_second": 0.102, | |
| "eval_wer": 0.3677930056710775, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 4.47, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.3273, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.65, | |
| "learning_rate": 5.522642316338268e-06, | |
| "loss": 0.324, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.65, | |
| "eval_loss": 0.3451197147369385, | |
| "eval_runtime": 628.249, | |
| "eval_samples_per_second": 1.592, | |
| "eval_steps_per_second": 0.1, | |
| "eval_wer": 0.36700535601764334, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.83, | |
| "learning_rate": 5.174497483512506e-06, | |
| "loss": 0.3219, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 4.825502516487497e-06, | |
| "loss": 0.3262, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.34306031465530396, | |
| "eval_runtime": 613.0254, | |
| "eval_samples_per_second": 1.631, | |
| "eval_steps_per_second": 0.103, | |
| "eval_wer": 0.3710223692501575, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 5.18, | |
| "learning_rate": 4.477357683661734e-06, | |
| "loss": 0.3203, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 5.36, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.3168, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 5.36, | |
| "eval_loss": 0.34187400341033936, | |
| "eval_runtime": 628.864, | |
| "eval_samples_per_second": 1.59, | |
| "eval_steps_per_second": 0.1, | |
| "eval_wer": 0.3847274732199118, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 5.54, | |
| "learning_rate": 3.790390522001662e-06, | |
| "loss": 0.3144, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 5.72, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.3178, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 5.72, | |
| "eval_loss": 0.34061843156814575, | |
| "eval_runtime": 627.3787, | |
| "eval_samples_per_second": 1.594, | |
| "eval_steps_per_second": 0.1, | |
| "eval_wer": 0.3832703213610586, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 5.9, | |
| "learning_rate": 3.12696703292044e-06, | |
| "loss": 0.3127, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 6.08, | |
| "learning_rate": 2.8081442660546126e-06, | |
| "loss": 0.3136, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 6.08, | |
| "eval_loss": 0.34004053473472595, | |
| "eval_runtime": 597.4949, | |
| "eval_samples_per_second": 1.674, | |
| "eval_steps_per_second": 0.105, | |
| "eval_wer": 0.3853182104599874, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 6.26, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.3126, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 6.43, | |
| "learning_rate": 2.204035482646267e-06, | |
| "loss": 0.3092, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 6.43, | |
| "eval_loss": 0.3392544984817505, | |
| "eval_runtime": 603.3789, | |
| "eval_samples_per_second": 1.657, | |
| "eval_steps_per_second": 0.104, | |
| "eval_wer": 0.38961090107120355, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 6.61, | |
| "learning_rate": 1.9216926233717087e-06, | |
| "loss": 0.3135, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 6.79, | |
| "learning_rate": 1.6543469682057105e-06, | |
| "loss": 0.3106, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 6.79, | |
| "eval_loss": 0.33891087770462036, | |
| "eval_runtime": 587.9522, | |
| "eval_samples_per_second": 1.701, | |
| "eval_steps_per_second": 0.107, | |
| "eval_wer": 0.3900047258979206, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 6.97, | |
| "learning_rate": 1.4033009983067454e-06, | |
| "loss": 0.3094, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 7.15, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.3057, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 7.15, | |
| "eval_loss": 0.33877310156822205, | |
| "eval_runtime": 590.6741, | |
| "eval_samples_per_second": 1.693, | |
| "eval_steps_per_second": 0.107, | |
| "eval_wer": 0.38031663516068054, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 7.33, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.3079, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 7.51, | |
| "learning_rate": 7.597595192178702e-07, | |
| "loss": 0.3087, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 7.51, | |
| "eval_loss": 0.33828216791152954, | |
| "eval_runtime": 586.9517, | |
| "eval_samples_per_second": 1.704, | |
| "eval_steps_per_second": 0.107, | |
| "eval_wer": 0.39406112161310647, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "learning_rate": 5.852620357053651e-07, | |
| "loss": 0.3082, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 7.86, | |
| "learning_rate": 4.322727117869951e-07, | |
| "loss": 0.308, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 7.86, | |
| "eval_loss": 0.3381615877151489, | |
| "eval_runtime": 590.3123, | |
| "eval_samples_per_second": 1.694, | |
| "eval_steps_per_second": 0.107, | |
| "eval_wer": 0.3873660995589162, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 8.04, | |
| "learning_rate": 3.015368960704584e-07, | |
| "loss": 0.3091, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 8.22, | |
| "learning_rate": 1.9369152030840553e-07, | |
| "loss": 0.3036, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 8.22, | |
| "eval_loss": 0.33812272548675537, | |
| "eval_runtime": 590.6229, | |
| "eval_samples_per_second": 1.693, | |
| "eval_steps_per_second": 0.107, | |
| "eval_wer": 0.38961090107120355, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "learning_rate": 1.0926199633097156e-07, | |
| "loss": 0.3049, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 8.58, | |
| "learning_rate": 4.865965629214819e-08, | |
| "loss": 0.3087, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 8.58, | |
| "eval_loss": 0.338046669960022, | |
| "eval_runtime": 593.4841, | |
| "eval_samples_per_second": 1.685, | |
| "eval_steps_per_second": 0.106, | |
| "eval_wer": 0.3909892879647133, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 8.76, | |
| "learning_rate": 1.2179748700879013e-08, | |
| "loss": 0.3092, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 8.94, | |
| "learning_rate": 0.0, | |
| "loss": 0.3079, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 8.94, | |
| "eval_loss": 0.3380787670612335, | |
| "eval_runtime": 604.744, | |
| "eval_samples_per_second": 1.654, | |
| "eval_steps_per_second": 0.104, | |
| "eval_wer": 0.38886263390044107, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 8.94, | |
| "step": 5000, | |
| "total_flos": 1.8645896628338688e+20, | |
| "train_loss": 0.4928088363647461, | |
| "train_runtime": 87940.1228, | |
| "train_samples_per_second": 7.278, | |
| "train_steps_per_second": 0.057 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 5000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9, | |
| "save_steps": 200, | |
| "total_flos": 1.8645896628338688e+20, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |