| { | |
| "best_metric": 48.88501742160279, | |
| "best_model_checkpoint": "./whisper-small-ar/checkpoint-60", | |
| "epoch": 34.285714285714285, | |
| "eval_steps": 10, | |
| "global_step": 60, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 58.168914794921875, | |
| "learning_rate": 1e-07, | |
| "loss": 2.3005, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 20.130443572998047, | |
| "learning_rate": 1.0080000000000002e-05, | |
| "loss": 1.829, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 21.1949405670166, | |
| "learning_rate": 2.006e-05, | |
| "loss": 1.8278, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 2.2857142857142856, | |
| "grad_norm": 17.402198791503906, | |
| "learning_rate": 3.0039999999999997e-05, | |
| "loss": 1.6203, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 8.070812225341797, | |
| "learning_rate": 4.0020000000000006e-05, | |
| "loss": 1.3838, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 3.4285714285714284, | |
| "grad_norm": 10.189924240112305, | |
| "learning_rate": 5e-05, | |
| "loss": 1.2624, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 11.481833457946777, | |
| "learning_rate": 4.9744102564102566e-05, | |
| "loss": 1.1801, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 4.571428571428571, | |
| "grad_norm": 8.908258438110352, | |
| "learning_rate": 4.948820512820513e-05, | |
| "loss": 1.036, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 5.142857142857143, | |
| "grad_norm": 6.949155807495117, | |
| "learning_rate": 4.92323076923077e-05, | |
| "loss": 0.9218, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "grad_norm": 5.213706016540527, | |
| "learning_rate": 4.897641025641026e-05, | |
| "loss": 0.821, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "eval_loss": 1.0209691524505615, | |
| "eval_runtime": 36.4465, | |
| "eval_samples_per_second": 2.771, | |
| "eval_steps_per_second": 0.192, | |
| "eval_wer": 93.58885017421603, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 6.285714285714286, | |
| "grad_norm": 5.321272850036621, | |
| "learning_rate": 4.872051282051282e-05, | |
| "loss": 0.7761, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 6.857142857142857, | |
| "grad_norm": 4.693880081176758, | |
| "learning_rate": 4.8464615384615386e-05, | |
| "loss": 0.7099, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 7.428571428571429, | |
| "grad_norm": 4.722415924072266, | |
| "learning_rate": 4.820871794871795e-05, | |
| "loss": 0.6289, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 3.9184811115264893, | |
| "learning_rate": 4.795282051282052e-05, | |
| "loss": 0.5725, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 8.571428571428571, | |
| "grad_norm": 3.7839274406433105, | |
| "learning_rate": 4.7696923076923084e-05, | |
| "loss": 0.5235, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 9.142857142857142, | |
| "grad_norm": 4.117933750152588, | |
| "learning_rate": 4.744102564102564e-05, | |
| "loss": 0.4812, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 9.714285714285714, | |
| "grad_norm": 3.781254529953003, | |
| "learning_rate": 4.7185128205128205e-05, | |
| "loss": 0.4397, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 10.285714285714286, | |
| "grad_norm": 3.4398269653320312, | |
| "learning_rate": 4.692923076923077e-05, | |
| "loss": 0.3749, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 10.857142857142858, | |
| "grad_norm": 3.5433449745178223, | |
| "learning_rate": 4.667333333333334e-05, | |
| "loss": 0.3506, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 11.428571428571429, | |
| "grad_norm": 3.3172426223754883, | |
| "learning_rate": 4.64174358974359e-05, | |
| "loss": 0.3185, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 11.428571428571429, | |
| "eval_loss": 0.83645099401474, | |
| "eval_runtime": 27.2985, | |
| "eval_samples_per_second": 3.7, | |
| "eval_steps_per_second": 0.256, | |
| "eval_wer": 109.26829268292684, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 3.2160873413085938, | |
| "learning_rate": 4.616153846153847e-05, | |
| "loss": 0.2779, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 12.571428571428571, | |
| "grad_norm": 3.265079975128174, | |
| "learning_rate": 4.5905641025641024e-05, | |
| "loss": 0.2553, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 13.142857142857142, | |
| "grad_norm": 3.1832408905029297, | |
| "learning_rate": 4.564974358974359e-05, | |
| "loss": 0.2157, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 13.714285714285714, | |
| "grad_norm": 2.9767119884490967, | |
| "learning_rate": 4.539384615384616e-05, | |
| "loss": 0.1921, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 14.285714285714286, | |
| "grad_norm": 2.942561388015747, | |
| "learning_rate": 4.513794871794872e-05, | |
| "loss": 0.1741, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 14.857142857142858, | |
| "grad_norm": 2.837989091873169, | |
| "learning_rate": 4.4882051282051286e-05, | |
| "loss": 0.1422, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 15.428571428571429, | |
| "grad_norm": 2.894479513168335, | |
| "learning_rate": 4.462615384615385e-05, | |
| "loss": 0.1257, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 2.841754198074341, | |
| "learning_rate": 4.437025641025641e-05, | |
| "loss": 0.1039, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 16.571428571428573, | |
| "grad_norm": 2.616755485534668, | |
| "learning_rate": 4.411435897435898e-05, | |
| "loss": 0.0847, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 17.142857142857142, | |
| "grad_norm": 2.05346417427063, | |
| "learning_rate": 4.385846153846154e-05, | |
| "loss": 0.0606, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 17.142857142857142, | |
| "eval_loss": 0.8186278939247131, | |
| "eval_runtime": 26.7925, | |
| "eval_samples_per_second": 3.77, | |
| "eval_steps_per_second": 0.261, | |
| "eval_wer": 54.843205574912886, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 17.714285714285715, | |
| "grad_norm": 1.4241214990615845, | |
| "learning_rate": 4.3602564102564106e-05, | |
| "loss": 0.051, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 18.285714285714285, | |
| "grad_norm": 0.806209921836853, | |
| "learning_rate": 4.334666666666667e-05, | |
| "loss": 0.0336, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 18.857142857142858, | |
| "grad_norm": 0.7847088575363159, | |
| "learning_rate": 4.309076923076923e-05, | |
| "loss": 0.0282, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 19.428571428571427, | |
| "grad_norm": 0.9520515203475952, | |
| "learning_rate": 4.28348717948718e-05, | |
| "loss": 0.0204, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.37412479519844055, | |
| "learning_rate": 4.257897435897436e-05, | |
| "loss": 0.0163, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 20.571428571428573, | |
| "grad_norm": 0.3346174657344818, | |
| "learning_rate": 4.2323076923076925e-05, | |
| "loss": 0.0122, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 21.142857142857142, | |
| "grad_norm": 0.23975922167301178, | |
| "learning_rate": 4.206717948717949e-05, | |
| "loss": 0.0105, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 21.714285714285715, | |
| "grad_norm": 0.2158885896205902, | |
| "learning_rate": 4.181128205128205e-05, | |
| "loss": 0.0078, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 22.285714285714285, | |
| "grad_norm": 0.15689243376255035, | |
| "learning_rate": 4.155538461538462e-05, | |
| "loss": 0.0065, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 22.857142857142858, | |
| "grad_norm": 0.10141926258802414, | |
| "learning_rate": 4.129948717948718e-05, | |
| "loss": 0.0056, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 22.857142857142858, | |
| "eval_loss": 0.9213815927505493, | |
| "eval_runtime": 26.8113, | |
| "eval_samples_per_second": 3.767, | |
| "eval_steps_per_second": 0.261, | |
| "eval_wer": 50.139372822299656, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 23.428571428571427, | |
| "grad_norm": 0.07463113218545914, | |
| "learning_rate": 4.1043589743589744e-05, | |
| "loss": 0.0047, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 0.09193433821201324, | |
| "learning_rate": 4.078769230769231e-05, | |
| "loss": 0.0041, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 24.571428571428573, | |
| "grad_norm": 0.06787554919719696, | |
| "learning_rate": 4.053179487179487e-05, | |
| "loss": 0.0035, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 25.142857142857142, | |
| "grad_norm": 0.07034426182508469, | |
| "learning_rate": 4.027589743589744e-05, | |
| "loss": 0.0031, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 25.714285714285715, | |
| "grad_norm": 0.06088101118803024, | |
| "learning_rate": 4.0020000000000006e-05, | |
| "loss": 0.0028, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 26.285714285714285, | |
| "grad_norm": 0.03634655103087425, | |
| "learning_rate": 3.9764102564102564e-05, | |
| "loss": 0.0024, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 26.857142857142858, | |
| "grad_norm": 0.13105067610740662, | |
| "learning_rate": 3.950820512820513e-05, | |
| "loss": 0.0022, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 27.428571428571427, | |
| "grad_norm": 0.02541457489132881, | |
| "learning_rate": 3.925230769230769e-05, | |
| "loss": 0.002, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "grad_norm": 0.04503984376788139, | |
| "learning_rate": 3.899641025641026e-05, | |
| "loss": 0.0017, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 28.571428571428573, | |
| "grad_norm": 0.08539522439241409, | |
| "learning_rate": 3.8740512820512826e-05, | |
| "loss": 0.0019, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 28.571428571428573, | |
| "eval_loss": 1.0016592741012573, | |
| "eval_runtime": 26.5804, | |
| "eval_samples_per_second": 3.8, | |
| "eval_steps_per_second": 0.263, | |
| "eval_wer": 50.87108013937283, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 29.142857142857142, | |
| "grad_norm": 0.06402863562107086, | |
| "learning_rate": 3.848461538461539e-05, | |
| "loss": 0.0017, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 29.714285714285715, | |
| "grad_norm": 0.016784947365522385, | |
| "learning_rate": 3.822871794871795e-05, | |
| "loss": 0.0013, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 30.285714285714285, | |
| "grad_norm": 0.030725400894880295, | |
| "learning_rate": 3.797282051282051e-05, | |
| "loss": 0.0013, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 30.857142857142858, | |
| "grad_norm": 0.015844004228711128, | |
| "learning_rate": 3.771692307692308e-05, | |
| "loss": 0.0011, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 31.428571428571427, | |
| "grad_norm": 0.015942782163619995, | |
| "learning_rate": 3.7461025641025645e-05, | |
| "loss": 0.0011, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "grad_norm": 0.18212947249412537, | |
| "learning_rate": 3.720512820512821e-05, | |
| "loss": 0.0012, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 32.57142857142857, | |
| "grad_norm": 0.016346724703907967, | |
| "learning_rate": 3.694923076923077e-05, | |
| "loss": 0.0009, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 33.142857142857146, | |
| "grad_norm": 0.010927367024123669, | |
| "learning_rate": 3.669333333333333e-05, | |
| "loss": 0.0009, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 33.714285714285715, | |
| "grad_norm": 0.013612424023449421, | |
| "learning_rate": 3.64374358974359e-05, | |
| "loss": 0.0008, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 34.285714285714285, | |
| "grad_norm": 0.10505015403032303, | |
| "learning_rate": 3.6181538461538464e-05, | |
| "loss": 0.0009, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 34.285714285714285, | |
| "eval_loss": 1.0416876077651978, | |
| "eval_runtime": 26.3087, | |
| "eval_samples_per_second": 3.839, | |
| "eval_steps_per_second": 0.266, | |
| "eval_wer": 48.88501742160279, | |
| "step": 60 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 200, | |
| "save_steps": 10, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.432671678450893e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |