| { | |
| "best_global_step": 600, | |
| "best_metric": 44.61419121291129, | |
| "best_model_checkpoint": "./HAUSA_B/checkpoint-600", | |
| "epoch": 0.8124576844955992, | |
| "eval_steps": 200, | |
| "global_step": 600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.027081922816519974, | |
| "grad_norm": 106.59215545654297, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 6.1915, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05416384563303995, | |
| "grad_norm": 30.563518524169922, | |
| "learning_rate": 1.8000000000000001e-06, | |
| "loss": 4.5886, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.08124576844955991, | |
| "grad_norm": 16.528457641601562, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 3.0338, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1083276912660799, | |
| "grad_norm": 13.852096557617188, | |
| "learning_rate": 3.8000000000000005e-06, | |
| "loss": 2.3188, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.13540961408259986, | |
| "grad_norm": 13.284646987915039, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 1.9725, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.16249153689911983, | |
| "grad_norm": 13.212055206298828, | |
| "learning_rate": 5.8e-06, | |
| "loss": 1.7839, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.1895734597156398, | |
| "grad_norm": 12.006990432739258, | |
| "learning_rate": 6.800000000000001e-06, | |
| "loss": 1.5991, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.2166553825321598, | |
| "grad_norm": 12.490514755249023, | |
| "learning_rate": 7.800000000000002e-06, | |
| "loss": 1.4411, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.24373730534867977, | |
| "grad_norm": 11.587646484375, | |
| "learning_rate": 8.8e-06, | |
| "loss": 1.3487, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2708192281651997, | |
| "grad_norm": 11.574933052062988, | |
| "learning_rate": 9.800000000000001e-06, | |
| "loss": 1.2119, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2708192281651997, | |
| "eval_loss": 1.1157827377319336, | |
| "eval_runtime": 1601.2991, | |
| "eval_samples_per_second": 3.688, | |
| "eval_steps_per_second": 0.462, | |
| "eval_wer": 64.22461160550755, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2979011509817197, | |
| "grad_norm": 8.368565559387207, | |
| "learning_rate": 9.920556107249256e-06, | |
| "loss": 0.9452, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.32498307379823965, | |
| "grad_norm": 7.2616682052612305, | |
| "learning_rate": 9.821251241310825e-06, | |
| "loss": 0.9298, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.35206499661475965, | |
| "grad_norm": 7.266394138336182, | |
| "learning_rate": 9.721946375372395e-06, | |
| "loss": 0.8567, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.3791469194312796, | |
| "grad_norm": 8.120067596435547, | |
| "learning_rate": 9.622641509433963e-06, | |
| "loss": 0.8251, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.4062288422477996, | |
| "grad_norm": 7.483547687530518, | |
| "learning_rate": 9.523336643495532e-06, | |
| "loss": 0.7774, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4333107650643196, | |
| "grad_norm": 7.521896839141846, | |
| "learning_rate": 9.4240317775571e-06, | |
| "loss": 0.7266, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.46039268788083954, | |
| "grad_norm": 7.073266506195068, | |
| "learning_rate": 9.32472691161867e-06, | |
| "loss": 0.7248, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.48747461069735953, | |
| "grad_norm": 6.335423469543457, | |
| "learning_rate": 9.22542204568024e-06, | |
| "loss": 0.7151, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5145565335138795, | |
| "grad_norm": 6.936922550201416, | |
| "learning_rate": 9.126117179741808e-06, | |
| "loss": 0.6881, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.5416384563303994, | |
| "grad_norm": 7.596807479858398, | |
| "learning_rate": 9.026812313803377e-06, | |
| "loss": 0.6995, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5416384563303994, | |
| "eval_loss": 0.6352065801620483, | |
| "eval_runtime": 1597.3635, | |
| "eval_samples_per_second": 3.697, | |
| "eval_steps_per_second": 0.463, | |
| "eval_wer": 51.63653601672089, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5687203791469194, | |
| "grad_norm": 6.981812477111816, | |
| "learning_rate": 8.927507447864945e-06, | |
| "loss": 0.6795, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.5958023019634394, | |
| "grad_norm": 6.481506824493408, | |
| "learning_rate": 8.828202581926516e-06, | |
| "loss": 0.6643, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6228842247799594, | |
| "grad_norm": 6.537086009979248, | |
| "learning_rate": 8.728897715988084e-06, | |
| "loss": 0.6459, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.6499661475964793, | |
| "grad_norm": 6.739567756652832, | |
| "learning_rate": 8.629592850049653e-06, | |
| "loss": 0.6736, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.6770480704129993, | |
| "grad_norm": 7.422546863555908, | |
| "learning_rate": 8.530287984111221e-06, | |
| "loss": 0.6591, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7041299932295193, | |
| "grad_norm": 5.7051215171813965, | |
| "learning_rate": 8.430983118172792e-06, | |
| "loss": 0.6442, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.7312119160460393, | |
| "grad_norm": 7.166143417358398, | |
| "learning_rate": 8.33167825223436e-06, | |
| "loss": 0.6326, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.7582938388625592, | |
| "grad_norm": 7.759460926055908, | |
| "learning_rate": 8.232373386295929e-06, | |
| "loss": 0.6328, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.7853757616790792, | |
| "grad_norm": 5.876537799835205, | |
| "learning_rate": 8.133068520357497e-06, | |
| "loss": 0.5945, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.8124576844955992, | |
| "grad_norm": 6.475106716156006, | |
| "learning_rate": 8.033763654419066e-06, | |
| "loss": 0.6038, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.8124576844955992, | |
| "eval_loss": 0.5318673849105835, | |
| "eval_runtime": 1614.0202, | |
| "eval_samples_per_second": 3.659, | |
| "eval_steps_per_second": 0.458, | |
| "eval_wer": 44.61419121291129, | |
| "step": 600 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 2214, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.540839686144e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |