| { | |
| "best_metric": 0.2065383642911911, | |
| "best_model_checkpoint": "./whisper-small-arabic-finetuned-on-halabi_daataset_no-diacritics-2/checkpoint-200", | |
| "epoch": 17.548672566371682, | |
| "eval_steps": 200, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.4424778761061947, | |
| "grad_norm": 56.8260383605957, | |
| "learning_rate": 1.25e-06, | |
| "loss": 1.151, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.8849557522123894, | |
| "grad_norm": 13.073741912841797, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.5879, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.3185840707964602, | |
| "grad_norm": 7.271528720855713, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.2198, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.7610619469026547, | |
| "grad_norm": 5.315987586975098, | |
| "learning_rate": 5e-06, | |
| "loss": 0.1285, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.1946902654867255, | |
| "grad_norm": 4.4957075119018555, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.0915, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.6371681415929205, | |
| "grad_norm": 5.721312999725342, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.0912, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.0707964601769913, | |
| "grad_norm": 3.7948434352874756, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.068, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 3.5132743362831858, | |
| "grad_norm": 4.056038856506348, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0426, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.5132743362831858, | |
| "eval_loss": 0.2065383642911911, | |
| "eval_runtime": 53.2112, | |
| "eval_samples_per_second": 1.861, | |
| "eval_steps_per_second": 0.132, | |
| "eval_wer": 0.24910007199424047, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.9557522123893807, | |
| "grad_norm": 2.0509068965911865, | |
| "learning_rate": 9.6875e-06, | |
| "loss": 0.0415, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 4.389380530973451, | |
| "grad_norm": 2.137742280960083, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 0.032, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.831858407079646, | |
| "grad_norm": 4.375368595123291, | |
| "learning_rate": 9.0625e-06, | |
| "loss": 0.0302, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 5.265486725663717, | |
| "grad_norm": 4.673452854156494, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.0195, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.707964601769912, | |
| "grad_norm": 1.6245851516723633, | |
| "learning_rate": 8.4375e-06, | |
| "loss": 0.0171, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 6.1415929203539825, | |
| "grad_norm": 1.5340633392333984, | |
| "learning_rate": 8.125000000000001e-06, | |
| "loss": 0.0149, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 6.584070796460177, | |
| "grad_norm": 2.1918094158172607, | |
| "learning_rate": 7.8125e-06, | |
| "loss": 0.0116, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 7.017699115044247, | |
| "grad_norm": 2.3824124336242676, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.0069, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 7.017699115044247, | |
| "eval_loss": 0.23826299607753754, | |
| "eval_runtime": 53.5165, | |
| "eval_samples_per_second": 1.85, | |
| "eval_steps_per_second": 0.131, | |
| "eval_wer": 0.2584593232541397, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 7.460176991150442, | |
| "grad_norm": 0.031165067106485367, | |
| "learning_rate": 7.1875e-06, | |
| "loss": 0.0053, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 7.902654867256637, | |
| "grad_norm": 0.7720983624458313, | |
| "learning_rate": 6.875e-06, | |
| "loss": 0.0067, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 8.336283185840708, | |
| "grad_norm": 0.6943212747573853, | |
| "learning_rate": 6.5625e-06, | |
| "loss": 0.004, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 8.778761061946902, | |
| "grad_norm": 3.529315948486328, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.0048, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 9.212389380530974, | |
| "grad_norm": 0.01847268082201481, | |
| "learning_rate": 5.9375e-06, | |
| "loss": 0.0036, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 9.654867256637168, | |
| "grad_norm": 0.5968337059020996, | |
| "learning_rate": 5.625e-06, | |
| "loss": 0.0048, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 10.08849557522124, | |
| "grad_norm": 0.6061381697654724, | |
| "learning_rate": 5.3125e-06, | |
| "loss": 0.0055, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 10.530973451327434, | |
| "grad_norm": 1.1333274841308594, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0021, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 10.530973451327434, | |
| "eval_loss": 0.24962136149406433, | |
| "eval_runtime": 52.7604, | |
| "eval_samples_per_second": 1.876, | |
| "eval_steps_per_second": 0.133, | |
| "eval_wer": 0.2735781137508999, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 10.973451327433628, | |
| "grad_norm": 0.0274394191801548, | |
| "learning_rate": 4.6875000000000004e-06, | |
| "loss": 0.0017, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 11.4070796460177, | |
| "grad_norm": 0.05795137956738472, | |
| "learning_rate": 4.3750000000000005e-06, | |
| "loss": 0.0016, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 11.849557522123893, | |
| "grad_norm": 0.04924352467060089, | |
| "learning_rate": 4.0625000000000005e-06, | |
| "loss": 0.0016, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 12.283185840707965, | |
| "grad_norm": 0.11852666735649109, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.0011, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 12.725663716814159, | |
| "grad_norm": 0.01910184696316719, | |
| "learning_rate": 3.4375e-06, | |
| "loss": 0.0008, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 13.15929203539823, | |
| "grad_norm": 0.046200644224882126, | |
| "learning_rate": 3.125e-06, | |
| "loss": 0.001, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 13.601769911504425, | |
| "grad_norm": 0.02231488563120365, | |
| "learning_rate": 2.8125e-06, | |
| "loss": 0.0009, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 14.035398230088495, | |
| "grad_norm": 0.08958712220191956, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.0007, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 14.035398230088495, | |
| "eval_loss": 0.25818175077438354, | |
| "eval_runtime": 55.2869, | |
| "eval_samples_per_second": 1.791, | |
| "eval_steps_per_second": 0.127, | |
| "eval_wer": 0.2786177105831533, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 14.47787610619469, | |
| "grad_norm": 0.035060103982686996, | |
| "learning_rate": 2.1875000000000002e-06, | |
| "loss": 0.0006, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 14.920353982300885, | |
| "grad_norm": 0.02991672046482563, | |
| "learning_rate": 1.8750000000000003e-06, | |
| "loss": 0.0008, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 15.353982300884956, | |
| "grad_norm": 0.019653476774692535, | |
| "learning_rate": 1.5625e-06, | |
| "loss": 0.0007, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 15.79646017699115, | |
| "grad_norm": 0.07724698632955551, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.0007, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 16.23008849557522, | |
| "grad_norm": 0.01434712577611208, | |
| "learning_rate": 9.375000000000001e-07, | |
| "loss": 0.0006, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 16.672566371681416, | |
| "grad_norm": 0.012539554387331009, | |
| "learning_rate": 6.25e-07, | |
| "loss": 0.0007, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 17.106194690265486, | |
| "grad_norm": 0.027095399796962738, | |
| "learning_rate": 3.125e-07, | |
| "loss": 0.0007, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 17.548672566371682, | |
| "grad_norm": 0.06335917860269547, | |
| "learning_rate": 0.0, | |
| "loss": 0.0006, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 17.548672566371682, | |
| "eval_loss": 0.25996866822242737, | |
| "eval_runtime": 54.379, | |
| "eval_samples_per_second": 1.821, | |
| "eval_steps_per_second": 0.129, | |
| "eval_wer": 0.27645788336933047, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 17.548672566371682, | |
| "step": 1000, | |
| "total_flos": 9.15623758135296e+18, | |
| "train_loss": 0.06514614920597524, | |
| "train_runtime": 15536.8267, | |
| "train_samples_per_second": 2.06, | |
| "train_steps_per_second": 0.064 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 1000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 18, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.15623758135296e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |