| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0289389067524115, | |
| "eval_steps": 500, | |
| "global_step": 8000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06430868167202572, | |
| "grad_norm": 6.638989448547363, | |
| "learning_rate": 0.00018487499999999998, | |
| "loss": 4.9036, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.06430868167202572, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 109.3576, | |
| "eval_samples_per_second": 35.672, | |
| "eval_steps_per_second": 0.558, | |
| "eval_wer": 0.9911607013476308, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.12861736334405144, | |
| "grad_norm": 5.176253795623779, | |
| "learning_rate": 0.000292, | |
| "loss": 1.3889, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.12861736334405144, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 104.5051, | |
| "eval_samples_per_second": 37.328, | |
| "eval_steps_per_second": 0.584, | |
| "eval_wer": 0.793387431773173, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.19292604501607716, | |
| "grad_norm": 8.991605758666992, | |
| "learning_rate": 0.0002711666666666666, | |
| "loss": 1.0207, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.19292604501607716, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 104.6871, | |
| "eval_samples_per_second": 37.263, | |
| "eval_steps_per_second": 0.583, | |
| "eval_wer": 0.6828961986185577, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.2572347266881029, | |
| "grad_norm": 5.992569446563721, | |
| "learning_rate": 0.000250375, | |
| "loss": 0.8901, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.2572347266881029, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 104.7623, | |
| "eval_samples_per_second": 37.237, | |
| "eval_steps_per_second": 0.582, | |
| "eval_wer": 0.6450031396415978, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.3215434083601286, | |
| "grad_norm": 5.538565635681152, | |
| "learning_rate": 0.00022954166666666666, | |
| "loss": 0.8467, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.3215434083601286, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 105.5938, | |
| "eval_samples_per_second": 36.943, | |
| "eval_steps_per_second": 0.578, | |
| "eval_wer": 0.5457904651499783, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.3858520900321543, | |
| "grad_norm": 10.923360824584961, | |
| "learning_rate": 0.00020874999999999998, | |
| "loss": 0.7813, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.3858520900321543, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 107.2937, | |
| "eval_samples_per_second": 36.358, | |
| "eval_steps_per_second": 0.569, | |
| "eval_wer": 0.5366855045162536, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.45016077170418006, | |
| "grad_norm": 8.696784973144531, | |
| "learning_rate": 0.00018791666666666664, | |
| "loss": 0.7456, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.45016077170418006, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 105.1616, | |
| "eval_samples_per_second": 37.095, | |
| "eval_steps_per_second": 0.58, | |
| "eval_wer": 0.5156016036323238, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.5144694533762058, | |
| "grad_norm": 25.500185012817383, | |
| "learning_rate": 0.00016708333333333332, | |
| "loss": 0.7312, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5144694533762058, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 106.9175, | |
| "eval_samples_per_second": 36.486, | |
| "eval_steps_per_second": 0.571, | |
| "eval_wer": 0.48007535139834806, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5787781350482315, | |
| "grad_norm": 4.240171909332275, | |
| "learning_rate": 0.00014629166666666667, | |
| "loss": 0.6929, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.5787781350482315, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 107.3751, | |
| "eval_samples_per_second": 36.331, | |
| "eval_steps_per_second": 0.568, | |
| "eval_wer": 0.4659469642080858, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.6430868167202572, | |
| "grad_norm": 7.269958019256592, | |
| "learning_rate": 0.0001255, | |
| "loss": 0.6567, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.6430868167202572, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 106.0885, | |
| "eval_samples_per_second": 36.771, | |
| "eval_steps_per_second": 0.575, | |
| "eval_wer": 0.44049171617639954, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.707395498392283, | |
| "grad_norm": 9.208508491516113, | |
| "learning_rate": 0.00010466666666666667, | |
| "loss": 0.6413, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.707395498392283, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 107.3705, | |
| "eval_samples_per_second": 36.332, | |
| "eval_steps_per_second": 0.568, | |
| "eval_wer": 0.4199391392551804, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.7717041800643086, | |
| "grad_norm": 12.485828399658203, | |
| "learning_rate": 8.387499999999999e-05, | |
| "loss": 0.6187, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.7717041800643086, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 105.1802, | |
| "eval_samples_per_second": 37.089, | |
| "eval_steps_per_second": 0.58, | |
| "eval_wer": 0.41023040139110273, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.8360128617363344, | |
| "grad_norm": 2.41277813911438, | |
| "learning_rate": 6.304166666666666e-05, | |
| "loss": 0.6086, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.8360128617363344, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 105.1613, | |
| "eval_samples_per_second": 37.095, | |
| "eval_steps_per_second": 0.58, | |
| "eval_wer": 0.40366130512486115, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.9003215434083601, | |
| "grad_norm": 6.724209308624268, | |
| "learning_rate": 4.2208333333333334e-05, | |
| "loss": 0.585, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.9003215434083601, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 107.3364, | |
| "eval_samples_per_second": 36.344, | |
| "eval_steps_per_second": 0.568, | |
| "eval_wer": 0.38716611119161476, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.9646302250803859, | |
| "grad_norm": 7.223489761352539, | |
| "learning_rate": 2.1374999999999998e-05, | |
| "loss": 0.5509, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.9646302250803859, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 107.0445, | |
| "eval_samples_per_second": 36.443, | |
| "eval_steps_per_second": 0.57, | |
| "eval_wer": 0.3831570303820702, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.0289389067524115, | |
| "grad_norm": 2.0498945713043213, | |
| "learning_rate": 5.416666666666666e-07, | |
| "loss": 0.5263, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.0289389067524115, | |
| "eval_loss": Infinity, | |
| "eval_runtime": 108.0404, | |
| "eval_samples_per_second": 36.107, | |
| "eval_steps_per_second": 0.565, | |
| "eval_wer": 0.37269960875235475, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.0289389067524115, | |
| "step": 8000, | |
| "total_flos": 8.75110352693014e+18, | |
| "train_loss": 1.0117860870361328, | |
| "train_runtime": 5013.4693, | |
| "train_samples_per_second": 12.766, | |
| "train_steps_per_second": 1.596 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 8000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.75110352693014e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |