| { | |
| "best_metric": 0.7981651376146789, | |
| "best_model_checkpoint": "tiny-bert-sst2/run-0/checkpoint-1500", | |
| "epoch": 3.0, | |
| "eval_steps": 100, | |
| "global_step": 1581, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.18975332068311196, | |
| "grad_norm": 0.7187711000442505, | |
| "learning_rate": 2.4275077059878603e-05, | |
| "loss": 1.2604, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.18975332068311196, | |
| "eval_accuracy": 0.6169724770642202, | |
| "eval_loss": 1.181114673614502, | |
| "eval_runtime": 2.4577, | |
| "eval_samples_per_second": 354.806, | |
| "eval_steps_per_second": 2.848, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3795066413662239, | |
| "grad_norm": 1.8982616662979126, | |
| "learning_rate": 2.3789299271608236e-05, | |
| "loss": 1.1808, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3795066413662239, | |
| "eval_accuracy": 0.7121559633027523, | |
| "eval_loss": 1.0743426084518433, | |
| "eval_runtime": 2.4969, | |
| "eval_samples_per_second": 349.235, | |
| "eval_steps_per_second": 2.803, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5692599620493358, | |
| "grad_norm": 2.6504175662994385, | |
| "learning_rate": 2.2812529230666033e-05, | |
| "loss": 1.0583, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5692599620493358, | |
| "eval_accuracy": 0.7591743119266054, | |
| "eval_loss": 0.9464107155799866, | |
| "eval_runtime": 2.4114, | |
| "eval_samples_per_second": 361.611, | |
| "eval_steps_per_second": 2.903, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7590132827324478, | |
| "grad_norm": 2.9992282390594482, | |
| "learning_rate": 2.138601927608547e-05, | |
| "loss": 0.9339, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7590132827324478, | |
| "eval_accuracy": 0.7775229357798165, | |
| "eval_loss": 0.8676859736442566, | |
| "eval_runtime": 2.5023, | |
| "eval_samples_per_second": 348.476, | |
| "eval_steps_per_second": 2.797, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9487666034155597, | |
| "grad_norm": 2.979876756668091, | |
| "learning_rate": 1.9570015801402756e-05, | |
| "loss": 0.8783, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.9487666034155597, | |
| "eval_accuracy": 0.786697247706422, | |
| "eval_loss": 0.836247980594635, | |
| "eval_runtime": 2.4279, | |
| "eval_samples_per_second": 359.162, | |
| "eval_steps_per_second": 2.883, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.1385199240986716, | |
| "grad_norm": 3.8972060680389404, | |
| "learning_rate": 1.7441214843436652e-05, | |
| "loss": 0.8286, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.1385199240986716, | |
| "eval_accuracy": 0.786697247706422, | |
| "eval_loss": 0.8169480562210083, | |
| "eval_runtime": 2.4359, | |
| "eval_samples_per_second": 357.985, | |
| "eval_steps_per_second": 2.874, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.3282732447817835, | |
| "grad_norm": 4.094536304473877, | |
| "learning_rate": 1.5089522946380597e-05, | |
| "loss": 0.7808, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.3282732447817835, | |
| "eval_accuracy": 0.7878440366972477, | |
| "eval_loss": 0.8050708770751953, | |
| "eval_runtime": 2.4399, | |
| "eval_samples_per_second": 357.397, | |
| "eval_steps_per_second": 2.869, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.5180265654648957, | |
| "grad_norm": 6.467775344848633, | |
| "learning_rate": 1.2614260100993157e-05, | |
| "loss": 0.765, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.5180265654648957, | |
| "eval_accuracy": 0.7889908256880734, | |
| "eval_loss": 0.7935003042221069, | |
| "eval_runtime": 2.4349, | |
| "eval_samples_per_second": 358.121, | |
| "eval_steps_per_second": 2.875, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.7077798861480076, | |
| "grad_norm": 7.3176445960998535, | |
| "learning_rate": 1.0119965121749416e-05, | |
| "loss": 0.742, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.7077798861480076, | |
| "eval_accuracy": 0.7901376146788991, | |
| "eval_loss": 0.7913622260093689, | |
| "eval_runtime": 2.505, | |
| "eval_samples_per_second": 348.1, | |
| "eval_steps_per_second": 2.794, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.8975332068311195, | |
| "grad_norm": 4.671020030975342, | |
| "learning_rate": 7.711980615220306e-06, | |
| "loss": 0.7379, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.8975332068311195, | |
| "eval_accuracy": 0.7970183486238532, | |
| "eval_loss": 0.7775447964668274, | |
| "eval_runtime": 2.4368, | |
| "eval_samples_per_second": 357.839, | |
| "eval_steps_per_second": 2.873, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0872865275142316, | |
| "grad_norm": 3.548794984817505, | |
| "learning_rate": 5.492004001563045e-06, | |
| "loss": 0.7169, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.0872865275142316, | |
| "eval_accuracy": 0.7981651376146789, | |
| "eval_loss": 0.7763888239860535, | |
| "eval_runtime": 2.4327, | |
| "eval_samples_per_second": 358.447, | |
| "eval_steps_per_second": 2.877, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.2770398481973433, | |
| "grad_norm": 2.657698392868042, | |
| "learning_rate": 3.5537924846990686e-06, | |
| "loss": 0.7053, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.2770398481973433, | |
| "eval_accuracy": 0.8004587155963303, | |
| "eval_loss": 0.7718890905380249, | |
| "eval_runtime": 2.4731, | |
| "eval_samples_per_second": 352.601, | |
| "eval_steps_per_second": 2.831, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.4667931688804554, | |
| "grad_norm": 8.931977272033691, | |
| "learning_rate": 1.9792033649762944e-06, | |
| "loss": 0.7001, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.4667931688804554, | |
| "eval_accuracy": 0.7958715596330275, | |
| "eval_loss": 0.7761191129684448, | |
| "eval_runtime": 2.4453, | |
| "eval_samples_per_second": 356.608, | |
| "eval_steps_per_second": 2.863, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.656546489563567, | |
| "grad_norm": 5.575241565704346, | |
| "learning_rate": 8.347369254527042e-07, | |
| "loss": 0.6988, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.656546489563567, | |
| "eval_accuracy": 0.7981651376146789, | |
| "eval_loss": 0.7722569108009338, | |
| "eval_runtime": 2.492, | |
| "eval_samples_per_second": 349.916, | |
| "eval_steps_per_second": 2.809, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.846299810246679, | |
| "grad_norm": 6.262284278869629, | |
| "learning_rate": 1.6872789753562043e-07, | |
| "loss": 0.7135, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.846299810246679, | |
| "eval_accuracy": 0.7981651376146789, | |
| "eval_loss": 0.7717686891555786, | |
| "eval_runtime": 2.443, | |
| "eval_samples_per_second": 356.945, | |
| "eval_steps_per_second": 2.865, | |
| "step": 1500 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 1581, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "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": 23032282372320.0, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.8670086094179865, | |
| "learning_rate": 2.432723709409012e-05, | |
| "num_train_epochs": 3, | |
| "temperature": 13, | |
| "warmup_ratio": 0.03434495454544933 | |
| } | |
| } | |