| { | |
| "best_metric": 0.7607843137254902, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-18/checkpoint-1120", | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 1280, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 2.1976242065429688, | |
| "learning_rate": 0.0004979101297109036, | |
| "loss": 0.5537, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_f1": 0.0, | |
| "eval_loss": 0.5193991661071777, | |
| "eval_mcc": 0.0, | |
| "eval_precision": 0.0, | |
| "eval_recall": 0.0, | |
| "eval_runtime": 1.8654, | |
| "eval_samples_per_second": 683.512, | |
| "eval_steps_per_second": 21.444, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.7457181215286255, | |
| "learning_rate": 0.00043567136349704073, | |
| "loss": 0.517, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7019607843137254, | |
| "eval_f1": 0.27480916030534347, | |
| "eval_loss": 0.4912988245487213, | |
| "eval_mcc": 0.24246530650972198, | |
| "eval_precision": 0.7272727272727273, | |
| "eval_recall": 0.16941176470588235, | |
| "eval_runtime": 2.0814, | |
| "eval_samples_per_second": 612.558, | |
| "eval_steps_per_second": 19.218, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.177762031555176, | |
| "learning_rate": 0.00037343259728317776, | |
| "loss": 0.4984, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7294117647058823, | |
| "eval_f1": 0.5161290322580645, | |
| "eval_loss": 0.4763549864292145, | |
| "eval_mcc": 0.3501338577435767, | |
| "eval_precision": 0.6388888888888888, | |
| "eval_recall": 0.4329411764705882, | |
| "eval_runtime": 1.8808, | |
| "eval_samples_per_second": 677.886, | |
| "eval_steps_per_second": 21.267, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.1199020147323608, | |
| "learning_rate": 0.00031119383106931485, | |
| "loss": 0.4888, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7278431372549019, | |
| "eval_f1": 0.4376012965964343, | |
| "eval_loss": 0.47154995799064636, | |
| "eval_mcc": 0.3302932991068742, | |
| "eval_precision": 0.703125, | |
| "eval_recall": 0.3176470588235294, | |
| "eval_runtime": 1.8859, | |
| "eval_samples_per_second": 676.058, | |
| "eval_steps_per_second": 21.21, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.9632917642593384, | |
| "learning_rate": 0.0002489550648554518, | |
| "loss": 0.4797, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7388235294117647, | |
| "eval_f1": 0.5963636363636363, | |
| "eval_loss": 0.4733050763607025, | |
| "eval_mcc": 0.4039872699550251, | |
| "eval_precision": 0.615, | |
| "eval_recall": 0.5788235294117647, | |
| "eval_runtime": 1.8712, | |
| "eval_samples_per_second": 681.38, | |
| "eval_steps_per_second": 21.377, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 2.4523799419403076, | |
| "learning_rate": 0.00018671629864158888, | |
| "loss": 0.472, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7403921568627451, | |
| "eval_f1": 0.46353322528363045, | |
| "eval_loss": 0.4695047438144684, | |
| "eval_mcc": 0.36750944548511355, | |
| "eval_precision": 0.7447916666666666, | |
| "eval_recall": 0.33647058823529413, | |
| "eval_runtime": 1.8728, | |
| "eval_samples_per_second": 680.799, | |
| "eval_steps_per_second": 21.358, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 3.7260565757751465, | |
| "learning_rate": 0.0001244775324277259, | |
| "loss": 0.4686, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7607843137254902, | |
| "eval_f1": 0.5861601085481682, | |
| "eval_loss": 0.47005850076675415, | |
| "eval_mcc": 0.43344484177930304, | |
| "eval_precision": 0.6923076923076923, | |
| "eval_recall": 0.508235294117647, | |
| "eval_runtime": 1.8699, | |
| "eval_samples_per_second": 681.865, | |
| "eval_steps_per_second": 21.392, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 1.6883962154388428, | |
| "learning_rate": 6.223876621386296e-05, | |
| "loss": 0.4597, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7529411764705882, | |
| "eval_f1": 0.5903771131339401, | |
| "eval_loss": 0.47118470072746277, | |
| "eval_mcc": 0.42107634181861553, | |
| "eval_precision": 0.6598837209302325, | |
| "eval_recall": 0.5341176470588235, | |
| "eval_runtime": 1.8712, | |
| "eval_samples_per_second": 681.374, | |
| "eval_steps_per_second": 21.376, | |
| "step": 1280 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 1440, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9, | |
| "save_steps": 500, | |
| "total_flos": 2334043503360.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.8696755067403684, | |
| "learning_rate": 0.0005601488959247667, | |
| "num_train_epochs": 9, | |
| "temperature": 29 | |
| } | |
| } | |