| { | |
| "best_metric": 0.7694117647058824, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-17/checkpoint-1440", | |
| "epoch": 9.0, | |
| "eval_steps": 500, | |
| "global_step": 1440, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.2461354732513428, | |
| "learning_rate": 0.0004470795307643556, | |
| "loss": 0.5079, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_f1": 0.0, | |
| "eval_loss": 0.48573797941207886, | |
| "eval_mcc": 0.0, | |
| "eval_precision": 0.0, | |
| "eval_recall": 0.0, | |
| "eval_runtime": 1.875, | |
| "eval_samples_per_second": 679.994, | |
| "eval_steps_per_second": 21.333, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.359627366065979, | |
| "learning_rate": 0.00039119458941881116, | |
| "loss": 0.4809, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.691764705882353, | |
| "eval_f1": 0.18295218295218293, | |
| "eval_loss": 0.46239688992500305, | |
| "eval_mcc": 0.20568505072418322, | |
| "eval_precision": 0.7857142857142857, | |
| "eval_recall": 0.10352941176470588, | |
| "eval_runtime": 2.0666, | |
| "eval_samples_per_second": 616.947, | |
| "eval_steps_per_second": 19.355, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.3378326892852783, | |
| "learning_rate": 0.0003353096480732667, | |
| "loss": 0.4671, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7223529411764706, | |
| "eval_f1": 0.470059880239521, | |
| "eval_loss": 0.45456141233444214, | |
| "eval_mcc": 0.321941426257721, | |
| "eval_precision": 0.6460905349794238, | |
| "eval_recall": 0.36941176470588233, | |
| "eval_runtime": 1.8782, | |
| "eval_samples_per_second": 678.832, | |
| "eval_steps_per_second": 21.297, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.1498600244522095, | |
| "learning_rate": 0.00027942470672772226, | |
| "loss": 0.4542, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7286274509803922, | |
| "eval_f1": 0.41156462585034015, | |
| "eval_loss": 0.4409485161304474, | |
| "eval_mcc": 0.33217642136796893, | |
| "eval_precision": 0.7423312883435583, | |
| "eval_recall": 0.2847058823529412, | |
| "eval_runtime": 1.8716, | |
| "eval_samples_per_second": 681.245, | |
| "eval_steps_per_second": 21.372, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.308472752571106, | |
| "learning_rate": 0.0002235397653821778, | |
| "loss": 0.4461, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.735686274509804, | |
| "eval_f1": 0.587515299877601, | |
| "eval_loss": 0.4530380070209503, | |
| "eval_mcc": 0.3942171016225521, | |
| "eval_precision": 0.6122448979591837, | |
| "eval_recall": 0.5647058823529412, | |
| "eval_runtime": 1.8727, | |
| "eval_samples_per_second": 680.848, | |
| "eval_steps_per_second": 21.36, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.423608422279358, | |
| "learning_rate": 0.00016765482403663335, | |
| "loss": 0.4408, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7474509803921568, | |
| "eval_f1": 0.48562300319488827, | |
| "eval_loss": 0.43300551176071167, | |
| "eval_mcc": 0.3880833523891067, | |
| "eval_precision": 0.7562189054726368, | |
| "eval_recall": 0.35764705882352943, | |
| "eval_runtime": 1.867, | |
| "eval_samples_per_second": 682.912, | |
| "eval_steps_per_second": 21.425, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 1.9654881954193115, | |
| "learning_rate": 0.0001117698826910889, | |
| "loss": 0.4337, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.756078431372549, | |
| "eval_f1": 0.5997425997425998, | |
| "eval_loss": 0.4409618675708771, | |
| "eval_mcc": 0.4304694096409861, | |
| "eval_precision": 0.6619318181818182, | |
| "eval_recall": 0.548235294117647, | |
| "eval_runtime": 1.8658, | |
| "eval_samples_per_second": 683.346, | |
| "eval_steps_per_second": 21.438, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 1.4317703247070312, | |
| "learning_rate": 5.588494134554445e-05, | |
| "loss": 0.4256, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7647058823529411, | |
| "eval_f1": 0.5945945945945945, | |
| "eval_loss": 0.4346446692943573, | |
| "eval_mcc": 0.44362213114101423, | |
| "eval_precision": 0.6984126984126984, | |
| "eval_recall": 0.5176470588235295, | |
| "eval_runtime": 1.8629, | |
| "eval_samples_per_second": 684.431, | |
| "eval_steps_per_second": 21.472, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 1.2457499504089355, | |
| "learning_rate": 0.0, | |
| "loss": 0.421, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.7694117647058824, | |
| "eval_f1": 0.5939226519337016, | |
| "eval_loss": 0.4335412383079529, | |
| "eval_mcc": 0.452898203483245, | |
| "eval_precision": 0.7190635451505016, | |
| "eval_recall": 0.5058823529411764, | |
| "eval_runtime": 1.8642, | |
| "eval_samples_per_second": 683.942, | |
| "eval_steps_per_second": 21.457, | |
| "step": 1440 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 1440, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9, | |
| "save_steps": 500, | |
| "total_flos": 2625798941280.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.7930879892159086, | |
| "learning_rate": 0.0005029644721099001, | |
| "num_train_epochs": 9, | |
| "temperature": 30 | |
| } | |
| } | |