| { | |
| "best_metric": 0.8989918573090345, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-0/checkpoint-6450", | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 6450, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.770919680595398, | |
| "learning_rate": 0.00045575725544182215, | |
| "loss": 0.2361, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8898797983714618, | |
| "eval_f1": 0.003508771929824561, | |
| "eval_loss": 0.22947192192077637, | |
| "eval_mcc": 0.039546160983878816, | |
| "eval_precision": 1.0, | |
| "eval_recall": 0.0017574692442882249, | |
| "eval_runtime": 9.9701, | |
| "eval_samples_per_second": 517.345, | |
| "eval_steps_per_second": 16.249, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.8797195553779602, | |
| "learning_rate": 0.00040511756039273084, | |
| "loss": 0.2287, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8896859247770453, | |
| "eval_f1": 0.0, | |
| "eval_loss": 0.22863532602787018, | |
| "eval_mcc": 0.0, | |
| "eval_precision": 0.0, | |
| "eval_recall": 0.0, | |
| "eval_runtime": 10.133, | |
| "eval_samples_per_second": 509.028, | |
| "eval_steps_per_second": 15.987, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.5287452340126038, | |
| "learning_rate": 0.0003544778653436394, | |
| "loss": 0.2263, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8896859247770453, | |
| "eval_f1": 0.0, | |
| "eval_loss": 0.2284112125635147, | |
| "eval_mcc": 0.0, | |
| "eval_precision": 0.0, | |
| "eval_recall": 0.0, | |
| "eval_runtime": 9.9193, | |
| "eval_samples_per_second": 519.998, | |
| "eval_steps_per_second": 16.332, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.02663254737854, | |
| "learning_rate": 0.0003038381702945481, | |
| "loss": 0.2233, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8966653741760372, | |
| "eval_f1": 0.19607843137254902, | |
| "eval_loss": 0.2262517362833023, | |
| "eval_mcc": 0.25274955080099837, | |
| "eval_precision": 0.6914893617021277, | |
| "eval_recall": 0.11423550087873462, | |
| "eval_runtime": 9.9602, | |
| "eval_samples_per_second": 517.859, | |
| "eval_steps_per_second": 16.265, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.33138540387153625, | |
| "learning_rate": 0.00025319847524545675, | |
| "loss": 0.2197, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8970531213648701, | |
| "eval_f1": 0.18683001531393567, | |
| "eval_loss": 0.21906551718711853, | |
| "eval_mcc": 0.25294386512776135, | |
| "eval_precision": 0.7261904761904762, | |
| "eval_recall": 0.10720562390158173, | |
| "eval_runtime": 9.9303, | |
| "eval_samples_per_second": 519.421, | |
| "eval_steps_per_second": 16.314, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.3962567448616028, | |
| "learning_rate": 0.00020255878019636542, | |
| "loss": 0.2188, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8968592477704537, | |
| "eval_f1": 0.1501597444089457, | |
| "eval_loss": 0.21888957917690277, | |
| "eval_mcc": 0.24100417515941616, | |
| "eval_precision": 0.8245614035087719, | |
| "eval_recall": 0.08260105448154657, | |
| "eval_runtime": 9.9381, | |
| "eval_samples_per_second": 519.011, | |
| "eval_steps_per_second": 16.301, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.816288948059082, | |
| "learning_rate": 0.00015191908514727406, | |
| "loss": 0.217, | |
| "step": 4515 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8986041101202016, | |
| "eval_f1": 0.20877458396369136, | |
| "eval_loss": 0.2174820899963379, | |
| "eval_mcc": 0.27516569645731026, | |
| "eval_precision": 0.75, | |
| "eval_recall": 0.12126537785588752, | |
| "eval_runtime": 9.926, | |
| "eval_samples_per_second": 519.645, | |
| "eval_steps_per_second": 16.321, | |
| "step": 4515 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.9780616760253906, | |
| "learning_rate": 0.00010127939009818271, | |
| "loss": 0.2152, | |
| "step": 5160 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8972469949592865, | |
| "eval_f1": 0.16927899686520376, | |
| "eval_loss": 0.2163967788219452, | |
| "eval_mcc": 0.2498815158419147, | |
| "eval_precision": 0.782608695652174, | |
| "eval_recall": 0.09490333919156414, | |
| "eval_runtime": 9.9496, | |
| "eval_samples_per_second": 518.41, | |
| "eval_steps_per_second": 16.282, | |
| "step": 5160 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.3984989523887634, | |
| "learning_rate": 5.0639695049091355e-05, | |
| "loss": 0.2142, | |
| "step": 5805 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.8986041101202016, | |
| "eval_f1": 0.22288261515601784, | |
| "eval_loss": 0.21662580966949463, | |
| "eval_mcc": 0.2797000424158355, | |
| "eval_precision": 0.7211538461538461, | |
| "eval_recall": 0.13181019332161686, | |
| "eval_runtime": 9.9343, | |
| "eval_samples_per_second": 519.211, | |
| "eval_steps_per_second": 16.307, | |
| "step": 5805 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.5117802023887634, | |
| "learning_rate": 0.0, | |
| "loss": 0.2133, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.8989918573090345, | |
| "eval_f1": 0.21654135338345867, | |
| "eval_loss": 0.21611623466014862, | |
| "eval_mcc": 0.2811949499154691, | |
| "eval_precision": 0.75, | |
| "eval_recall": 0.1265377855887522, | |
| "eval_runtime": 9.9337, | |
| "eval_samples_per_second": 519.243, | |
| "eval_steps_per_second": 16.308, | |
| "step": 6450 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 6450, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 15869425326000.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.6185480031675171, | |
| "learning_rate": 0.0005063969504909135, | |
| "num_train_epochs": 10, | |
| "temperature": 26 | |
| } | |
| } | |