| { | |
| "best_metric": 96.2, | |
| "best_model_checkpoint": "outputs/bitfit/t5-base/qnli/checkpoint-3800", | |
| "epoch": 3.0, | |
| "global_step": 9726, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06, | |
| "eval_accuracy": 92.5, | |
| "eval_average_metrics": 92.5, | |
| "eval_loss": 0.07770004123449326, | |
| "eval_runtime": 4.8406, | |
| "eval_samples_per_second": 206.585, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "eval_accuracy": 94.89999999999999, | |
| "eval_average_metrics": 94.89999999999999, | |
| "eval_loss": 0.05564185231924057, | |
| "eval_runtime": 4.8573, | |
| "eval_samples_per_second": 205.876, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 0.0002845774213448488, | |
| "loss": 0.1452, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "eval_accuracy": 94.89999999999999, | |
| "eval_average_metrics": 94.89999999999999, | |
| "eval_loss": 0.058011941611766815, | |
| "eval_runtime": 4.8049, | |
| "eval_samples_per_second": 208.12, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_accuracy": 94.69999999999999, | |
| "eval_average_metrics": 94.69999999999999, | |
| "eval_loss": 0.054659225046634674, | |
| "eval_runtime": 4.7475, | |
| "eval_samples_per_second": 210.637, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 0.0002691548426896977, | |
| "loss": 0.0755, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "eval_accuracy": 95.5, | |
| "eval_average_metrics": 95.5, | |
| "eval_loss": 0.050559960305690765, | |
| "eval_runtime": 4.7617, | |
| "eval_samples_per_second": 210.008, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "eval_accuracy": 95.1, | |
| "eval_average_metrics": 95.1, | |
| "eval_loss": 0.05192100629210472, | |
| "eval_runtime": 4.7418, | |
| "eval_samples_per_second": 210.892, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "eval_accuracy": 95.1, | |
| "eval_average_metrics": 95.1, | |
| "eval_loss": 0.05229589715600014, | |
| "eval_runtime": 4.8058, | |
| "eval_samples_per_second": 208.082, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 0.00025373226403454656, | |
| "loss": 0.0741, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "eval_accuracy": 95.0, | |
| "eval_average_metrics": 95.0, | |
| "eval_loss": 0.054607994854450226, | |
| "eval_runtime": 4.7778, | |
| "eval_samples_per_second": 209.3, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "eval_accuracy": 94.5, | |
| "eval_average_metrics": 94.5, | |
| "eval_loss": 0.061842530965805054, | |
| "eval_runtime": 4.6161, | |
| "eval_samples_per_second": 216.634, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 0.0002383096853793954, | |
| "loss": 0.0767, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "eval_accuracy": 95.19999999999999, | |
| "eval_average_metrics": 95.19999999999999, | |
| "eval_loss": 0.05252940207719803, | |
| "eval_runtime": 4.9106, | |
| "eval_samples_per_second": 203.642, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "eval_accuracy": 95.5, | |
| "eval_average_metrics": 95.5, | |
| "eval_loss": 0.05460799112915993, | |
| "eval_runtime": 4.7646, | |
| "eval_samples_per_second": 209.881, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "eval_accuracy": 95.89999999999999, | |
| "eval_average_metrics": 95.89999999999999, | |
| "eval_loss": 0.05030712112784386, | |
| "eval_runtime": 4.7399, | |
| "eval_samples_per_second": 210.974, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 0.00022288710672424428, | |
| "loss": 0.0742, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_accuracy": 95.8, | |
| "eval_average_metrics": 95.8, | |
| "eval_loss": 0.048599060624837875, | |
| "eval_runtime": 4.7193, | |
| "eval_samples_per_second": 211.896, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "eval_accuracy": 95.6, | |
| "eval_average_metrics": 95.6, | |
| "eval_loss": 0.05030570924282074, | |
| "eval_runtime": 4.7668, | |
| "eval_samples_per_second": 209.784, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 0.00020746452806909312, | |
| "loss": 0.075, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "eval_accuracy": 95.8, | |
| "eval_average_metrics": 95.8, | |
| "eval_loss": 0.047494787722826004, | |
| "eval_runtime": 4.7477, | |
| "eval_samples_per_second": 210.63, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "eval_accuracy": 95.19999999999999, | |
| "eval_average_metrics": 95.19999999999999, | |
| "eval_loss": 0.04927437752485275, | |
| "eval_runtime": 4.7206, | |
| "eval_samples_per_second": 211.839, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "eval_accuracy": 95.6, | |
| "eval_average_metrics": 95.6, | |
| "eval_loss": 0.05033966153860092, | |
| "eval_runtime": 4.6755, | |
| "eval_samples_per_second": 213.88, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 0.000192041949413942, | |
| "loss": 0.0715, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "eval_accuracy": 95.0, | |
| "eval_average_metrics": 95.0, | |
| "eval_loss": 0.05600408464670181, | |
| "eval_runtime": 4.8028, | |
| "eval_samples_per_second": 208.211, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "eval_accuracy": 96.2, | |
| "eval_average_metrics": 96.2, | |
| "eval_loss": 0.04570974037051201, | |
| "eval_runtime": 4.76, | |
| "eval_samples_per_second": 210.082, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 0.00017661937075879084, | |
| "loss": 0.0709, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "eval_accuracy": 95.7, | |
| "eval_average_metrics": 95.7, | |
| "eval_loss": 0.04873451963067055, | |
| "eval_runtime": 4.7588, | |
| "eval_samples_per_second": 210.137, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "eval_accuracy": 95.7, | |
| "eval_average_metrics": 95.7, | |
| "eval_loss": 0.04785784333944321, | |
| "eval_runtime": 4.7192, | |
| "eval_samples_per_second": 211.9, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "eval_accuracy": 95.89999999999999, | |
| "eval_average_metrics": 95.89999999999999, | |
| "eval_loss": 0.05093304440379143, | |
| "eval_runtime": 4.68, | |
| "eval_samples_per_second": 213.674, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 0.0001611967921036397, | |
| "loss": 0.0693, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "eval_accuracy": 95.19999999999999, | |
| "eval_average_metrics": 95.19999999999999, | |
| "eval_loss": 0.05333869531750679, | |
| "eval_runtime": 4.6769, | |
| "eval_samples_per_second": 213.817, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "eval_accuracy": 95.19999999999999, | |
| "eval_average_metrics": 95.19999999999999, | |
| "eval_loss": 0.05299631133675575, | |
| "eval_runtime": 4.8038, | |
| "eval_samples_per_second": 208.169, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "learning_rate": 0.00014577421344848858, | |
| "loss": 0.0701, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "eval_accuracy": 95.39999999999999, | |
| "eval_average_metrics": 95.39999999999999, | |
| "eval_loss": 0.049973100423812866, | |
| "eval_runtime": 4.7662, | |
| "eval_samples_per_second": 209.812, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "eval_accuracy": 95.39999999999999, | |
| "eval_average_metrics": 95.39999999999999, | |
| "eval_loss": 0.04949893057346344, | |
| "eval_runtime": 3.1699, | |
| "eval_samples_per_second": 315.466, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "eval_accuracy": 95.39999999999999, | |
| "eval_average_metrics": 95.39999999999999, | |
| "eval_loss": 0.0472724474966526, | |
| "eval_runtime": 4.7592, | |
| "eval_samples_per_second": 210.12, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 0.00013035163479333743, | |
| "loss": 0.0701, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "eval_accuracy": 95.7, | |
| "eval_average_metrics": 95.7, | |
| "eval_loss": 0.04894798621535301, | |
| "eval_runtime": 4.9844, | |
| "eval_samples_per_second": 200.627, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "eval_accuracy": 95.39999999999999, | |
| "eval_average_metrics": 95.39999999999999, | |
| "eval_loss": 0.046521034091711044, | |
| "eval_runtime": 4.8064, | |
| "eval_samples_per_second": 208.058, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "learning_rate": 0.0001149290561381863, | |
| "loss": 0.0703, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "eval_accuracy": 95.89999999999999, | |
| "eval_average_metrics": 95.89999999999999, | |
| "eval_loss": 0.04965585470199585, | |
| "eval_runtime": 4.7332, | |
| "eval_samples_per_second": 211.272, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "eval_accuracy": 95.7, | |
| "eval_average_metrics": 95.7, | |
| "eval_loss": 0.04986129701137543, | |
| "eval_runtime": 4.7614, | |
| "eval_samples_per_second": 210.02, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "eval_accuracy": 95.39999999999999, | |
| "eval_average_metrics": 95.39999999999999, | |
| "eval_loss": 0.04970705509185791, | |
| "eval_runtime": 4.7354, | |
| "eval_samples_per_second": 211.177, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 9.950647748303516e-05, | |
| "loss": 0.0683, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "eval_accuracy": 95.5, | |
| "eval_average_metrics": 95.5, | |
| "eval_loss": 0.05166413262486458, | |
| "eval_runtime": 4.7531, | |
| "eval_samples_per_second": 210.389, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "eval_accuracy": 95.6, | |
| "eval_average_metrics": 95.6, | |
| "eval_loss": 0.04985863342881203, | |
| "eval_runtime": 4.6875, | |
| "eval_samples_per_second": 213.333, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 8.408389882788402e-05, | |
| "loss": 0.0639, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "eval_accuracy": 95.6, | |
| "eval_average_metrics": 95.6, | |
| "eval_loss": 0.04906700178980827, | |
| "eval_runtime": 4.7251, | |
| "eval_samples_per_second": 211.637, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "eval_accuracy": 95.39999999999999, | |
| "eval_average_metrics": 95.39999999999999, | |
| "eval_loss": 0.05032121390104294, | |
| "eval_runtime": 4.6901, | |
| "eval_samples_per_second": 213.216, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "eval_accuracy": 95.6, | |
| "eval_average_metrics": 95.6, | |
| "eval_loss": 0.049571845680475235, | |
| "eval_runtime": 4.7849, | |
| "eval_samples_per_second": 208.989, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "learning_rate": 6.866132017273288e-05, | |
| "loss": 0.0628, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "eval_accuracy": 95.7, | |
| "eval_average_metrics": 95.7, | |
| "eval_loss": 0.05029569938778877, | |
| "eval_runtime": 4.7631, | |
| "eval_samples_per_second": 209.949, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "eval_accuracy": 95.39999999999999, | |
| "eval_average_metrics": 95.39999999999999, | |
| "eval_loss": 0.05099053308367729, | |
| "eval_runtime": 4.6959, | |
| "eval_samples_per_second": 212.954, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "learning_rate": 5.323874151758173e-05, | |
| "loss": 0.0629, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "eval_accuracy": 95.7, | |
| "eval_average_metrics": 95.7, | |
| "eval_loss": 0.04943836107850075, | |
| "eval_runtime": 4.8312, | |
| "eval_samples_per_second": 206.989, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "eval_accuracy": 95.39999999999999, | |
| "eval_average_metrics": 95.39999999999999, | |
| "eval_loss": 0.04987151920795441, | |
| "eval_runtime": 4.7857, | |
| "eval_samples_per_second": 208.955, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "eval_accuracy": 95.89999999999999, | |
| "eval_average_metrics": 95.89999999999999, | |
| "eval_loss": 0.04786397144198418, | |
| "eval_runtime": 4.6689, | |
| "eval_samples_per_second": 214.182, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "learning_rate": 3.781616286243059e-05, | |
| "loss": 0.0644, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "eval_accuracy": 96.0, | |
| "eval_average_metrics": 96.0, | |
| "eval_loss": 0.04664640501141548, | |
| "eval_runtime": 4.8983, | |
| "eval_samples_per_second": 204.153, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "eval_accuracy": 95.89999999999999, | |
| "eval_average_metrics": 95.89999999999999, | |
| "eval_loss": 0.04768410325050354, | |
| "eval_runtime": 4.7197, | |
| "eval_samples_per_second": 211.878, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "learning_rate": 2.2393584207279457e-05, | |
| "loss": 0.0654, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "eval_accuracy": 95.8, | |
| "eval_average_metrics": 95.8, | |
| "eval_loss": 0.04792344942688942, | |
| "eval_runtime": 4.8006, | |
| "eval_samples_per_second": 208.308, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "eval_accuracy": 95.6, | |
| "eval_average_metrics": 95.6, | |
| "eval_loss": 0.04864838346838951, | |
| "eval_runtime": 4.7035, | |
| "eval_samples_per_second": 212.608, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "eval_accuracy": 95.89999999999999, | |
| "eval_average_metrics": 95.89999999999999, | |
| "eval_loss": 0.0480988509953022, | |
| "eval_runtime": 4.7424, | |
| "eval_samples_per_second": 210.863, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 6.971005552128316e-06, | |
| "loss": 0.0601, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "eval_accuracy": 95.89999999999999, | |
| "eval_average_metrics": 95.89999999999999, | |
| "eval_loss": 0.04831766337156296, | |
| "eval_runtime": 4.7844, | |
| "eval_samples_per_second": 209.012, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 9726, | |
| "total_flos": 4.195152360543514e+16, | |
| "train_loss": 0.07298248127461363, | |
| "train_runtime": 3523.2101, | |
| "train_samples_per_second": 88.337, | |
| "train_steps_per_second": 2.761 | |
| } | |
| ], | |
| "max_steps": 9726, | |
| "num_train_epochs": 3, | |
| "total_flos": 4.195152360543514e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |