| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 19.946666666666665, | |
| "eval_steps": 500, | |
| "global_step": 3740, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 0.0001, | |
| "loss": 0.5411, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7903686274509804, | |
| "eval_loss": 0.39375776052474976, | |
| "eval_runtime": 7.7091, | |
| "eval_samples_per_second": 64.858, | |
| "eval_steps_per_second": 8.172, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4081, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 0.0001, | |
| "loss": 0.362, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7917960784313726, | |
| "eval_loss": 0.3804325461387634, | |
| "eval_runtime": 7.4651, | |
| "eval_samples_per_second": 66.979, | |
| "eval_steps_per_second": 8.439, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3376, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3047, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7891450980392157, | |
| "eval_loss": 0.39341700077056885, | |
| "eval_runtime": 7.1191, | |
| "eval_samples_per_second": 70.233, | |
| "eval_steps_per_second": 8.849, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 0.0001, | |
| "loss": 0.2848, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "learning_rate": 0.0001, | |
| "loss": 0.2469, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7846274509803921, | |
| "eval_loss": 0.42259183526039124, | |
| "eval_runtime": 6.9941, | |
| "eval_samples_per_second": 71.489, | |
| "eval_steps_per_second": 9.008, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "learning_rate": 0.0001, | |
| "loss": 0.2215, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "learning_rate": 0.0001, | |
| "loss": 0.2022, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7803058823529412, | |
| "eval_loss": 0.4660645127296448, | |
| "eval_runtime": 6.6511, | |
| "eval_samples_per_second": 75.175, | |
| "eval_steps_per_second": 9.472, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 5.33, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1783, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 5.87, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1681, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7761333333333333, | |
| "eval_loss": 0.5122596621513367, | |
| "eval_runtime": 7.6517, | |
| "eval_samples_per_second": 65.345, | |
| "eval_steps_per_second": 8.233, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1437, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 6.93, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1404, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7720627450980392, | |
| "eval_loss": 0.5730750560760498, | |
| "eval_runtime": 7.0342, | |
| "eval_samples_per_second": 71.081, | |
| "eval_steps_per_second": 8.956, | |
| "step": 1312 | |
| }, | |
| { | |
| "epoch": 7.47, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1179, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1197, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.770078431372549, | |
| "eval_loss": 0.6074877977371216, | |
| "eval_runtime": 6.9934, | |
| "eval_samples_per_second": 71.496, | |
| "eval_steps_per_second": 9.008, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 8.53, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.7688470588235294, | |
| "eval_loss": 0.631741464138031, | |
| "eval_runtime": 7.3864, | |
| "eval_samples_per_second": 67.692, | |
| "eval_steps_per_second": 8.529, | |
| "step": 1687 | |
| }, | |
| { | |
| "epoch": 9.07, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1027, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "learning_rate": 0.0001, | |
| "loss": 0.089, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.7663764705882353, | |
| "eval_loss": 0.6717921495437622, | |
| "eval_runtime": 6.6214, | |
| "eval_samples_per_second": 75.513, | |
| "eval_steps_per_second": 9.515, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 10.13, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0911, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 10.67, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0837, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.7653333333333333, | |
| "eval_loss": 0.6921772956848145, | |
| "eval_runtime": 6.6402, | |
| "eval_samples_per_second": 75.299, | |
| "eval_steps_per_second": 9.488, | |
| "step": 2062 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0823, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 11.73, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0788, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.7631764705882353, | |
| "eval_loss": 0.7253576517105103, | |
| "eval_runtime": 7.6814, | |
| "eval_samples_per_second": 65.092, | |
| "eval_steps_per_second": 8.202, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 12.27, | |
| "learning_rate": 0.0001, | |
| "loss": 0.078, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0761, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.7628627450980392, | |
| "eval_loss": 0.7256088256835938, | |
| "eval_runtime": 7.6806, | |
| "eval_samples_per_second": 65.099, | |
| "eval_steps_per_second": 8.202, | |
| "step": 2437 | |
| }, | |
| { | |
| "epoch": 13.33, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0743, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 13.87, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0749, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.7620627450980392, | |
| "eval_loss": 0.7533740401268005, | |
| "eval_runtime": 7.644, | |
| "eval_samples_per_second": 65.41, | |
| "eval_steps_per_second": 8.242, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 14.4, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0719, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 14.93, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0741, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.7620078431372549, | |
| "eval_loss": 0.7529163360595703, | |
| "eval_runtime": 7.4208, | |
| "eval_samples_per_second": 67.378, | |
| "eval_steps_per_second": 8.49, | |
| "step": 2812 | |
| }, | |
| { | |
| "epoch": 15.47, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0704, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0726, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.7610745098039216, | |
| "eval_loss": 0.7678206562995911, | |
| "eval_runtime": 6.6228, | |
| "eval_samples_per_second": 75.497, | |
| "eval_steps_per_second": 9.513, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 16.53, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0687, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.7610274509803922, | |
| "eval_loss": 0.7728469371795654, | |
| "eval_runtime": 7.7738, | |
| "eval_samples_per_second": 64.319, | |
| "eval_steps_per_second": 8.104, | |
| "step": 3187 | |
| }, | |
| { | |
| "epoch": 17.07, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0706, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 17.6, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0682, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.7603058823529412, | |
| "eval_loss": 0.7807328104972839, | |
| "eval_runtime": 7.6425, | |
| "eval_samples_per_second": 65.424, | |
| "eval_steps_per_second": 8.243, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 18.13, | |
| "learning_rate": 0.0001, | |
| "loss": 0.069, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 18.67, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0682, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.7609882352941176, | |
| "eval_loss": 0.7872016429901123, | |
| "eval_runtime": 7.0778, | |
| "eval_samples_per_second": 70.644, | |
| "eval_steps_per_second": 8.901, | |
| "step": 3562 | |
| }, | |
| { | |
| "epoch": 19.2, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0682, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 19.73, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0682, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 19.95, | |
| "eval_accuracy": 0.7597490196078431, | |
| "eval_loss": 0.8054783940315247, | |
| "eval_runtime": 7.0912, | |
| "eval_samples_per_second": 70.51, | |
| "eval_steps_per_second": 8.884, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 19.95, | |
| "step": 3740, | |
| "total_flos": 8.667972359988183e+17, | |
| "train_loss": 0.1472176224152672, | |
| "train_runtime": 2965.7973, | |
| "train_samples_per_second": 40.461, | |
| "train_steps_per_second": 1.261 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 3740, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 8.667972359988183e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |