| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3e-05, | |
| "loss": 2.3698, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8255, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6054430379746836, | |
| "eval_loss": 1.83919358253479, | |
| "eval_runtime": 4.6476, | |
| "eval_samples_per_second": 107.584, | |
| "eval_steps_per_second": 13.556, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7843, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7546, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 3e-05, | |
| "loss": 1.7368, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6078227848101266, | |
| "eval_loss": 1.81111478805542, | |
| "eval_runtime": 4.7063, | |
| "eval_samples_per_second": 106.242, | |
| "eval_steps_per_second": 13.386, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6749, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6689, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.607493670886076, | |
| "eval_loss": 1.8103100061416626, | |
| "eval_runtime": 5.2341, | |
| "eval_samples_per_second": 95.528, | |
| "eval_steps_per_second": 12.037, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6205, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "learning_rate": 3e-05, | |
| "loss": 1.556, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 3e-05, | |
| "loss": 1.5555, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.6067341772151899, | |
| "eval_loss": 1.8414338827133179, | |
| "eval_runtime": 4.8146, | |
| "eval_samples_per_second": 103.85, | |
| "eval_steps_per_second": 13.085, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "learning_rate": 3e-05, | |
| "loss": 1.4289, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "learning_rate": 3e-05, | |
| "loss": 1.4559, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.6037721518987341, | |
| "eval_loss": 1.8992472887039185, | |
| "eval_runtime": 5.1125, | |
| "eval_samples_per_second": 97.799, | |
| "eval_steps_per_second": 12.323, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3828, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3271, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3514, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.6018227848101266, | |
| "eval_loss": 1.9584064483642578, | |
| "eval_runtime": 4.4025, | |
| "eval_samples_per_second": 113.572, | |
| "eval_steps_per_second": 14.31, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "learning_rate": 3e-05, | |
| "loss": 1.2239, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "learning_rate": 3e-05, | |
| "loss": 1.2491, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.5999746835443038, | |
| "eval_loss": 2.030003070831299, | |
| "eval_runtime": 4.7046, | |
| "eval_samples_per_second": 106.279, | |
| "eval_steps_per_second": 13.391, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "learning_rate": 3e-05, | |
| "loss": 1.1873, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "learning_rate": 3e-05, | |
| "loss": 1.1455, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 3e-05, | |
| "loss": 1.1749, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.5981518987341772, | |
| "eval_loss": 2.1050591468811035, | |
| "eval_runtime": 4.5572, | |
| "eval_samples_per_second": 109.717, | |
| "eval_steps_per_second": 13.824, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0398, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0769, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.5953924050632912, | |
| "eval_loss": 2.194838762283325, | |
| "eval_runtime": 5.1306, | |
| "eval_samples_per_second": 97.455, | |
| "eval_steps_per_second": 12.279, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0208, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9809, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0134, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.594253164556962, | |
| "eval_loss": 2.2515170574188232, | |
| "eval_runtime": 4.7037, | |
| "eval_samples_per_second": 106.3, | |
| "eval_steps_per_second": 13.394, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8808, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9209, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.592126582278481, | |
| "eval_loss": 2.3421294689178467, | |
| "eval_runtime": 4.5581, | |
| "eval_samples_per_second": 109.695, | |
| "eval_steps_per_second": 13.822, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "learning_rate": 3e-05, | |
| "loss": 0.881, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 11.6, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8321, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8636, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.5905063291139241, | |
| "eval_loss": 2.4442591667175293, | |
| "eval_runtime": 5.1497, | |
| "eval_samples_per_second": 97.093, | |
| "eval_steps_per_second": 12.234, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 12.4, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7437, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7866, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.588, | |
| "eval_loss": 2.557358741760254, | |
| "eval_runtime": 4.558, | |
| "eval_samples_per_second": 109.698, | |
| "eval_steps_per_second": 13.822, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 13.2, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7408, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 13.6, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7067, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7448, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.5866835443037974, | |
| "eval_loss": 2.579989194869995, | |
| "eval_runtime": 4.4082, | |
| "eval_samples_per_second": 113.424, | |
| "eval_steps_per_second": 14.291, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 14.4, | |
| "learning_rate": 3e-05, | |
| "loss": 0.637, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 14.8, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6709, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.5845569620253165, | |
| "eval_loss": 2.6911704540252686, | |
| "eval_runtime": 5.1406, | |
| "eval_samples_per_second": 97.265, | |
| "eval_steps_per_second": 12.255, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 15.2, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6375, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 15.6, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6088, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6439, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.5853164556962025, | |
| "eval_loss": 2.7545602321624756, | |
| "eval_runtime": 4.7065, | |
| "eval_samples_per_second": 106.237, | |
| "eval_steps_per_second": 13.386, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 16.4, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5552, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 16.8, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5869, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.5831139240506329, | |
| "eval_loss": 2.799652338027954, | |
| "eval_runtime": 5.213, | |
| "eval_samples_per_second": 95.914, | |
| "eval_steps_per_second": 12.085, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 17.2, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5547, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 17.6, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5336, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5596, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.5832911392405064, | |
| "eval_loss": 2.843494176864624, | |
| "eval_runtime": 4.6373, | |
| "eval_samples_per_second": 107.822, | |
| "eval_steps_per_second": 13.586, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 18.4, | |
| "learning_rate": 3e-05, | |
| "loss": 0.4871, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 18.8, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5205, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.5832911392405064, | |
| "eval_loss": 2.9509618282318115, | |
| "eval_runtime": 4.4063, | |
| "eval_samples_per_second": 113.473, | |
| "eval_steps_per_second": 14.298, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 19.2, | |
| "learning_rate": 3e-05, | |
| "loss": 0.4924, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 19.6, | |
| "learning_rate": 3e-05, | |
| "loss": 0.4789, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5045, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.5824050632911393, | |
| "eval_loss": 2.9796953201293945, | |
| "eval_runtime": 5.1055, | |
| "eval_samples_per_second": 97.933, | |
| "eval_steps_per_second": 12.34, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 5000, | |
| "total_flos": 3.1967425075347456e+17, | |
| "train_loss": 1.043557526397705, | |
| "train_runtime": 3605.2187, | |
| "train_samples_per_second": 44.38, | |
| "train_steps_per_second": 1.387 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 5000, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 3.1967425075347456e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |