| { | |
| "best_metric": 0.6984126984126984, | |
| "best_model_checkpoint": "vit-base-patch16-224/checkpoint-245", | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 350, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 6.601813793182373, | |
| "learning_rate": 1.4285714285714286e-06, | |
| "loss": 0.785, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 5.436784267425537, | |
| "learning_rate": 2.8571428571428573e-06, | |
| "loss": 0.7224, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 5.551754474639893, | |
| "learning_rate": 4.2857142857142855e-06, | |
| "loss": 0.6893, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.4126984126984127, | |
| "eval_loss": 0.7319461703300476, | |
| "eval_runtime": 12.0068, | |
| "eval_samples_per_second": 5.247, | |
| "eval_steps_per_second": 1.333, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 7.985645294189453, | |
| "learning_rate": 4.920634920634921e-06, | |
| "loss": 0.7348, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 4.910062313079834, | |
| "learning_rate": 4.761904761904762e-06, | |
| "loss": 0.7049, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 4.96162223815918, | |
| "learning_rate": 4.603174603174604e-06, | |
| "loss": 0.7185, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 7.957433223724365, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 0.702, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5238095238095238, | |
| "eval_loss": 0.686264157295227, | |
| "eval_runtime": 10.8952, | |
| "eval_samples_per_second": 5.782, | |
| "eval_steps_per_second": 1.469, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.2857142857142856, | |
| "grad_norm": 5.780089378356934, | |
| "learning_rate": 4.2857142857142855e-06, | |
| "loss": 0.6473, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.571428571428571, | |
| "grad_norm": 3.8612518310546875, | |
| "learning_rate": 4.126984126984127e-06, | |
| "loss": 0.6843, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 5.266788482666016, | |
| "learning_rate": 3.968253968253968e-06, | |
| "loss": 0.6644, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5873015873015873, | |
| "eval_loss": 0.6796107888221741, | |
| "eval_runtime": 10.7178, | |
| "eval_samples_per_second": 5.878, | |
| "eval_steps_per_second": 1.493, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 3.142857142857143, | |
| "grad_norm": 4.334442615509033, | |
| "learning_rate": 3.80952380952381e-06, | |
| "loss": 0.6628, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 3.4285714285714284, | |
| "grad_norm": 4.810235500335693, | |
| "learning_rate": 3.6507936507936507e-06, | |
| "loss": 0.6903, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.7142857142857144, | |
| "grad_norm": 7.596153736114502, | |
| "learning_rate": 3.492063492063492e-06, | |
| "loss": 0.6859, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 8.868383407592773, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 0.645, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5714285714285714, | |
| "eval_loss": 0.6722006797790527, | |
| "eval_runtime": 10.7432, | |
| "eval_samples_per_second": 5.864, | |
| "eval_steps_per_second": 1.489, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 4.285714285714286, | |
| "grad_norm": 4.674743175506592, | |
| "learning_rate": 3.1746031746031746e-06, | |
| "loss": 0.6336, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 4.571428571428571, | |
| "grad_norm": 5.1306023597717285, | |
| "learning_rate": 3.015873015873016e-06, | |
| "loss": 0.6607, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 4.857142857142857, | |
| "grad_norm": 5.9358015060424805, | |
| "learning_rate": 2.8571428571428573e-06, | |
| "loss": 0.6455, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.6507936507936508, | |
| "eval_loss": 0.6544514298439026, | |
| "eval_runtime": 10.8969, | |
| "eval_samples_per_second": 5.781, | |
| "eval_steps_per_second": 1.468, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 5.142857142857143, | |
| "grad_norm": 4.706002712249756, | |
| "learning_rate": 2.6984126984126986e-06, | |
| "loss": 0.6039, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 5.428571428571429, | |
| "grad_norm": 5.049504280090332, | |
| "learning_rate": 2.53968253968254e-06, | |
| "loss": 0.6881, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "grad_norm": 4.565422534942627, | |
| "learning_rate": 2.380952380952381e-06, | |
| "loss": 0.5841, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 5.591720104217529, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 0.6456, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.6507936507936508, | |
| "eval_loss": 0.6535508036613464, | |
| "eval_runtime": 10.8606, | |
| "eval_samples_per_second": 5.801, | |
| "eval_steps_per_second": 1.473, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 6.285714285714286, | |
| "grad_norm": 5.09126091003418, | |
| "learning_rate": 2.0634920634920634e-06, | |
| "loss": 0.6349, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 6.571428571428571, | |
| "grad_norm": 4.8962812423706055, | |
| "learning_rate": 1.904761904761905e-06, | |
| "loss": 0.5735, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 6.857142857142857, | |
| "grad_norm": 4.864034175872803, | |
| "learning_rate": 1.746031746031746e-06, | |
| "loss": 0.6745, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.6984126984126984, | |
| "eval_loss": 0.6463457942008972, | |
| "eval_runtime": 10.8225, | |
| "eval_samples_per_second": 5.821, | |
| "eval_steps_per_second": 1.478, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 7.142857142857143, | |
| "grad_norm": 7.017258167266846, | |
| "learning_rate": 1.5873015873015873e-06, | |
| "loss": 0.6431, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 7.428571428571429, | |
| "grad_norm": 4.79843282699585, | |
| "learning_rate": 1.4285714285714286e-06, | |
| "loss": 0.6164, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 7.714285714285714, | |
| "grad_norm": 6.843749523162842, | |
| "learning_rate": 1.26984126984127e-06, | |
| "loss": 0.644, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 5.495963096618652, | |
| "learning_rate": 1.111111111111111e-06, | |
| "loss": 0.6369, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.6666666666666666, | |
| "eval_loss": 0.6524515748023987, | |
| "eval_runtime": 11.0752, | |
| "eval_samples_per_second": 5.688, | |
| "eval_steps_per_second": 1.445, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 8.285714285714286, | |
| "grad_norm": 7.264893531799316, | |
| "learning_rate": 9.523809523809525e-07, | |
| "loss": 0.6661, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 8.571428571428571, | |
| "grad_norm": 3.773362636566162, | |
| "learning_rate": 7.936507936507937e-07, | |
| "loss": 0.607, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 8.857142857142858, | |
| "grad_norm": 6.338254451751709, | |
| "learning_rate": 6.34920634920635e-07, | |
| "loss": 0.6012, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.6984126984126984, | |
| "eval_loss": 0.6485763788223267, | |
| "eval_runtime": 10.9179, | |
| "eval_samples_per_second": 5.77, | |
| "eval_steps_per_second": 1.465, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 9.142857142857142, | |
| "grad_norm": 5.557713031768799, | |
| "learning_rate": 4.7619047619047623e-07, | |
| "loss": 0.625, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 9.428571428571429, | |
| "grad_norm": 5.070873737335205, | |
| "learning_rate": 3.174603174603175e-07, | |
| "loss": 0.6012, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 9.714285714285714, | |
| "grad_norm": 6.369267463684082, | |
| "learning_rate": 1.5873015873015874e-07, | |
| "loss": 0.6237, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 8.102092742919922, | |
| "learning_rate": 0.0, | |
| "loss": 0.6219, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.6984126984126984, | |
| "eval_loss": 0.646605372428894, | |
| "eval_runtime": 11.2095, | |
| "eval_samples_per_second": 5.62, | |
| "eval_steps_per_second": 1.427, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 350, | |
| "total_flos": 4.3240530204942336e+17, | |
| "train_loss": 0.6562235273633684, | |
| "train_runtime": 1176.6645, | |
| "train_samples_per_second": 4.742, | |
| "train_steps_per_second": 0.297 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 350, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.3240530204942336e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |