| { | |
| "best_metric": 0.8625, | |
| "best_model_checkpoint": "vit-msn-small-finetuned-alzheimers/checkpoint-220", | |
| "epoch": 9.777777777777779, | |
| "eval_steps": 500, | |
| "global_step": 220, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 4.161161422729492, | |
| "learning_rate": 2.272727272727273e-05, | |
| "loss": 1.0688, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 19.670530319213867, | |
| "learning_rate": 4.545454545454546e-05, | |
| "loss": 0.9297, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.9777777777777777, | |
| "eval_accuracy": 0.615625, | |
| "eval_loss": 0.8769256472587585, | |
| "eval_runtime": 3.9355, | |
| "eval_samples_per_second": 162.623, | |
| "eval_steps_per_second": 2.541, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 7.473578929901123, | |
| "learning_rate": 4.797979797979798e-05, | |
| "loss": 0.8854, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 7.05064058303833, | |
| "learning_rate": 4.545454545454546e-05, | |
| "loss": 0.8601, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.634375, | |
| "eval_loss": 0.7798857688903809, | |
| "eval_runtime": 3.5552, | |
| "eval_samples_per_second": 180.019, | |
| "eval_steps_per_second": 2.813, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 12.152384757995605, | |
| "learning_rate": 4.292929292929293e-05, | |
| "loss": 0.8571, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 11.653566360473633, | |
| "learning_rate": 4.0404040404040405e-05, | |
| "loss": 0.7954, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.977777777777778, | |
| "eval_accuracy": 0.6828125, | |
| "eval_loss": 0.7196512222290039, | |
| "eval_runtime": 3.5735, | |
| "eval_samples_per_second": 179.098, | |
| "eval_steps_per_second": 2.798, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 3.111111111111111, | |
| "grad_norm": 10.308629989624023, | |
| "learning_rate": 3.787878787878788e-05, | |
| "loss": 0.7808, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 15.87501335144043, | |
| "learning_rate": 3.535353535353535e-05, | |
| "loss": 0.7552, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 9.567431449890137, | |
| "learning_rate": 3.282828282828283e-05, | |
| "loss": 0.7468, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.6734375, | |
| "eval_loss": 0.7003158330917358, | |
| "eval_runtime": 3.5895, | |
| "eval_samples_per_second": 178.3, | |
| "eval_steps_per_second": 2.786, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 4.444444444444445, | |
| "grad_norm": 6.920608043670654, | |
| "learning_rate": 3.0303030303030306e-05, | |
| "loss": 0.683, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 4.888888888888889, | |
| "grad_norm": 8.078107833862305, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.6935, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.977777777777778, | |
| "eval_accuracy": 0.7546875, | |
| "eval_loss": 0.6063631772994995, | |
| "eval_runtime": 3.8885, | |
| "eval_samples_per_second": 164.589, | |
| "eval_steps_per_second": 2.572, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 5.333333333333333, | |
| "grad_norm": 11.370922088623047, | |
| "learning_rate": 2.5252525252525256e-05, | |
| "loss": 0.6469, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 5.777777777777778, | |
| "grad_norm": 12.74963665008545, | |
| "learning_rate": 2.272727272727273e-05, | |
| "loss": 0.6271, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.76875, | |
| "eval_loss": 0.5647965669631958, | |
| "eval_runtime": 3.9273, | |
| "eval_samples_per_second": 162.963, | |
| "eval_steps_per_second": 2.546, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 6.222222222222222, | |
| "grad_norm": 11.86828327178955, | |
| "learning_rate": 2.0202020202020203e-05, | |
| "loss": 0.5635, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 13.892380714416504, | |
| "learning_rate": 1.7676767676767676e-05, | |
| "loss": 0.5622, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 6.977777777777778, | |
| "eval_accuracy": 0.809375, | |
| "eval_loss": 0.48242831230163574, | |
| "eval_runtime": 3.5816, | |
| "eval_samples_per_second": 178.692, | |
| "eval_steps_per_second": 2.792, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 7.111111111111111, | |
| "grad_norm": 8.864046096801758, | |
| "learning_rate": 1.5151515151515153e-05, | |
| "loss": 0.525, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 7.555555555555555, | |
| "grad_norm": 8.477625846862793, | |
| "learning_rate": 1.2626262626262628e-05, | |
| "loss": 0.4967, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 12.289462089538574, | |
| "learning_rate": 1.0101010101010101e-05, | |
| "loss": 0.4815, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8609375, | |
| "eval_loss": 0.4012059271335602, | |
| "eval_runtime": 3.613, | |
| "eval_samples_per_second": 177.138, | |
| "eval_steps_per_second": 2.768, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 8.444444444444445, | |
| "grad_norm": 8.446834564208984, | |
| "learning_rate": 7.5757575757575764e-06, | |
| "loss": 0.45, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 8.88888888888889, | |
| "grad_norm": 15.683026313781738, | |
| "learning_rate": 5.050505050505051e-06, | |
| "loss": 0.4771, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 8.977777777777778, | |
| "eval_accuracy": 0.85625, | |
| "eval_loss": 0.3798871636390686, | |
| "eval_runtime": 3.5689, | |
| "eval_samples_per_second": 179.325, | |
| "eval_steps_per_second": 2.802, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 9.333333333333334, | |
| "grad_norm": 9.576162338256836, | |
| "learning_rate": 2.5252525252525253e-06, | |
| "loss": 0.4376, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 9.777777777777779, | |
| "grad_norm": 9.685094833374023, | |
| "learning_rate": 0.0, | |
| "loss": 0.4171, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 9.777777777777779, | |
| "eval_accuracy": 0.8625, | |
| "eval_loss": 0.3611668050289154, | |
| "eval_runtime": 3.9731, | |
| "eval_samples_per_second": 161.083, | |
| "eval_steps_per_second": 2.517, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 9.777777777777779, | |
| "step": 220, | |
| "total_flos": 1.102133137023959e+18, | |
| "train_loss": 0.670017595724626, | |
| "train_runtime": 696.1068, | |
| "train_samples_per_second": 82.746, | |
| "train_steps_per_second": 0.316 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 220, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 1.102133137023959e+18, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |