| { | |
| "best_metric": 0.7738095238095238, | |
| "best_model_checkpoint": "Human-action-swin/checkpoint-234", | |
| "epoch": 2.9714285714285715, | |
| "eval_steps": 500, | |
| "global_step": 234, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 8.177011489868164, | |
| "learning_rate": 2.0833333333333336e-05, | |
| "loss": 2.8077, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 7.242084980010986, | |
| "learning_rate": 4.166666666666667e-05, | |
| "loss": 2.6464, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 6.426877498626709, | |
| "learning_rate": 4.8571428571428576e-05, | |
| "loss": 2.402, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 12.246893882751465, | |
| "learning_rate": 4.6190476190476194e-05, | |
| "loss": 2.0132, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 20.22879981994629, | |
| "learning_rate": 4.380952380952381e-05, | |
| "loss": 1.6849, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 11.127490043640137, | |
| "learning_rate": 4.1428571428571437e-05, | |
| "loss": 1.5244, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 8.744277954101562, | |
| "learning_rate": 3.904761904761905e-05, | |
| "loss": 1.4273, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "eval_accuracy": 0.671031746031746, | |
| "eval_loss": 1.046990990638733, | |
| "eval_runtime": 38.8701, | |
| "eval_samples_per_second": 64.831, | |
| "eval_steps_per_second": 2.032, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 7.494091033935547, | |
| "learning_rate": 3.6666666666666666e-05, | |
| "loss": 1.3184, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 6.497593879699707, | |
| "learning_rate": 3.428571428571429e-05, | |
| "loss": 1.2244, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 10.031980514526367, | |
| "learning_rate": 3.19047619047619e-05, | |
| "loss": 1.2127, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 8.93614673614502, | |
| "learning_rate": 2.9523809523809526e-05, | |
| "loss": 1.1732, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 8.977005958557129, | |
| "learning_rate": 2.714285714285714e-05, | |
| "loss": 1.1427, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 8.760551452636719, | |
| "learning_rate": 2.4761904761904762e-05, | |
| "loss": 1.1373, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 8.87807846069336, | |
| "learning_rate": 2.2380952380952384e-05, | |
| "loss": 1.1024, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 7.670358657836914, | |
| "learning_rate": 2e-05, | |
| "loss": 1.0813, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "eval_accuracy": 0.7511904761904762, | |
| "eval_loss": 0.7847604155540466, | |
| "eval_runtime": 40.8525, | |
| "eval_samples_per_second": 61.685, | |
| "eval_steps_per_second": 1.934, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 9.62759780883789, | |
| "learning_rate": 1.761904761904762e-05, | |
| "loss": 1.0885, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 8.1283540725708, | |
| "learning_rate": 1.5238095238095241e-05, | |
| "loss": 1.0769, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 8.88747501373291, | |
| "learning_rate": 1.2857142857142857e-05, | |
| "loss": 1.0129, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "grad_norm": 9.25979995727539, | |
| "learning_rate": 1.0476190476190477e-05, | |
| "loss": 1.0048, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 7.371565818786621, | |
| "learning_rate": 8.095238095238097e-06, | |
| "loss": 0.9917, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 6.723045349121094, | |
| "learning_rate": 5.7142857142857145e-06, | |
| "loss": 0.9229, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "grad_norm": 7.953981399536133, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 1.0518, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 7.315853118896484, | |
| "learning_rate": 9.523809523809526e-07, | |
| "loss": 0.9644, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "eval_accuracy": 0.7738095238095238, | |
| "eval_loss": 0.7221077084541321, | |
| "eval_runtime": 40.8941, | |
| "eval_samples_per_second": 61.623, | |
| "eval_steps_per_second": 1.932, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "step": 234, | |
| "total_flos": 7.447570178286551e+17, | |
| "train_loss": 1.3846357653283665, | |
| "train_runtime": 446.4628, | |
| "train_samples_per_second": 67.732, | |
| "train_steps_per_second": 0.524 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 234, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 7.447570178286551e+17, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |