| { | |
| "best_metric": 0.9583333333333334, | |
| "best_model_checkpoint": "timesformer-base-finetuned-k400-finetuned-kinetic-subset-three-local-temporal-with-spatial/checkpoint-150", | |
| "epoch": 5.166666666666667, | |
| "eval_steps": 500, | |
| "global_step": 300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.01975257694721222, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.0009, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.014412298798561096, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.0005, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.048455920070409775, | |
| "learning_rate": 5e-05, | |
| "loss": 0.0018, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.003363919910043478, | |
| "learning_rate": 4.814814814814815e-05, | |
| "loss": 0.0012, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.024853387847542763, | |
| "learning_rate": 4.62962962962963e-05, | |
| "loss": 0.0003, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "eval_accuracy": 0.8472222222222222, | |
| "eval_loss": 0.4552870988845825, | |
| "eval_runtime": 58.6213, | |
| "eval_samples_per_second": 1.228, | |
| "eval_steps_per_second": 1.228, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 0.004102244041860104, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 0.3192, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 0.003750218078494072, | |
| "learning_rate": 4.259259259259259e-05, | |
| "loss": 0.0002, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.002901990432292223, | |
| "learning_rate": 4.074074074074074e-05, | |
| "loss": 0.5754, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 0.019022488966584206, | |
| "learning_rate": 3.888888888888889e-05, | |
| "loss": 0.0002, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 0.004384249914437532, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 0.0004, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "eval_accuracy": 0.9027777777777778, | |
| "eval_loss": 0.33492106199264526, | |
| "eval_runtime": 58.3208, | |
| "eval_samples_per_second": 1.235, | |
| "eval_steps_per_second": 1.235, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 0.008162049576640129, | |
| "learning_rate": 3.518518518518519e-05, | |
| "loss": 0.0003, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "grad_norm": 0.0084863705560565, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.0002, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 0.005603179335594177, | |
| "learning_rate": 3.148148148148148e-05, | |
| "loss": 0.0004, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 0.035682689398527145, | |
| "learning_rate": 2.962962962962963e-05, | |
| "loss": 0.0003, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 0.07813193649053574, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.0004, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "eval_accuracy": 0.9583333333333334, | |
| "eval_loss": 0.12790486216545105, | |
| "eval_runtime": 58.0641, | |
| "eval_samples_per_second": 1.24, | |
| "eval_steps_per_second": 1.24, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 0.0019370431546121836, | |
| "learning_rate": 2.5925925925925925e-05, | |
| "loss": 0.0003, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "grad_norm": 0.0020284331403672695, | |
| "learning_rate": 2.4074074074074074e-05, | |
| "loss": 0.5874, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 0.0013429196551442146, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.019, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "grad_norm": 3.899311065673828, | |
| "learning_rate": 2.037037037037037e-05, | |
| "loss": 0.0037, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "grad_norm": 0.001104432623833418, | |
| "learning_rate": 1.8518518518518518e-05, | |
| "loss": 0.0995, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "eval_accuracy": 0.9305555555555556, | |
| "eval_loss": 0.1180504709482193, | |
| "eval_runtime": 57.5974, | |
| "eval_samples_per_second": 1.25, | |
| "eval_steps_per_second": 1.25, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "grad_norm": 0.003288655774667859, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.0003, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "grad_norm": 0.018307456746697426, | |
| "learning_rate": 1.4814814814814815e-05, | |
| "loss": 0.0002, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 0.012207794934511185, | |
| "learning_rate": 1.2962962962962962e-05, | |
| "loss": 0.0004, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 4.13, | |
| "grad_norm": 0.001230477704666555, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 0.0007, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "grad_norm": 0.001645821612328291, | |
| "learning_rate": 9.259259259259259e-06, | |
| "loss": 0.0001, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "eval_accuracy": 0.9166666666666666, | |
| "eval_loss": 0.17973226308822632, | |
| "eval_runtime": 57.4339, | |
| "eval_samples_per_second": 1.254, | |
| "eval_steps_per_second": 1.254, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "grad_norm": 0.04015364125370979, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 0.0002, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 5.07, | |
| "grad_norm": 0.10828184336423874, | |
| "learning_rate": 5.555555555555556e-06, | |
| "loss": 0.0002, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 0.0032853526063263416, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 0.0006, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 5.13, | |
| "grad_norm": 0.005327914375811815, | |
| "learning_rate": 1.8518518518518519e-06, | |
| "loss": 0.0825, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 5.17, | |
| "grad_norm": 0.004417418502271175, | |
| "learning_rate": 0.0, | |
| "loss": 0.0003, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.17, | |
| "eval_accuracy": 0.9444444444444444, | |
| "eval_loss": 0.13933943212032318, | |
| "eval_runtime": 56.7608, | |
| "eval_samples_per_second": 1.268, | |
| "eval_steps_per_second": 1.268, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.17, | |
| "step": 300, | |
| "total_flos": 3.28565653420032e+17, | |
| "train_loss": 0.056568971110973504, | |
| "train_runtime": 562.6502, | |
| "train_samples_per_second": 0.533, | |
| "train_steps_per_second": 0.533 | |
| }, | |
| { | |
| "epoch": 5.17, | |
| "eval_accuracy": 0.9117647058823529, | |
| "eval_loss": 0.37310144305229187, | |
| "eval_runtime": 34.9776, | |
| "eval_samples_per_second": 1.944, | |
| "eval_steps_per_second": 1.944, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.17, | |
| "eval_accuracy": 0.9117647058823529, | |
| "eval_loss": 0.37310147285461426, | |
| "eval_runtime": 33.7759, | |
| "eval_samples_per_second": 2.013, | |
| "eval_steps_per_second": 2.013, | |
| "step": 300 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 300, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 500, | |
| "total_flos": 3.28565653420032e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |