| { | |
| "best_metric": 0.775, | |
| "best_model_checkpoint": "videomae-base-finetuned-rwf2000-subset___v1/checkpoint-500", | |
| "epoch": 4.2, | |
| "eval_steps": 500, | |
| "global_step": 500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 6.027842044830322, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.6934, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 5.1828837394714355, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.7343, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 6.022104740142822, | |
| "learning_rate": 6e-06, | |
| "loss": 0.6867, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.9819915294647217, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.6967, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 3.980625867843628, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6699, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 6.501969337463379, | |
| "learning_rate": 9.777777777777779e-06, | |
| "loss": 0.6533, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 3.093120813369751, | |
| "learning_rate": 9.555555555555556e-06, | |
| "loss": 0.627, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 6.481435775756836, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 0.6128, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 5.1239118576049805, | |
| "learning_rate": 9.111111111111112e-06, | |
| "loss": 0.5995, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 3.803208351135254, | |
| "learning_rate": 8.888888888888888e-06, | |
| "loss": 0.5205, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "eval_accuracy": 0.68375, | |
| "eval_loss": 0.5434452891349792, | |
| "eval_runtime": 225.7874, | |
| "eval_samples_per_second": 3.543, | |
| "eval_steps_per_second": 0.221, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 4.416550159454346, | |
| "learning_rate": 8.666666666666668e-06, | |
| "loss": 0.5088, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 6.156940460205078, | |
| "learning_rate": 8.444444444444446e-06, | |
| "loss": 0.5568, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 5.02749490737915, | |
| "learning_rate": 8.222222222222222e-06, | |
| "loss": 0.5033, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 5.65871524810791, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.4887, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 5.758727550506592, | |
| "learning_rate": 7.77777777777778e-06, | |
| "loss": 0.5106, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 11.959117889404297, | |
| "learning_rate": 7.555555555555556e-06, | |
| "loss": 0.5293, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.1400000000000001, | |
| "grad_norm": 5.07788610458374, | |
| "learning_rate": 7.333333333333333e-06, | |
| "loss": 0.4665, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 4.521602153778076, | |
| "learning_rate": 7.111111111111112e-06, | |
| "loss": 0.4658, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 5.800256252288818, | |
| "learning_rate": 6.88888888888889e-06, | |
| "loss": 0.5002, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 3.6714704036712646, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.4084, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "eval_accuracy": 0.655, | |
| "eval_loss": 0.5904867053031921, | |
| "eval_runtime": 247.6204, | |
| "eval_samples_per_second": 3.231, | |
| "eval_steps_per_second": 0.202, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 6.782428741455078, | |
| "learning_rate": 6.444444444444445e-06, | |
| "loss": 0.4441, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 4.966807842254639, | |
| "learning_rate": 6.222222222222223e-06, | |
| "loss": 0.4307, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 8.04571533203125, | |
| "learning_rate": 6e-06, | |
| "loss": 0.4022, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 6.805935859680176, | |
| "learning_rate": 5.777777777777778e-06, | |
| "loss": 0.4298, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 6.641063213348389, | |
| "learning_rate": 5.555555555555557e-06, | |
| "loss": 0.4494, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 6.542612552642822, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 0.4735, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "grad_norm": 6.421093940734863, | |
| "learning_rate": 5.1111111111111115e-06, | |
| "loss": 0.3762, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 5.869024753570557, | |
| "learning_rate": 4.888888888888889e-06, | |
| "loss": 0.4387, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 9.899517059326172, | |
| "learning_rate": 4.666666666666667e-06, | |
| "loss": 0.4453, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 11.099763870239258, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 0.4198, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "eval_accuracy": 0.74625, | |
| "eval_loss": 0.48142582178115845, | |
| "eval_runtime": 271.5798, | |
| "eval_samples_per_second": 2.946, | |
| "eval_steps_per_second": 0.184, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 8.095602989196777, | |
| "learning_rate": 4.222222222222223e-06, | |
| "loss": 0.4279, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "grad_norm": 5.949367046356201, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.3158, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "grad_norm": 10.842527389526367, | |
| "learning_rate": 3.777777777777778e-06, | |
| "loss": 0.3554, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "grad_norm": 5.070864677429199, | |
| "learning_rate": 3.555555555555556e-06, | |
| "loss": 0.2949, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 15.597376823425293, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 0.3548, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 4.660379409790039, | |
| "learning_rate": 3.1111111111111116e-06, | |
| "loss": 0.4349, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "grad_norm": 11.151165008544922, | |
| "learning_rate": 2.888888888888889e-06, | |
| "loss": 0.3602, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "grad_norm": 9.200762748718262, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 0.3891, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "grad_norm": 6.462916851043701, | |
| "learning_rate": 2.4444444444444447e-06, | |
| "loss": 0.2933, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 8.92557144165039, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 0.3188, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "eval_accuracy": 0.755, | |
| "eval_loss": 0.5160250663757324, | |
| "eval_runtime": 225.8295, | |
| "eval_samples_per_second": 3.542, | |
| "eval_steps_per_second": 0.221, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 7.164823532104492, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.2785, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 4.04, | |
| "grad_norm": 4.740599155426025, | |
| "learning_rate": 1.777777777777778e-06, | |
| "loss": 0.3388, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 4.06, | |
| "grad_norm": 8.506775856018066, | |
| "learning_rate": 1.5555555555555558e-06, | |
| "loss": 0.3093, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "grad_norm": 4.401611804962158, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 0.2399, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 7.568962574005127, | |
| "learning_rate": 1.111111111111111e-06, | |
| "loss": 0.3434, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 4.12, | |
| "grad_norm": 9.326196670532227, | |
| "learning_rate": 8.88888888888889e-07, | |
| "loss": 0.2818, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 4.14, | |
| "grad_norm": 6.332729816436768, | |
| "learning_rate": 6.666666666666667e-07, | |
| "loss": 0.3154, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 11.177745819091797, | |
| "learning_rate": 4.444444444444445e-07, | |
| "loss": 0.3217, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "grad_norm": 6.826053619384766, | |
| "learning_rate": 2.2222222222222224e-07, | |
| "loss": 0.3167, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 2.244813919067383, | |
| "learning_rate": 0.0, | |
| "loss": 0.2687, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "eval_accuracy": 0.775, | |
| "eval_loss": 0.48075544834136963, | |
| "eval_runtime": 229.5259, | |
| "eval_samples_per_second": 3.485, | |
| "eval_steps_per_second": 0.218, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "step": 500, | |
| "total_flos": 9.968516715773952e+18, | |
| "train_loss": 0.4460222806930542, | |
| "train_runtime": 4184.8427, | |
| "train_samples_per_second": 1.912, | |
| "train_steps_per_second": 0.119 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "eval_accuracy": 0.775, | |
| "eval_loss": 0.48075541853904724, | |
| "eval_runtime": 225.4225, | |
| "eval_samples_per_second": 3.549, | |
| "eval_steps_per_second": 0.222, | |
| "step": 500 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "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": 9.968516715773952e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |