{ "best_metric": 0.9586206896551724, "best_model_checkpoint": "videomae-base-finetuned-soccer-action-recognition/checkpoint-1790", "epoch": 31.03125, "eval_steps": 500, "global_step": 2728, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.03, "learning_rate": 1.4652014652014653e-05, "loss": 1.7115, "step": 80 }, { "epoch": 0.03, "eval_accuracy": 0.4, "eval_loss": 1.4195746183395386, "eval_runtime": 150.0417, "eval_samples_per_second": 0.966, "eval_steps_per_second": 0.966, "step": 85 }, { "epoch": 1.03, "learning_rate": 2.9304029304029305e-05, "loss": 1.0097, "step": 160 }, { "epoch": 1.03, "eval_accuracy": 0.6758620689655173, "eval_loss": 0.7807376980781555, "eval_runtime": 30.3768, "eval_samples_per_second": 4.773, "eval_steps_per_second": 4.773, "step": 170 }, { "epoch": 2.03, "learning_rate": 4.3956043956043955e-05, "loss": 0.6192, "step": 240 }, { "epoch": 2.03, "eval_accuracy": 0.7034482758620689, "eval_loss": 0.795186460018158, "eval_runtime": 30.3396, "eval_samples_per_second": 4.779, "eval_steps_per_second": 4.779, "step": 255 }, { "epoch": 3.02, "learning_rate": 4.904276985743381e-05, "loss": 0.4713, "step": 320 }, { "epoch": 3.03, "eval_accuracy": 0.7931034482758621, "eval_loss": 0.6536410450935364, "eval_runtime": 28.0955, "eval_samples_per_second": 5.161, "eval_steps_per_second": 5.161, "step": 341 }, { "epoch": 4.02, "learning_rate": 4.7413441955193486e-05, "loss": 0.3973, "step": 400 }, { "epoch": 4.03, "eval_accuracy": 0.8689655172413793, "eval_loss": 0.3637762665748596, "eval_runtime": 28.0719, "eval_samples_per_second": 5.165, "eval_steps_per_second": 5.165, "step": 426 }, { "epoch": 5.02, "learning_rate": 4.578411405295316e-05, "loss": 0.3633, "step": 480 }, { "epoch": 5.03, "eval_accuracy": 0.896551724137931, "eval_loss": 0.36160850524902344, "eval_runtime": 28.2429, "eval_samples_per_second": 5.134, "eval_steps_per_second": 5.134, "step": 511 }, { "epoch": 6.02, "learning_rate": 4.415478615071284e-05, "loss": 0.2336, "step": 560 }, { "epoch": 6.03, "eval_accuracy": 0.896551724137931, "eval_loss": 0.4579198360443115, "eval_runtime": 30.0911, "eval_samples_per_second": 4.819, "eval_steps_per_second": 4.819, "step": 596 }, { "epoch": 7.02, "learning_rate": 4.2525458248472504e-05, "loss": 0.1997, "step": 640 }, { "epoch": 7.03, "eval_accuracy": 0.6068965517241379, "eval_loss": 1.597042202949524, "eval_runtime": 27.9534, "eval_samples_per_second": 5.187, "eval_steps_per_second": 5.187, "step": 682 }, { "epoch": 8.01, "learning_rate": 4.089613034623218e-05, "loss": 0.2738, "step": 720 }, { "epoch": 8.03, "eval_accuracy": 0.8689655172413793, "eval_loss": 0.4101611375808716, "eval_runtime": 27.8973, "eval_samples_per_second": 5.198, "eval_steps_per_second": 5.198, "step": 767 }, { "epoch": 9.01, "learning_rate": 3.9266802443991856e-05, "loss": 0.2492, "step": 800 }, { "epoch": 9.03, "eval_accuracy": 0.8344827586206897, "eval_loss": 0.7651154398918152, "eval_runtime": 29.2625, "eval_samples_per_second": 4.955, "eval_steps_per_second": 4.955, "step": 852 }, { "epoch": 10.01, "learning_rate": 3.763747454175153e-05, "loss": 0.1568, "step": 880 }, { "epoch": 10.03, "eval_accuracy": 0.8137931034482758, "eval_loss": 0.8560731410980225, "eval_runtime": 30.5796, "eval_samples_per_second": 4.742, "eval_steps_per_second": 4.742, "step": 937 }, { "epoch": 11.01, "learning_rate": 3.60081466395112e-05, "loss": 0.1856, "step": 960 }, { "epoch": 11.03, "eval_accuracy": 0.9241379310344827, "eval_loss": 0.2810556888580322, "eval_runtime": 27.9183, "eval_samples_per_second": 5.194, "eval_steps_per_second": 5.194, "step": 1023 }, { "epoch": 12.01, "learning_rate": 3.437881873727088e-05, "loss": 0.1296, "step": 1040 }, { "epoch": 12.03, "eval_accuracy": 0.9172413793103448, "eval_loss": 0.3444097340106964, "eval_runtime": 29.2588, "eval_samples_per_second": 4.956, "eval_steps_per_second": 4.956, "step": 1108 }, { "epoch": 13.0, "learning_rate": 3.274949083503055e-05, "loss": 0.0782, "step": 1120 }, { "epoch": 13.03, "eval_accuracy": 0.9241379310344827, "eval_loss": 0.34232720732688904, "eval_runtime": 29.8187, "eval_samples_per_second": 4.863, "eval_steps_per_second": 4.863, "step": 1193 }, { "epoch": 14.0, "learning_rate": 3.1120162932790225e-05, "loss": 0.14, "step": 1200 }, { "epoch": 14.03, "eval_accuracy": 0.9241379310344827, "eval_loss": 0.31215447187423706, "eval_runtime": 30.9503, "eval_samples_per_second": 4.685, "eval_steps_per_second": 4.685, "step": 1278 }, { "epoch": 15.0, "learning_rate": 2.94908350305499e-05, "loss": 0.0802, "step": 1280 }, { "epoch": 15.03, "learning_rate": 2.786150712830957e-05, "loss": 0.0689, "step": 1360 }, { "epoch": 15.03, "eval_accuracy": 0.9172413793103448, "eval_loss": 0.35335081815719604, "eval_runtime": 30.933, "eval_samples_per_second": 4.688, "eval_steps_per_second": 4.688, "step": 1364 }, { "epoch": 16.03, "learning_rate": 2.6232179226069247e-05, "loss": 0.036, "step": 1440 }, { "epoch": 16.03, "eval_accuracy": 0.9103448275862069, "eval_loss": 0.4814639985561371, "eval_runtime": 30.7939, "eval_samples_per_second": 4.709, "eval_steps_per_second": 4.709, "step": 1449 }, { "epoch": 17.03, "learning_rate": 2.4602851323828923e-05, "loss": 0.0695, "step": 1520 }, { "epoch": 17.03, "eval_accuracy": 0.8827586206896552, "eval_loss": 0.5698482394218445, "eval_runtime": 30.7434, "eval_samples_per_second": 4.716, "eval_steps_per_second": 4.716, "step": 1534 }, { "epoch": 18.02, "learning_rate": 2.29735234215886e-05, "loss": 0.0618, "step": 1600 }, { "epoch": 18.03, "eval_accuracy": 0.9310344827586207, "eval_loss": 0.3052562177181244, "eval_runtime": 29.2874, "eval_samples_per_second": 4.951, "eval_steps_per_second": 4.951, "step": 1619 }, { "epoch": 19.02, "learning_rate": 2.134419551934827e-05, "loss": 0.0553, "step": 1680 }, { "epoch": 19.03, "eval_accuracy": 0.9241379310344827, "eval_loss": 0.3443007469177246, "eval_runtime": 30.1383, "eval_samples_per_second": 4.811, "eval_steps_per_second": 4.811, "step": 1705 }, { "epoch": 20.02, "learning_rate": 1.9714867617107944e-05, "loss": 0.0301, "step": 1760 }, { "epoch": 20.03, "eval_accuracy": 0.9586206896551724, "eval_loss": 0.1426996886730194, "eval_runtime": 29.78, "eval_samples_per_second": 4.869, "eval_steps_per_second": 4.869, "step": 1790 }, { "epoch": 21.02, "learning_rate": 1.808553971486762e-05, "loss": 0.0412, "step": 1840 }, { "epoch": 21.03, "eval_accuracy": 0.8689655172413793, "eval_loss": 0.5618650317192078, "eval_runtime": 29.9124, "eval_samples_per_second": 4.847, "eval_steps_per_second": 4.847, "step": 1875 }, { "epoch": 22.02, "learning_rate": 1.6456211812627292e-05, "loss": 0.0492, "step": 1920 }, { "epoch": 22.03, "eval_accuracy": 0.8896551724137931, "eval_loss": 0.5700664520263672, "eval_runtime": 27.7486, "eval_samples_per_second": 5.225, "eval_steps_per_second": 5.225, "step": 1960 }, { "epoch": 23.01, "learning_rate": 1.4826883910386965e-05, "loss": 0.0171, "step": 2000 }, { "epoch": 23.03, "eval_accuracy": 0.8689655172413793, "eval_loss": 0.6377372145652771, "eval_runtime": 29.1792, "eval_samples_per_second": 4.969, "eval_steps_per_second": 4.969, "step": 2046 }, { "epoch": 24.01, "learning_rate": 1.3197556008146641e-05, "loss": 0.0181, "step": 2080 }, { "epoch": 24.03, "eval_accuracy": 0.8827586206896552, "eval_loss": 0.5981259346008301, "eval_runtime": 30.3638, "eval_samples_per_second": 4.775, "eval_steps_per_second": 4.775, "step": 2131 }, { "epoch": 25.01, "learning_rate": 1.1568228105906315e-05, "loss": 0.0305, "step": 2160 }, { "epoch": 25.03, "eval_accuracy": 0.9448275862068966, "eval_loss": 0.3177809715270996, "eval_runtime": 30.2408, "eval_samples_per_second": 4.795, "eval_steps_per_second": 4.795, "step": 2216 }, { "epoch": 26.01, "learning_rate": 9.938900203665988e-06, "loss": 0.0393, "step": 2240 }, { "epoch": 26.03, "eval_accuracy": 0.9103448275862069, "eval_loss": 0.5434169173240662, "eval_runtime": 29.452, "eval_samples_per_second": 4.923, "eval_steps_per_second": 4.923, "step": 2301 }, { "epoch": 27.01, "learning_rate": 8.309572301425662e-06, "loss": 0.0248, "step": 2320 }, { "epoch": 27.03, "eval_accuracy": 0.9241379310344827, "eval_loss": 0.40973013639450073, "eval_runtime": 30.9246, "eval_samples_per_second": 4.689, "eval_steps_per_second": 4.689, "step": 2387 }, { "epoch": 28.0, "learning_rate": 6.6802443991853366e-06, "loss": 0.0146, "step": 2400 }, { "epoch": 28.03, "eval_accuracy": 0.9103448275862069, "eval_loss": 0.44273480772972107, "eval_runtime": 29.0928, "eval_samples_per_second": 4.984, "eval_steps_per_second": 4.984, "step": 2472 }, { "epoch": 29.0, "learning_rate": 5.05091649694501e-06, "loss": 0.012, "step": 2480 }, { "epoch": 29.03, "eval_accuracy": 0.903448275862069, "eval_loss": 0.5618546605110168, "eval_runtime": 29.5423, "eval_samples_per_second": 4.908, "eval_steps_per_second": 4.908, "step": 2557 }, { "epoch": 30.0, "learning_rate": 3.4215885947046847e-06, "loss": 0.0036, "step": 2560 }, { "epoch": 30.03, "learning_rate": 1.7922606924643586e-06, "loss": 0.0065, "step": 2640 }, { "epoch": 30.03, "eval_accuracy": 0.9103448275862069, "eval_loss": 0.5383774042129517, "eval_runtime": 30.5686, "eval_samples_per_second": 4.743, "eval_steps_per_second": 4.743, "step": 2642 }, { "epoch": 31.03, "learning_rate": 1.629327902240326e-07, "loss": 0.009, "step": 2720 }, { "epoch": 31.03, "eval_accuracy": 0.9172413793103448, "eval_loss": 0.5013585686683655, "eval_runtime": 30.4753, "eval_samples_per_second": 4.758, "eval_steps_per_second": 4.758, "step": 2728 }, { "epoch": 31.03, "step": 2728, "total_flos": 2.719508368433401e+19, "train_loss": 0.20195167288498247, "train_runtime": 11757.4892, "train_samples_per_second": 1.856, "train_steps_per_second": 0.232 }, { "epoch": 31.03, "eval_accuracy": 0.9470198675496688, "eval_loss": 0.2553568482398987, "eval_runtime": 163.4914, "eval_samples_per_second": 0.924, "eval_steps_per_second": 0.924, "step": 2728 }, { "epoch": 31.03, "eval_accuracy": 0.9470198675496688, "eval_loss": 0.2553568482398987, "eval_runtime": 31.773, "eval_samples_per_second": 4.752, "eval_steps_per_second": 4.752, "step": 2728 } ], "logging_steps": 80, "max_steps": 2728, "num_train_epochs": 9223372036854775807, "save_steps": 500, "total_flos": 2.719508368433401e+19, "trial_name": null, "trial_params": null }