{ "best_metric": 0.7972972972972973, "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-new_dataset_50e/checkpoint-108", "epoch": 49.94117647058823, "global_step": 200, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.94, "eval_accuracy": 0.6081081081081081, "eval_loss": 0.7080705165863037, "eval_runtime": 1.9058, "eval_samples_per_second": 38.828, "eval_steps_per_second": 1.574, "step": 4 }, { "epoch": 1.94, "eval_accuracy": 0.6351351351351351, "eval_loss": 0.7103850841522217, "eval_runtime": 1.9278, "eval_samples_per_second": 38.387, "eval_steps_per_second": 1.556, "step": 8 }, { "epoch": 2.47, "learning_rate": 2.5e-05, "loss": 0.5516, "step": 10 }, { "epoch": 2.94, "eval_accuracy": 0.6351351351351351, "eval_loss": 0.6910790801048279, "eval_runtime": 1.9434, "eval_samples_per_second": 38.078, "eval_steps_per_second": 1.544, "step": 12 }, { "epoch": 3.94, "eval_accuracy": 0.7027027027027027, "eval_loss": 0.7155552506446838, "eval_runtime": 1.7624, "eval_samples_per_second": 41.988, "eval_steps_per_second": 1.702, "step": 16 }, { "epoch": 4.94, "learning_rate": 5e-05, "loss": 0.537, "step": 20 }, { "epoch": 4.94, "eval_accuracy": 0.7297297297297297, "eval_loss": 0.7345479130744934, "eval_runtime": 1.372, "eval_samples_per_second": 53.938, "eval_steps_per_second": 2.187, "step": 20 }, { "epoch": 5.94, "eval_accuracy": 0.6891891891891891, "eval_loss": 0.6745122671127319, "eval_runtime": 1.2553, "eval_samples_per_second": 58.952, "eval_steps_per_second": 2.39, "step": 24 }, { "epoch": 6.94, "eval_accuracy": 0.7297297297297297, "eval_loss": 0.7145545482635498, "eval_runtime": 1.401, "eval_samples_per_second": 52.821, "eval_steps_per_second": 2.141, "step": 28 }, { "epoch": 7.47, "learning_rate": 4.722222222222222e-05, "loss": 0.5333, "step": 30 }, { "epoch": 7.94, "eval_accuracy": 0.6891891891891891, "eval_loss": 0.7056529521942139, "eval_runtime": 1.2629, "eval_samples_per_second": 58.596, "eval_steps_per_second": 2.376, "step": 32 }, { "epoch": 8.94, "eval_accuracy": 0.7027027027027027, "eval_loss": 0.653145432472229, "eval_runtime": 1.3937, "eval_samples_per_second": 53.096, "eval_steps_per_second": 2.153, "step": 36 }, { "epoch": 9.94, "learning_rate": 4.4444444444444447e-05, "loss": 0.4871, "step": 40 }, { "epoch": 9.94, "eval_accuracy": 0.7027027027027027, "eval_loss": 0.6404768228530884, "eval_runtime": 1.2576, "eval_samples_per_second": 58.844, "eval_steps_per_second": 2.386, "step": 40 }, { "epoch": 10.94, "eval_accuracy": 0.6891891891891891, "eval_loss": 0.612636148929596, "eval_runtime": 1.3946, "eval_samples_per_second": 53.061, "eval_steps_per_second": 2.151, "step": 44 }, { "epoch": 11.94, "eval_accuracy": 0.7027027027027027, "eval_loss": 0.6303472518920898, "eval_runtime": 1.625, "eval_samples_per_second": 45.538, "eval_steps_per_second": 1.846, "step": 48 }, { "epoch": 12.47, "learning_rate": 4.166666666666667e-05, "loss": 0.4432, "step": 50 }, { "epoch": 12.94, "eval_accuracy": 0.7027027027027027, "eval_loss": 0.6263724565505981, "eval_runtime": 1.9541, "eval_samples_per_second": 37.87, "eval_steps_per_second": 1.535, "step": 52 }, { "epoch": 13.94, "eval_accuracy": 0.7432432432432432, "eval_loss": 0.6347029209136963, "eval_runtime": 1.8037, "eval_samples_per_second": 41.026, "eval_steps_per_second": 1.663, "step": 56 }, { "epoch": 14.94, "learning_rate": 3.888888888888889e-05, "loss": 0.3669, "step": 60 }, { "epoch": 14.94, "eval_accuracy": 0.6621621621621622, "eval_loss": 0.6698160171508789, "eval_runtime": 1.964, "eval_samples_per_second": 37.678, "eval_steps_per_second": 1.527, "step": 60 }, { "epoch": 15.94, "eval_accuracy": 0.7567567567567568, "eval_loss": 0.6345672011375427, "eval_runtime": 1.9075, "eval_samples_per_second": 38.794, "eval_steps_per_second": 1.573, "step": 64 }, { "epoch": 16.94, "eval_accuracy": 0.6891891891891891, "eval_loss": 0.6509883999824524, "eval_runtime": 1.9133, "eval_samples_per_second": 38.676, "eval_steps_per_second": 1.568, "step": 68 }, { "epoch": 17.47, "learning_rate": 3.611111111111111e-05, "loss": 0.3704, "step": 70 }, { "epoch": 17.94, "eval_accuracy": 0.6891891891891891, "eval_loss": 0.6491345763206482, "eval_runtime": 1.8115, "eval_samples_per_second": 40.849, "eval_steps_per_second": 1.656, "step": 72 }, { "epoch": 18.94, "eval_accuracy": 0.7567567567567568, "eval_loss": 0.5946679711341858, "eval_runtime": 1.2941, "eval_samples_per_second": 57.183, "eval_steps_per_second": 2.318, "step": 76 }, { "epoch": 19.94, "learning_rate": 3.3333333333333335e-05, "loss": 0.3624, "step": 80 }, { "epoch": 19.94, "eval_accuracy": 0.7027027027027027, "eval_loss": 0.6247528195381165, "eval_runtime": 1.3841, "eval_samples_per_second": 53.465, "eval_steps_per_second": 2.167, "step": 80 }, { "epoch": 20.94, "eval_accuracy": 0.7027027027027027, "eval_loss": 0.6580175161361694, "eval_runtime": 1.2688, "eval_samples_per_second": 58.325, "eval_steps_per_second": 2.365, "step": 84 }, { "epoch": 21.94, "eval_accuracy": 0.7162162162162162, "eval_loss": 0.6344525814056396, "eval_runtime": 1.4991, "eval_samples_per_second": 49.363, "eval_steps_per_second": 2.001, "step": 88 }, { "epoch": 22.47, "learning_rate": 3.055555555555556e-05, "loss": 0.3164, "step": 90 }, { "epoch": 22.94, "eval_accuracy": 0.7567567567567568, "eval_loss": 0.6092268228530884, "eval_runtime": 1.8236, "eval_samples_per_second": 40.578, "eval_steps_per_second": 1.645, "step": 92 }, { "epoch": 23.94, "eval_accuracy": 0.7162162162162162, "eval_loss": 0.6497542858123779, "eval_runtime": 1.3039, "eval_samples_per_second": 56.753, "eval_steps_per_second": 2.301, "step": 96 }, { "epoch": 24.94, "learning_rate": 2.777777777777778e-05, "loss": 0.2777, "step": 100 }, { "epoch": 24.94, "eval_accuracy": 0.7702702702702703, "eval_loss": 0.6914950013160706, "eval_runtime": 1.2986, "eval_samples_per_second": 56.986, "eval_steps_per_second": 2.31, "step": 100 }, { "epoch": 25.94, "eval_accuracy": 0.7837837837837838, "eval_loss": 0.6481638550758362, "eval_runtime": 1.3273, "eval_samples_per_second": 55.751, "eval_steps_per_second": 2.26, "step": 104 }, { "epoch": 26.94, "eval_accuracy": 0.7972972972972973, "eval_loss": 0.640661895275116, "eval_runtime": 1.8593, "eval_samples_per_second": 39.8, "eval_steps_per_second": 1.614, "step": 108 }, { "epoch": 27.47, "learning_rate": 2.5e-05, "loss": 0.2946, "step": 110 }, { "epoch": 27.94, "eval_accuracy": 0.7837837837837838, "eval_loss": 0.6135300993919373, "eval_runtime": 1.6274, "eval_samples_per_second": 45.472, "eval_steps_per_second": 1.843, "step": 112 }, { "epoch": 28.94, "eval_accuracy": 0.7567567567567568, "eval_loss": 0.6818699240684509, "eval_runtime": 1.2554, "eval_samples_per_second": 58.947, "eval_steps_per_second": 2.39, "step": 116 }, { "epoch": 29.94, "learning_rate": 2.2222222222222223e-05, "loss": 0.2546, "step": 120 }, { "epoch": 29.94, "eval_accuracy": 0.7567567567567568, "eval_loss": 0.6401073932647705, "eval_runtime": 1.3679, "eval_samples_per_second": 54.098, "eval_steps_per_second": 2.193, "step": 120 }, { "epoch": 30.94, "eval_accuracy": 0.7432432432432432, "eval_loss": 0.6369650363922119, "eval_runtime": 1.2605, "eval_samples_per_second": 58.708, "eval_steps_per_second": 2.38, "step": 124 }, { "epoch": 31.94, "eval_accuracy": 0.7702702702702703, "eval_loss": 0.6487568616867065, "eval_runtime": 1.3752, "eval_samples_per_second": 53.81, "eval_steps_per_second": 2.182, "step": 128 }, { "epoch": 32.47, "learning_rate": 1.9444444444444445e-05, "loss": 0.2477, "step": 130 }, { "epoch": 32.94, "eval_accuracy": 0.7972972972972973, "eval_loss": 0.6429384350776672, "eval_runtime": 1.86, "eval_samples_per_second": 39.785, "eval_steps_per_second": 1.613, "step": 132 }, { "epoch": 33.94, "eval_accuracy": 0.7702702702702703, "eval_loss": 0.6539828181266785, "eval_runtime": 1.2683, "eval_samples_per_second": 58.344, "eval_steps_per_second": 2.365, "step": 136 }, { "epoch": 34.94, "learning_rate": 1.6666666666666667e-05, "loss": 0.1968, "step": 140 }, { "epoch": 34.94, "eval_accuracy": 0.7972972972972973, "eval_loss": 0.5895383358001709, "eval_runtime": 1.3392, "eval_samples_per_second": 55.257, "eval_steps_per_second": 2.24, "step": 140 }, { "epoch": 35.94, "eval_accuracy": 0.7567567567567568, "eval_loss": 0.6242156624794006, "eval_runtime": 1.29, "eval_samples_per_second": 57.364, "eval_steps_per_second": 2.326, "step": 144 }, { "epoch": 36.94, "eval_accuracy": 0.7567567567567568, "eval_loss": 0.6575082540512085, "eval_runtime": 1.6409, "eval_samples_per_second": 45.097, "eval_steps_per_second": 1.828, "step": 148 }, { "epoch": 37.47, "learning_rate": 1.388888888888889e-05, "loss": 0.2235, "step": 150 }, { "epoch": 37.94, "eval_accuracy": 0.7702702702702703, "eval_loss": 0.6263442039489746, "eval_runtime": 1.655, "eval_samples_per_second": 44.714, "eval_steps_per_second": 1.813, "step": 152 }, { "epoch": 38.94, "eval_accuracy": 0.7837837837837838, "eval_loss": 0.6224556565284729, "eval_runtime": 1.2758, "eval_samples_per_second": 58.004, "eval_steps_per_second": 2.352, "step": 156 }, { "epoch": 39.94, "learning_rate": 1.1111111111111112e-05, "loss": 0.2005, "step": 160 }, { "epoch": 39.94, "eval_accuracy": 0.7702702702702703, "eval_loss": 0.673054039478302, "eval_runtime": 1.3797, "eval_samples_per_second": 53.637, "eval_steps_per_second": 2.174, "step": 160 }, { "epoch": 40.94, "eval_accuracy": 0.7702702702702703, "eval_loss": 0.684382438659668, "eval_runtime": 1.2635, "eval_samples_per_second": 58.566, "eval_steps_per_second": 2.374, "step": 164 }, { "epoch": 41.94, "eval_accuracy": 0.7702702702702703, "eval_loss": 0.65503990650177, "eval_runtime": 1.2866, "eval_samples_per_second": 57.518, "eval_steps_per_second": 2.332, "step": 168 }, { "epoch": 42.47, "learning_rate": 8.333333333333334e-06, "loss": 0.2062, "step": 170 }, { "epoch": 42.94, "eval_accuracy": 0.7702702702702703, "eval_loss": 0.6700405478477478, "eval_runtime": 1.2839, "eval_samples_per_second": 57.635, "eval_steps_per_second": 2.337, "step": 172 }, { "epoch": 43.94, "eval_accuracy": 0.7702702702702703, "eval_loss": 0.6660715341567993, "eval_runtime": 1.309, "eval_samples_per_second": 56.532, "eval_steps_per_second": 2.292, "step": 176 }, { "epoch": 44.94, "learning_rate": 5.555555555555556e-06, "loss": 0.1933, "step": 180 }, { "epoch": 44.94, "eval_accuracy": 0.7837837837837838, "eval_loss": 0.66064453125, "eval_runtime": 1.2556, "eval_samples_per_second": 58.934, "eval_steps_per_second": 2.389, "step": 180 }, { "epoch": 45.94, "eval_accuracy": 0.7702702702702703, "eval_loss": 0.6756728291511536, "eval_runtime": 1.3709, "eval_samples_per_second": 53.978, "eval_steps_per_second": 2.188, "step": 184 }, { "epoch": 46.94, "eval_accuracy": 0.7567567567567568, "eval_loss": 0.688907265663147, "eval_runtime": 1.2332, "eval_samples_per_second": 60.008, "eval_steps_per_second": 2.433, "step": 188 }, { "epoch": 47.47, "learning_rate": 2.777777777777778e-06, "loss": 0.1895, "step": 190 }, { "epoch": 47.94, "eval_accuracy": 0.7567567567567568, "eval_loss": 0.6939808130264282, "eval_runtime": 1.3438, "eval_samples_per_second": 55.068, "eval_steps_per_second": 2.232, "step": 192 }, { "epoch": 48.94, "eval_accuracy": 0.7567567567567568, "eval_loss": 0.6919070482254028, "eval_runtime": 1.2377, "eval_samples_per_second": 59.787, "eval_steps_per_second": 2.424, "step": 196 }, { "epoch": 49.94, "learning_rate": 0.0, "loss": 0.1666, "step": 200 }, { "epoch": 49.94, "eval_accuracy": 0.7432432432432432, "eval_loss": 0.6898728609085083, "eval_runtime": 1.3948, "eval_samples_per_second": 53.056, "eval_steps_per_second": 2.151, "step": 200 }, { "epoch": 49.94, "step": 200, "total_flos": 6.387670531495895e+17, "train_loss": 0.3209643447399139, "train_runtime": 2433.6037, "train_samples_per_second": 10.56, "train_steps_per_second": 0.082 } ], "max_steps": 200, "num_train_epochs": 50, "total_flos": 6.387670531495895e+17, "trial_name": null, "trial_params": null }