{ "best_metric": 0.898876404494382, "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-eurosat/checkpoint-52", "epoch": 27.692307692307693, "eval_steps": 500, "global_step": 90, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.9230769230769231, "eval_accuracy": 0.24719101123595505, "eval_loss": 1.8888534307479858, "eval_runtime": 0.8953, "eval_samples_per_second": 99.405, "eval_steps_per_second": 2.234, "step": 3 }, { "epoch": 1.8461538461538463, "eval_accuracy": 0.38202247191011235, "eval_loss": 1.7625091075897217, "eval_runtime": 0.4339, "eval_samples_per_second": 205.122, "eval_steps_per_second": 4.609, "step": 6 }, { "epoch": 2.769230769230769, "eval_accuracy": 0.449438202247191, "eval_loss": 1.5602569580078125, "eval_runtime": 0.4301, "eval_samples_per_second": 206.928, "eval_steps_per_second": 4.65, "step": 9 }, { "epoch": 3.076923076923077, "grad_norm": 24.15924644470215, "learning_rate": 4.938271604938271e-05, "loss": 1.7854, "step": 10 }, { "epoch": 4.0, "eval_accuracy": 0.5280898876404494, "eval_loss": 1.3004920482635498, "eval_runtime": 0.433, "eval_samples_per_second": 205.529, "eval_steps_per_second": 4.619, "step": 13 }, { "epoch": 4.923076923076923, "eval_accuracy": 0.6292134831460674, "eval_loss": 1.0408316850662231, "eval_runtime": 0.4328, "eval_samples_per_second": 205.62, "eval_steps_per_second": 4.621, "step": 16 }, { "epoch": 5.846153846153846, "eval_accuracy": 0.6853932584269663, "eval_loss": 0.8925455212593079, "eval_runtime": 0.4391, "eval_samples_per_second": 202.683, "eval_steps_per_second": 4.555, "step": 19 }, { "epoch": 6.153846153846154, "grad_norm": 42.52753448486328, "learning_rate": 4.3209876543209875e-05, "loss": 1.1431, "step": 20 }, { "epoch": 6.769230769230769, "eval_accuracy": 0.7303370786516854, "eval_loss": 0.761424720287323, "eval_runtime": 0.4374, "eval_samples_per_second": 203.488, "eval_steps_per_second": 4.573, "step": 22 }, { "epoch": 8.0, "eval_accuracy": 0.7752808988764045, "eval_loss": 0.634343683719635, "eval_runtime": 0.4474, "eval_samples_per_second": 198.926, "eval_steps_per_second": 4.47, "step": 26 }, { "epoch": 8.923076923076923, "eval_accuracy": 0.797752808988764, "eval_loss": 0.5810024738311768, "eval_runtime": 0.4478, "eval_samples_per_second": 198.744, "eval_steps_per_second": 4.466, "step": 29 }, { "epoch": 9.23076923076923, "grad_norm": 40.938568115234375, "learning_rate": 3.7037037037037037e-05, "loss": 0.7715, "step": 30 }, { "epoch": 9.846153846153847, "eval_accuracy": 0.8426966292134831, "eval_loss": 0.5550972819328308, "eval_runtime": 0.4609, "eval_samples_per_second": 193.105, "eval_steps_per_second": 4.339, "step": 32 }, { "epoch": 10.76923076923077, "eval_accuracy": 0.8539325842696629, "eval_loss": 0.5209007859230042, "eval_runtime": 0.4554, "eval_samples_per_second": 195.413, "eval_steps_per_second": 4.391, "step": 35 }, { "epoch": 12.0, "eval_accuracy": 0.8202247191011236, "eval_loss": 0.5689600110054016, "eval_runtime": 0.4644, "eval_samples_per_second": 191.632, "eval_steps_per_second": 4.306, "step": 39 }, { "epoch": 12.307692307692308, "grad_norm": 41.01618957519531, "learning_rate": 3.08641975308642e-05, "loss": 0.5645, "step": 40 }, { "epoch": 12.923076923076923, "eval_accuracy": 0.8876404494382022, "eval_loss": 0.4430869519710541, "eval_runtime": 0.5046, "eval_samples_per_second": 176.377, "eval_steps_per_second": 3.964, "step": 42 }, { "epoch": 13.846153846153847, "eval_accuracy": 0.8202247191011236, "eval_loss": 0.4921651780605316, "eval_runtime": 0.4657, "eval_samples_per_second": 191.097, "eval_steps_per_second": 4.294, "step": 45 }, { "epoch": 14.76923076923077, "eval_accuracy": 0.8314606741573034, "eval_loss": 0.4914401173591614, "eval_runtime": 0.5376, "eval_samples_per_second": 165.548, "eval_steps_per_second": 3.72, "step": 48 }, { "epoch": 15.384615384615385, "grad_norm": 45.103057861328125, "learning_rate": 2.4691358024691357e-05, "loss": 0.4999, "step": 50 }, { "epoch": 16.0, "eval_accuracy": 0.898876404494382, "eval_loss": 0.3767582178115845, "eval_runtime": 0.4592, "eval_samples_per_second": 193.816, "eval_steps_per_second": 4.355, "step": 52 }, { "epoch": 16.923076923076923, "eval_accuracy": 0.8539325842696629, "eval_loss": 0.42921027541160583, "eval_runtime": 0.4521, "eval_samples_per_second": 196.876, "eval_steps_per_second": 4.424, "step": 55 }, { "epoch": 17.846153846153847, "eval_accuracy": 0.8651685393258427, "eval_loss": 0.3845795691013336, "eval_runtime": 0.4679, "eval_samples_per_second": 190.193, "eval_steps_per_second": 4.274, "step": 58 }, { "epoch": 18.46153846153846, "grad_norm": 36.70952224731445, "learning_rate": 1.8518518518518518e-05, "loss": 0.4555, "step": 60 }, { "epoch": 18.76923076923077, "eval_accuracy": 0.8876404494382022, "eval_loss": 0.3498174548149109, "eval_runtime": 0.488, "eval_samples_per_second": 182.38, "eval_steps_per_second": 4.098, "step": 61 }, { "epoch": 20.0, "eval_accuracy": 0.8651685393258427, "eval_loss": 0.3523467183113098, "eval_runtime": 0.4442, "eval_samples_per_second": 200.357, "eval_steps_per_second": 4.502, "step": 65 }, { "epoch": 20.923076923076923, "eval_accuracy": 0.8876404494382022, "eval_loss": 0.3540913164615631, "eval_runtime": 0.4584, "eval_samples_per_second": 194.157, "eval_steps_per_second": 4.363, "step": 68 }, { "epoch": 21.53846153846154, "grad_norm": 48.17884063720703, "learning_rate": 1.2345679012345678e-05, "loss": 0.3941, "step": 70 }, { "epoch": 21.846153846153847, "eval_accuracy": 0.898876404494382, "eval_loss": 0.32396575808525085, "eval_runtime": 0.4491, "eval_samples_per_second": 198.193, "eval_steps_per_second": 4.454, "step": 71 }, { "epoch": 22.76923076923077, "eval_accuracy": 0.898876404494382, "eval_loss": 0.31694793701171875, "eval_runtime": 0.4762, "eval_samples_per_second": 186.89, "eval_steps_per_second": 4.2, "step": 74 }, { "epoch": 24.0, "eval_accuracy": 0.8764044943820225, "eval_loss": 0.3317168951034546, "eval_runtime": 0.4808, "eval_samples_per_second": 185.122, "eval_steps_per_second": 4.16, "step": 78 }, { "epoch": 24.615384615384617, "grad_norm": 53.90333938598633, "learning_rate": 6.172839506172839e-06, "loss": 0.361, "step": 80 }, { "epoch": 24.923076923076923, "eval_accuracy": 0.8876404494382022, "eval_loss": 0.32509633898735046, "eval_runtime": 0.4625, "eval_samples_per_second": 192.419, "eval_steps_per_second": 4.324, "step": 81 }, { "epoch": 25.846153846153847, "eval_accuracy": 0.8764044943820225, "eval_loss": 0.3197612166404724, "eval_runtime": 0.4547, "eval_samples_per_second": 195.751, "eval_steps_per_second": 4.399, "step": 84 }, { "epoch": 26.76923076923077, "eval_accuracy": 0.8764044943820225, "eval_loss": 0.3116617500782013, "eval_runtime": 0.4618, "eval_samples_per_second": 192.737, "eval_steps_per_second": 4.331, "step": 87 }, { "epoch": 27.692307692307693, "grad_norm": 35.301124572753906, "learning_rate": 0.0, "loss": 0.3485, "step": 90 }, { "epoch": 27.692307692307693, "eval_accuracy": 0.8764044943820225, "eval_loss": 0.3100605309009552, "eval_runtime": 0.4918, "eval_samples_per_second": 180.964, "eval_steps_per_second": 4.067, "step": 90 }, { "epoch": 27.692307692307693, "step": 90, "total_flos": 5.512834219233116e+17, "train_loss": 0.7025861872567071, "train_runtime": 285.963, "train_samples_per_second": 83.927, "train_steps_per_second": 0.315 } ], "logging_steps": 10, "max_steps": 90, "num_input_tokens_seen": 0, "num_train_epochs": 30, "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": 5.512834219233116e+17, "train_batch_size": 64, "trial_name": null, "trial_params": null }