{ "best_metric": 0.8847062587738037, "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned/checkpoint-320", "epoch": 9.846153846153847, "eval_steps": 500, "global_step": 320, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.3076923076923077, "grad_norm": 4.222843647003174, "learning_rate": 1.5625e-05, "loss": 1.2776, "step": 10 }, { "epoch": 0.6153846153846154, "grad_norm": 4.994299411773682, "learning_rate": 3.125e-05, "loss": 1.1793, "step": 20 }, { "epoch": 0.9230769230769231, "grad_norm": 5.790884971618652, "learning_rate": 4.6875e-05, "loss": 1.089, "step": 30 }, { "epoch": 0.9846153846153847, "eval_accuracy": 0.591915303176131, "eval_f1": 0.5663015617463389, "eval_loss": 1.043319582939148, "eval_precision": 0.570578116425331, "eval_recall": 0.591915303176131, "eval_runtime": 972.9147, "eval_samples_per_second": 1.068, "eval_steps_per_second": 0.034, "step": 32 }, { "epoch": 1.2307692307692308, "grad_norm": 5.5890069007873535, "learning_rate": 4.8611111111111115e-05, "loss": 1.0389, "step": 40 }, { "epoch": 1.5384615384615383, "grad_norm": 7.14047908782959, "learning_rate": 4.6875e-05, "loss": 1.0614, "step": 50 }, { "epoch": 1.8461538461538463, "grad_norm": 5.124348163604736, "learning_rate": 4.5138888888888894e-05, "loss": 1.0165, "step": 60 }, { "epoch": 2.0, "eval_accuracy": 0.5928777670837344, "eval_f1": 0.5521015886078269, "eval_loss": 1.0114457607269287, "eval_precision": 0.600779215992248, "eval_recall": 0.5928777670837344, "eval_runtime": 23.8377, "eval_samples_per_second": 43.586, "eval_steps_per_second": 1.384, "step": 65 }, { "epoch": 2.1538461538461537, "grad_norm": 3.7715022563934326, "learning_rate": 4.340277777777778e-05, "loss": 0.9634, "step": 70 }, { "epoch": 2.4615384615384617, "grad_norm": 6.994911193847656, "learning_rate": 4.166666666666667e-05, "loss": 0.9796, "step": 80 }, { "epoch": 2.769230769230769, "grad_norm": 5.106964111328125, "learning_rate": 3.993055555555556e-05, "loss": 0.935, "step": 90 }, { "epoch": 2.9846153846153847, "eval_accuracy": 0.6371511068334937, "eval_f1": 0.6069098097960562, "eval_loss": 0.9436932802200317, "eval_precision": 0.6626840278536283, "eval_recall": 0.6371511068334937, "eval_runtime": 24.0859, "eval_samples_per_second": 43.137, "eval_steps_per_second": 1.37, "step": 97 }, { "epoch": 3.076923076923077, "grad_norm": 5.561572551727295, "learning_rate": 3.8194444444444444e-05, "loss": 0.908, "step": 100 }, { "epoch": 3.3846153846153846, "grad_norm": 5.457753658294678, "learning_rate": 3.6458333333333336e-05, "loss": 0.8504, "step": 110 }, { "epoch": 3.6923076923076925, "grad_norm": 6.219572067260742, "learning_rate": 3.472222222222222e-05, "loss": 0.8522, "step": 120 }, { "epoch": 4.0, "grad_norm": 5.555025577545166, "learning_rate": 3.2986111111111115e-05, "loss": 0.9051, "step": 130 }, { "epoch": 4.0, "eval_accuracy": 0.6400384985563041, "eval_f1": 0.6327783476656897, "eval_loss": 0.9238916635513306, "eval_precision": 0.6380923158137368, "eval_recall": 0.6400384985563041, "eval_runtime": 23.787, "eval_samples_per_second": 43.679, "eval_steps_per_second": 1.387, "step": 130 }, { "epoch": 4.3076923076923075, "grad_norm": 4.2132954597473145, "learning_rate": 3.125e-05, "loss": 0.8394, "step": 140 }, { "epoch": 4.615384615384615, "grad_norm": 7.527224540710449, "learning_rate": 2.951388888888889e-05, "loss": 0.8278, "step": 150 }, { "epoch": 4.923076923076923, "grad_norm": 8.78996467590332, "learning_rate": 2.777777777777778e-05, "loss": 0.856, "step": 160 }, { "epoch": 4.984615384615385, "eval_accuracy": 0.6381135707410972, "eval_f1": 0.631926098346969, "eval_loss": 0.9268857836723328, "eval_precision": 0.6475643517149878, "eval_recall": 0.6381135707410972, "eval_runtime": 23.9493, "eval_samples_per_second": 43.383, "eval_steps_per_second": 1.378, "step": 162 }, { "epoch": 5.230769230769231, "grad_norm": 4.454049587249756, "learning_rate": 2.604166666666667e-05, "loss": 0.7715, "step": 170 }, { "epoch": 5.538461538461538, "grad_norm": 5.65101957321167, "learning_rate": 2.4305555555555558e-05, "loss": 0.8265, "step": 180 }, { "epoch": 5.846153846153846, "grad_norm": 6.742358207702637, "learning_rate": 2.2569444444444447e-05, "loss": 0.8317, "step": 190 }, { "epoch": 6.0, "eval_accuracy": 0.6487006737247353, "eval_f1": 0.6366752317854656, "eval_loss": 0.911486804485321, "eval_precision": 0.653599096662139, "eval_recall": 0.6487006737247353, "eval_runtime": 23.7936, "eval_samples_per_second": 43.667, "eval_steps_per_second": 1.387, "step": 195 }, { "epoch": 6.153846153846154, "grad_norm": 6.174919128417969, "learning_rate": 2.0833333333333336e-05, "loss": 0.7994, "step": 200 }, { "epoch": 6.461538461538462, "grad_norm": 6.239418983459473, "learning_rate": 1.9097222222222222e-05, "loss": 0.7792, "step": 210 }, { "epoch": 6.769230769230769, "grad_norm": 7.731762886047363, "learning_rate": 1.736111111111111e-05, "loss": 0.7914, "step": 220 }, { "epoch": 6.984615384615385, "eval_accuracy": 0.6660250240615977, "eval_f1": 0.6558142271285787, "eval_loss": 0.8913388848304749, "eval_precision": 0.6622428442076029, "eval_recall": 0.6660250240615977, "eval_runtime": 24.0291, "eval_samples_per_second": 43.239, "eval_steps_per_second": 1.373, "step": 227 }, { "epoch": 7.076923076923077, "grad_norm": 5.157851219177246, "learning_rate": 1.5625e-05, "loss": 0.744, "step": 230 }, { "epoch": 7.384615384615385, "grad_norm": 8.33967113494873, "learning_rate": 1.388888888888889e-05, "loss": 0.723, "step": 240 }, { "epoch": 7.6923076923076925, "grad_norm": 6.409264087677002, "learning_rate": 1.2152777777777779e-05, "loss": 0.7512, "step": 250 }, { "epoch": 8.0, "grad_norm": 6.12069034576416, "learning_rate": 1.0416666666666668e-05, "loss": 0.763, "step": 260 }, { "epoch": 8.0, "eval_accuracy": 0.6631376323387873, "eval_f1": 0.6567791543544663, "eval_loss": 0.8967146873474121, "eval_precision": 0.6610152533977022, "eval_recall": 0.6631376323387873, "eval_runtime": 23.8361, "eval_samples_per_second": 43.589, "eval_steps_per_second": 1.384, "step": 260 }, { "epoch": 8.307692307692308, "grad_norm": 7.385721683502197, "learning_rate": 8.680555555555556e-06, "loss": 0.7283, "step": 270 }, { "epoch": 8.615384615384615, "grad_norm": 4.541664123535156, "learning_rate": 6.944444444444445e-06, "loss": 0.7345, "step": 280 }, { "epoch": 8.923076923076923, "grad_norm": 7.139510631561279, "learning_rate": 5.208333333333334e-06, "loss": 0.7079, "step": 290 }, { "epoch": 8.984615384615385, "eval_accuracy": 0.6612127045235804, "eval_f1": 0.6518917882622605, "eval_loss": 0.9004968404769897, "eval_precision": 0.6637745288500654, "eval_recall": 0.6612127045235804, "eval_runtime": 24.1208, "eval_samples_per_second": 43.075, "eval_steps_per_second": 1.368, "step": 292 }, { "epoch": 9.23076923076923, "grad_norm": 4.807325839996338, "learning_rate": 3.4722222222222224e-06, "loss": 0.7071, "step": 300 }, { "epoch": 9.538461538461538, "grad_norm": 5.517625331878662, "learning_rate": 1.7361111111111112e-06, "loss": 0.6935, "step": 310 }, { "epoch": 9.846153846153847, "grad_norm": 5.944396495819092, "learning_rate": 0.0, "loss": 0.6984, "step": 320 }, { "epoch": 9.846153846153847, "eval_accuracy": 0.6612127045235804, "eval_f1": 0.6503718951301083, "eval_loss": 0.8847062587738037, "eval_precision": 0.6589546644463634, "eval_recall": 0.6612127045235804, "eval_runtime": 24.299, "eval_samples_per_second": 42.759, "eval_steps_per_second": 1.358, "step": 320 }, { "epoch": 9.846153846153847, "step": 320, "total_flos": 1.0354878719936594e+18, "train_loss": 0.8634449824690819, "train_runtime": 4124.7516, "train_samples_per_second": 10.066, "train_steps_per_second": 0.078 } ], "logging_steps": 10, "max_steps": 320, "num_input_tokens_seen": 0, "num_train_epochs": 10, "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": 1.0354878719936594e+18, "train_batch_size": 32, "trial_name": null, "trial_params": null }