{ "best_metric": 0.9117647058823529, "best_model_checkpoint": "deit-tiny-patch16-224-finetuned-papsmear/checkpoint-269", "epoch": 14.805194805194805, "eval_steps": 500, "global_step": 285, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.5194805194805194, "grad_norm": 11.024985313415527, "learning_rate": 1.7241379310344828e-05, "loss": 0.3521, "step": 10 }, { "epoch": 0.987012987012987, "eval_accuracy": 0.8235294117647058, "eval_loss": 0.4595417082309723, "eval_runtime": 28.5401, "eval_samples_per_second": 4.765, "eval_steps_per_second": 0.315, "step": 19 }, { "epoch": 1.0389610389610389, "grad_norm": 14.813636779785156, "learning_rate": 3.4482758620689657e-05, "loss": 0.3115, "step": 20 }, { "epoch": 1.5584415584415585, "grad_norm": 16.087106704711914, "learning_rate": 4.9804687500000004e-05, "loss": 0.335, "step": 30 }, { "epoch": 1.974025974025974, "eval_accuracy": 0.8602941176470589, "eval_loss": 0.4491410553455353, "eval_runtime": 27.9762, "eval_samples_per_second": 4.861, "eval_steps_per_second": 0.322, "step": 38 }, { "epoch": 2.0779220779220777, "grad_norm": 19.62188720703125, "learning_rate": 4.78515625e-05, "loss": 0.4021, "step": 40 }, { "epoch": 2.5974025974025974, "grad_norm": 12.998429298400879, "learning_rate": 4.58984375e-05, "loss": 0.3248, "step": 50 }, { "epoch": 2.961038961038961, "eval_accuracy": 0.875, "eval_loss": 0.4195605218410492, "eval_runtime": 28.3986, "eval_samples_per_second": 4.789, "eval_steps_per_second": 0.317, "step": 57 }, { "epoch": 3.116883116883117, "grad_norm": 8.330426216125488, "learning_rate": 4.3945312500000005e-05, "loss": 0.3613, "step": 60 }, { "epoch": 3.6363636363636362, "grad_norm": 10.42120361328125, "learning_rate": 4.1992187500000003e-05, "loss": 0.3271, "step": 70 }, { "epoch": 4.0, "eval_accuracy": 0.8014705882352942, "eval_loss": 0.5466737151145935, "eval_runtime": 28.1607, "eval_samples_per_second": 4.829, "eval_steps_per_second": 0.32, "step": 77 }, { "epoch": 4.1558441558441555, "grad_norm": 14.403099060058594, "learning_rate": 4.00390625e-05, "loss": 0.3451, "step": 80 }, { "epoch": 4.675324675324675, "grad_norm": 18.443952560424805, "learning_rate": 3.80859375e-05, "loss": 0.3286, "step": 90 }, { "epoch": 4.987012987012987, "eval_accuracy": 0.8161764705882353, "eval_loss": 0.4768396317958832, "eval_runtime": 28.7873, "eval_samples_per_second": 4.724, "eval_steps_per_second": 0.313, "step": 96 }, { "epoch": 5.194805194805195, "grad_norm": 8.146430969238281, "learning_rate": 3.6132812500000005e-05, "loss": 0.2577, "step": 100 }, { "epoch": 5.714285714285714, "grad_norm": 7.037234306335449, "learning_rate": 3.41796875e-05, "loss": 0.2854, "step": 110 }, { "epoch": 5.974025974025974, "eval_accuracy": 0.8676470588235294, "eval_loss": 0.41474443674087524, "eval_runtime": 28.6719, "eval_samples_per_second": 4.743, "eval_steps_per_second": 0.314, "step": 115 }, { "epoch": 6.233766233766234, "grad_norm": 10.498587608337402, "learning_rate": 3.22265625e-05, "loss": 0.2295, "step": 120 }, { "epoch": 6.753246753246753, "grad_norm": 13.865105628967285, "learning_rate": 3.02734375e-05, "loss": 0.2291, "step": 130 }, { "epoch": 6.961038961038961, "eval_accuracy": 0.8676470588235294, "eval_loss": 0.4321264624595642, "eval_runtime": 28.1356, "eval_samples_per_second": 4.834, "eval_steps_per_second": 0.32, "step": 134 }, { "epoch": 7.2727272727272725, "grad_norm": 14.981769561767578, "learning_rate": 2.83203125e-05, "loss": 0.2356, "step": 140 }, { "epoch": 7.792207792207792, "grad_norm": 17.98302459716797, "learning_rate": 2.63671875e-05, "loss": 0.2619, "step": 150 }, { "epoch": 8.0, "eval_accuracy": 0.8235294117647058, "eval_loss": 0.5726210474967957, "eval_runtime": 28.5909, "eval_samples_per_second": 4.757, "eval_steps_per_second": 0.315, "step": 154 }, { "epoch": 8.311688311688311, "grad_norm": 17.48947525024414, "learning_rate": 2.44140625e-05, "loss": 0.2315, "step": 160 }, { "epoch": 8.831168831168831, "grad_norm": 18.059572219848633, "learning_rate": 2.24609375e-05, "loss": 0.2196, "step": 170 }, { "epoch": 8.987012987012987, "eval_accuracy": 0.8676470588235294, "eval_loss": 0.4344201982021332, "eval_runtime": 28.8842, "eval_samples_per_second": 4.708, "eval_steps_per_second": 0.312, "step": 173 }, { "epoch": 9.35064935064935, "grad_norm": 13.211594581604004, "learning_rate": 2.05078125e-05, "loss": 0.22, "step": 180 }, { "epoch": 9.87012987012987, "grad_norm": 14.067967414855957, "learning_rate": 1.85546875e-05, "loss": 0.2116, "step": 190 }, { "epoch": 9.974025974025974, "eval_accuracy": 0.875, "eval_loss": 0.38092589378356934, "eval_runtime": 28.2538, "eval_samples_per_second": 4.814, "eval_steps_per_second": 0.319, "step": 192 }, { "epoch": 10.38961038961039, "grad_norm": 8.458534240722656, "learning_rate": 1.66015625e-05, "loss": 0.1843, "step": 200 }, { "epoch": 10.909090909090908, "grad_norm": 11.62593936920166, "learning_rate": 1.4648437500000001e-05, "loss": 0.1913, "step": 210 }, { "epoch": 10.96103896103896, "eval_accuracy": 0.8602941176470589, "eval_loss": 0.37573114037513733, "eval_runtime": 28.2614, "eval_samples_per_second": 4.812, "eval_steps_per_second": 0.318, "step": 211 }, { "epoch": 11.428571428571429, "grad_norm": 15.429533958435059, "learning_rate": 1.2695312500000001e-05, "loss": 0.1654, "step": 220 }, { "epoch": 11.948051948051948, "grad_norm": 8.531671524047852, "learning_rate": 1.0742187500000001e-05, "loss": 0.1604, "step": 230 }, { "epoch": 12.0, "eval_accuracy": 0.8897058823529411, "eval_loss": 0.35508498549461365, "eval_runtime": 28.7628, "eval_samples_per_second": 4.728, "eval_steps_per_second": 0.313, "step": 231 }, { "epoch": 12.467532467532468, "grad_norm": 7.988156318664551, "learning_rate": 8.789062500000001e-06, "loss": 0.1558, "step": 240 }, { "epoch": 12.987012987012987, "grad_norm": 6.887962341308594, "learning_rate": 6.8359375e-06, "loss": 0.1307, "step": 250 }, { "epoch": 12.987012987012987, "eval_accuracy": 0.8970588235294118, "eval_loss": 0.3330402374267578, "eval_runtime": 29.009, "eval_samples_per_second": 4.688, "eval_steps_per_second": 0.31, "step": 250 }, { "epoch": 13.506493506493506, "grad_norm": 7.4578070640563965, "learning_rate": 4.8828125e-06, "loss": 0.1425, "step": 260 }, { "epoch": 13.974025974025974, "eval_accuracy": 0.9117647058823529, "eval_loss": 0.34208229184150696, "eval_runtime": 28.868, "eval_samples_per_second": 4.711, "eval_steps_per_second": 0.312, "step": 269 }, { "epoch": 14.025974025974026, "grad_norm": 13.509991645812988, "learning_rate": 2.9296875e-06, "loss": 0.1368, "step": 270 }, { "epoch": 14.545454545454545, "grad_norm": 9.896185874938965, "learning_rate": 9.765625e-07, "loss": 0.141, "step": 280 }, { "epoch": 14.805194805194805, "eval_accuracy": 0.9117647058823529, "eval_loss": 0.3409328758716583, "eval_runtime": 28.479, "eval_samples_per_second": 4.775, "eval_steps_per_second": 0.316, "step": 285 }, { "epoch": 14.805194805194805, "step": 285, "total_flos": 9.04681758989353e+16, "train_loss": 0.2430994153022766, "train_runtime": 4010.606, "train_samples_per_second": 4.578, "train_steps_per_second": 0.071 } ], "logging_steps": 10, "max_steps": 285, "num_input_tokens_seen": 0, "num_train_epochs": 15, "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": 9.04681758989353e+16, "train_batch_size": 16, "trial_name": null, "trial_params": null }