| { | |
| "best_metric": 0.9862068965517241, | |
| "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-sealv1/checkpoint-94", | |
| "epoch": 28.571428571428573, | |
| "eval_steps": 500, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.95, | |
| "eval_accuracy": 0.496551724137931, | |
| "eval_loss": 1.2919867038726807, | |
| "eval_runtime": 3.1223, | |
| "eval_samples_per_second": 46.44, | |
| "eval_steps_per_second": 1.601, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 29.579647064208984, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 1.1379, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "eval_accuracy": 0.496551724137931, | |
| "eval_loss": 1.017675518989563, | |
| "eval_runtime": 3.0482, | |
| "eval_samples_per_second": 47.569, | |
| "eval_steps_per_second": 1.64, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "eval_accuracy": 0.8758620689655172, | |
| "eval_loss": 0.7626163959503174, | |
| "eval_runtime": 3.0592, | |
| "eval_samples_per_second": 47.398, | |
| "eval_steps_per_second": 1.634, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 3.81, | |
| "grad_norm": 11.914603233337402, | |
| "learning_rate": 4.814814814814815e-05, | |
| "loss": 0.6784, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9310344827586207, | |
| "eval_loss": 0.5387857556343079, | |
| "eval_runtime": 3.0489, | |
| "eval_samples_per_second": 47.558, | |
| "eval_steps_per_second": 1.64, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 4.95, | |
| "eval_accuracy": 0.9103448275862069, | |
| "eval_loss": 0.4190843105316162, | |
| "eval_runtime": 3.0202, | |
| "eval_samples_per_second": 48.01, | |
| "eval_steps_per_second": 1.656, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 5.71, | |
| "grad_norm": 7.094911575317383, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 0.3269, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.9, | |
| "eval_accuracy": 0.8896551724137931, | |
| "eval_loss": 0.39897656440734863, | |
| "eval_runtime": 3.0245, | |
| "eval_samples_per_second": 47.942, | |
| "eval_steps_per_second": 1.653, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 6.86, | |
| "eval_accuracy": 0.9517241379310345, | |
| "eval_loss": 0.20898482203483582, | |
| "eval_runtime": 3.0505, | |
| "eval_samples_per_second": 47.534, | |
| "eval_steps_per_second": 1.639, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 7.62, | |
| "grad_norm": 6.531408786773682, | |
| "learning_rate": 4.074074074074074e-05, | |
| "loss": 0.2068, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9586206896551724, | |
| "eval_loss": 0.18190069496631622, | |
| "eval_runtime": 3.053, | |
| "eval_samples_per_second": 47.494, | |
| "eval_steps_per_second": 1.638, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 8.95, | |
| "eval_accuracy": 0.9655172413793104, | |
| "eval_loss": 0.11921437084674835, | |
| "eval_runtime": 3.0392, | |
| "eval_samples_per_second": 47.71, | |
| "eval_steps_per_second": 1.645, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 9.52, | |
| "grad_norm": 5.500025272369385, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 0.1104, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 9.9, | |
| "eval_accuracy": 0.9724137931034482, | |
| "eval_loss": 0.06815902888774872, | |
| "eval_runtime": 3.0143, | |
| "eval_samples_per_second": 48.105, | |
| "eval_steps_per_second": 1.659, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 10.86, | |
| "eval_accuracy": 0.9724137931034482, | |
| "eval_loss": 0.08536545187234879, | |
| "eval_runtime": 3.0425, | |
| "eval_samples_per_second": 47.658, | |
| "eval_steps_per_second": 1.643, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 11.43, | |
| "grad_norm": 2.2279155254364014, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.0571, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.9655172413793104, | |
| "eval_loss": 0.08157803863286972, | |
| "eval_runtime": 3.0775, | |
| "eval_samples_per_second": 47.116, | |
| "eval_steps_per_second": 1.625, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 12.95, | |
| "eval_accuracy": 0.9793103448275862, | |
| "eval_loss": 0.053488027304410934, | |
| "eval_runtime": 3.0426, | |
| "eval_samples_per_second": 47.656, | |
| "eval_steps_per_second": 1.643, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 13.33, | |
| "grad_norm": 3.4508020877838135, | |
| "learning_rate": 2.962962962962963e-05, | |
| "loss": 0.0382, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 13.9, | |
| "eval_accuracy": 0.9793103448275862, | |
| "eval_loss": 0.04910058155655861, | |
| "eval_runtime": 3.0539, | |
| "eval_samples_per_second": 47.48, | |
| "eval_steps_per_second": 1.637, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 14.86, | |
| "eval_accuracy": 0.9793103448275862, | |
| "eval_loss": 0.05336721986532211, | |
| "eval_runtime": 3.0215, | |
| "eval_samples_per_second": 47.989, | |
| "eval_steps_per_second": 1.655, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 15.24, | |
| "grad_norm": 0.396097868680954, | |
| "learning_rate": 2.5925925925925925e-05, | |
| "loss": 0.0158, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.9793103448275862, | |
| "eval_loss": 0.03685032203793526, | |
| "eval_runtime": 3.0204, | |
| "eval_samples_per_second": 48.006, | |
| "eval_steps_per_second": 1.655, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 16.95, | |
| "eval_accuracy": 0.9724137931034482, | |
| "eval_loss": 0.11105308681726456, | |
| "eval_runtime": 3.0328, | |
| "eval_samples_per_second": 47.81, | |
| "eval_steps_per_second": 1.649, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 17.14, | |
| "grad_norm": 1.3435662984848022, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.0082, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 17.9, | |
| "eval_accuracy": 0.9862068965517241, | |
| "eval_loss": 0.0514516718685627, | |
| "eval_runtime": 3.0168, | |
| "eval_samples_per_second": 48.065, | |
| "eval_steps_per_second": 1.657, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 18.86, | |
| "eval_accuracy": 0.9793103448275862, | |
| "eval_loss": 0.07132605463266373, | |
| "eval_runtime": 3.0424, | |
| "eval_samples_per_second": 47.66, | |
| "eval_steps_per_second": 1.643, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 19.05, | |
| "grad_norm": 1.5692905187606812, | |
| "learning_rate": 1.8518518518518518e-05, | |
| "loss": 0.0105, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.9793103448275862, | |
| "eval_loss": 0.059764981269836426, | |
| "eval_runtime": 3.0213, | |
| "eval_samples_per_second": 47.993, | |
| "eval_steps_per_second": 1.655, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 20.95, | |
| "grad_norm": 0.6745745539665222, | |
| "learning_rate": 1.4814814814814815e-05, | |
| "loss": 0.009, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 20.95, | |
| "eval_accuracy": 0.9724137931034482, | |
| "eval_loss": 0.07587717473506927, | |
| "eval_runtime": 3.0482, | |
| "eval_samples_per_second": 47.569, | |
| "eval_steps_per_second": 1.64, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 21.9, | |
| "eval_accuracy": 0.9793103448275862, | |
| "eval_loss": 0.07691250741481781, | |
| "eval_runtime": 3.0353, | |
| "eval_samples_per_second": 47.771, | |
| "eval_steps_per_second": 1.647, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 22.86, | |
| "grad_norm": 5.058462142944336, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 0.0134, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 22.86, | |
| "eval_accuracy": 0.9793103448275862, | |
| "eval_loss": 0.070177361369133, | |
| "eval_runtime": 3.0247, | |
| "eval_samples_per_second": 47.938, | |
| "eval_steps_per_second": 1.653, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_accuracy": 0.9793103448275862, | |
| "eval_loss": 0.060489553958177567, | |
| "eval_runtime": 3.0525, | |
| "eval_samples_per_second": 47.501, | |
| "eval_steps_per_second": 1.638, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 24.76, | |
| "grad_norm": 1.4174549579620361, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 0.0042, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 24.95, | |
| "eval_accuracy": 0.9793103448275862, | |
| "eval_loss": 0.06210276484489441, | |
| "eval_runtime": 3.0244, | |
| "eval_samples_per_second": 47.944, | |
| "eval_steps_per_second": 1.653, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 25.9, | |
| "eval_accuracy": 0.9793103448275862, | |
| "eval_loss": 0.06535915285348892, | |
| "eval_runtime": 3.0417, | |
| "eval_samples_per_second": 47.671, | |
| "eval_steps_per_second": 1.644, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 26.67, | |
| "grad_norm": 0.7513312101364136, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 0.0027, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 26.86, | |
| "eval_accuracy": 0.9724137931034482, | |
| "eval_loss": 0.06658191233873367, | |
| "eval_runtime": 3.0308, | |
| "eval_samples_per_second": 47.843, | |
| "eval_steps_per_second": 1.65, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_accuracy": 0.9793103448275862, | |
| "eval_loss": 0.06654076278209686, | |
| "eval_runtime": 3.0747, | |
| "eval_samples_per_second": 47.159, | |
| "eval_steps_per_second": 1.626, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 28.57, | |
| "grad_norm": 2.4569530487060547, | |
| "learning_rate": 0.0, | |
| "loss": 0.0065, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 28.57, | |
| "eval_accuracy": 0.9793103448275862, | |
| "eval_loss": 0.06503264605998993, | |
| "eval_runtime": 3.0856, | |
| "eval_samples_per_second": 46.992, | |
| "eval_steps_per_second": 1.62, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 28.57, | |
| "step": 150, | |
| "total_flos": 1.4797115229312e+18, | |
| "train_loss": 0.175064886448284, | |
| "train_runtime": 1076.0914, | |
| "train_samples_per_second": 18.205, | |
| "train_steps_per_second": 0.139 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 150, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 500, | |
| "total_flos": 1.4797115229312e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |