标签：Loading data py ImageNet LT params print open
1.stage_1 print
/home/zy/anaconda3/envs/py/bin/python /home/zy/pycharm/project/OpenLongTailRecognition-OLTR/main.py
Loading dataset from: /home/zy/pycharm/project/ImageNet2012
{'criterions': {'PerformanceLoss': {'def_file': './loss/SoftmaxLoss.py',
'loss_params': {},
'optim_params': None,
'weight': 1.0}},
'memory': {'centroids': False, 'init_centroids': False},
'networks': {'classifier': {'def_file': './models/DotProductClassifier.py',
'optim_params': {'lr': 0.1,
'momentum': 0.9,
'weight_decay': 0.0005},
'params': {'dataset': 'ImageNet_LT',
'in_dim': 512,
'num_classes': 1000,
'stage1_weights': False}},
'feat_model': {'def_file': './models/ResNet10Feature.py',
'fix': False,
'optim_params': {'lr': 0.1,
'momentum': 0.9,
'weight_decay': 0.0005},
'params': {'dataset': 'ImageNet_LT',
'dropout': None,
'stage1_weights': False,
'use_fc': False,
'use_modulatedatt': False}}},
'training_opt': {'batch_size': 128,
'dataset': 'ImageNet_LT',
'display_step': 10,
'feature_dim': 512,
'log_dir': './logs/ImageNet_LT/stage1',
'num_classes': 1000,
'num_epochs': 30,
'num_workers': 8,
'open_threshold': 0.1,
'sampler': None,
'scheduler_params': {'gamma': 0.1, 'step_size': 10}}}
Enter load_data
I want to know phase:%s
Loading data from ./data/ImageNet_LT/ImageNet_LT_train.txt
No sampler.
Shuffle is True.
Enter load_data
I want to know phase:%s
Loading data from ./data/ImageNet_LT/ImageNet_LT_val.txt
No sampler.
Shuffle is True.
Using 1 GPUs.
Loading Scratch ResNet 10 Feature Model.
No Pretrained Weights For Feature Model.
Loading Dot Product Classifier.
Random initialized classifier weights.
------Enter run_networks train----
Using steps for training.
Initializing model optimizer.
Loading Softmax Loss.
Phase: train
Epoch: [1/30] Step: 0 Minibatch_loss_performance: 6.958 Minibatch_accuracy_micro: 0.000
Epoch: [1/30] Step: 900 Minibatch_loss_performance: 5.773 Minibatch_accuracy_micro: 0.039
Phase: val
100%|██████████| 157/157 [00:27<00:00, 5.66it/s]
Phase: val
总体准确率(Evaluation_accuracy_micro_top1): 0.016
计算F度量值(Averaged F-measure): 0.004
多样本准确率(Many_shot_accuracy_top1): 0.042 中等样本准确率(Median_shot_accuracy_top1): 0.000 少样本准确率(Low_shot_accuracy_top1): 0.000
Epoch: [2/30] Step: 0 Minibatch_loss_performance: 5.838 Minibatch_accuracy_micro: 0.070
Epoch: [2/30] Step: 900 Minibatch_loss_performance: 5.226 Minibatch_accuracy_micro: 0.102
Phase: val
100%|██████████| 157/157 [00:27<00:00, 5.64it/s]
Phase: val
总体准确率(Evaluation_accuracy_micro_top1): 0.022
计算F度量值(Averaged F-measure): 0.008
多样本准确率(Many_shot_accuracy_top1): 0.053 中等样本准确率(Median_shot_accuracy_top1): 0.002 少样本准确率(Low_shot_accuracy_top1): 0.000
Epoch: [3/30] Step: 0 Minibatch_loss_performance: 5.210 Minibatch_accuracy_micro: 0.062
Epoch: [3/30] Step: 900 Minibatch_loss_performance: 4.887 Minibatch_accuracy_micro: 0.133
Phase: val
100%|██████████| 157/157 [00:26<00:00, 5.92it/s]
Phase: val
总体准确率(Evaluation_accuracy_micro_top1): 0.033
计算F度量值(Averaged F-measure): 0.013
多样本准确率(Many_shot_accuracy_top1): 0.084 中等样本准确率(Median_shot_accuracy_top1): 0.001 少样本准确率(Low_shot_accuracy_top1): 0.000
Epoch: [4/30] Step: 0 Minibatch_loss_performance: 4.857 Minibatch_accuracy_micro: 0.117
Phase: val
100%|██████████| 157/157 [00:24<00:00, 9.53it/s]
Phase: val
总体准确率(Evaluation_accuracy_micro_top1): 0.259
计算F度量值(Averaged F-measure): 0.211
多样本准确率(Many_shot_accuracy_top1): 0.493 中等样本准确率(Median_shot_accuracy_top1): 0.143 少样本准确率(Low_shot_accuracy_top1): 0.006
Epoch: [30/30] Step: 0 Minibatch_loss_performance: 2.552 Minibatch_accuracy_micro: 0.445
Epoch: [30/30] Step: 10 Minibatch_loss_performance: 2.570 Minibatch_accuracy_micro: 0.469
Epoch: [30/30] Step: 20 Minibatch_loss_performance: 2.714 Minibatch_accuracy_micro: 0.453
Epoch: [30/30] Step: 30 Minibatch_loss_performance: 2.458 Minibatch_accuracy_micro: 0.492
Epoch: [30/30] Step: 40 Minibatch_loss_performance: 2.490 Minibatch_accuracy_micro: 0.477
Epoch: [30/30] Step: 50 Minibatch_loss_performance: 2.590 Minibatch_accuracy_micro: 0.398
Epoch: [30/30] Step: 60 Minibatch_loss_performance: 2.514 Minibatch_accuracy_micro: 0.492
Epoch: [30/30] Step: 70 Minibatch_loss_performance: 2.822 Minibatch_accuracy_micro: 0.398
Epoch: [30/30] Step: 80 Minibatch_loss_performance: 2.642 Minibatch_accuracy_micro: 0.375
Epoch: [30/30] Step: 90 Minibatch_loss_performance: 2.389 Minibatch_accuracy_micro: 0.477
Epoch: [30/30] Step: 100 Minibatch_loss_performance: 2.305 Minibatch_accuracy_micro: 0.500
Epoch: [30/30] Step: 110 Minibatch_loss_performance: 2.620 Minibatch_accuracy_micro: 0.430
Epoch: [30/30] Step: 120 Minibatch_loss_performance: 2.502 Minibatch_accuracy_micro: 0.453
Epoch: [30/30] Step: 130 Minibatch_loss_performance: 2.354 Minibatch_accuracy_micro: 0.461
Epoch: [30/30] Step: 140 Minibatch_loss_performance: 2.532 Minibatch_accuracy_micro: 0.469
Epoch: [30/30] Step: 150 Minibatch_loss_performance: 2.524 Minibatch_accuracy_micro: 0.492
Epoch: [30/30] Step: 160 Minibatch_loss_performance: 2.945 Minibatch_accuracy_micro: 0.391
Epoch: [30/30] Step: 170 Minibatch_loss_performance: 2.498 Minibatch_accuracy_micro: 0.461
Epoch: [30/30] Step: 180 Minibatch_loss_performance: 2.608 Minibatch_accuracy_micro: 0.445
Epoch: [30/30] Step: 190 Minibatch_loss_performance: 2.591 Minibatch_accuracy_micro: 0.453
Epoch: [30/30] Step: 200 Minibatch_loss_performance: 2.370 Minibatch_accuracy_micro: 0.531
Epoch: [30/30] Step: 210 Minibatch_loss_performance: 2.481 Minibatch_accuracy_micro: 0.523
Epoch: [30/30] Step: 220 Minibatch_loss_performance: 2.447 Minibatch_accuracy_micro: 0.469
Epoch: [30/30] Step: 230 Minibatch_loss_performance: 2.541 Minibatch_accuracy_micro: 0.461
Epoch: [30/30] Step: 240 Minibatch_loss_performance: 2.581 Minibatch_accuracy_micro: 0.492
Epoch: [30/30] Step: 250 Minibatch_loss_performance: 2.785 Minibatch_accuracy_micro: 0.438
Epoch: [30/30] Step: 260 Minibatch_loss_performance: 2.376 Minibatch_accuracy_micro: 0.500
Epoch: [30/30] Step: 270 Minibatch_loss_performance: 2.592 Minibatch_accuracy_micro: 0.414
Epoch: [30/30] Step: 280 Minibatch_loss_performance: 2.532 Minibatch_accuracy_micro: 0.445
Epoch: [30/30] Step: 290 Minibatch_loss_performance: 2.585 Minibatch_accuracy_micro: 0.438
Epoch: [30/30] Step: 300 Minibatch_loss_performance: 3.031 Minibatch_accuracy_micro: 0.414
Epoch: [30/30] Step: 310 Minibatch_loss_performance: 2.546 Minibatch_accuracy_micro: 0.398
Epoch: [30/30] Step: 320 Minibatch_loss_performance: 2.952 Minibatch_accuracy_micro: 0.422
Epoch: [30/30] Step: 330 Minibatch_loss_performance: 2.688 Minibatch_accuracy_micro: 0.516
Epoch: [30/30] Step: 340 Minibatch_loss_performance: 2.745 Minibatch_accuracy_micro: 0.359
Epoch: [30/30] Step: 350 Minibatch_loss_performance: 2.765 Minibatch_accuracy_micro: 0.422
Epoch: [30/30] Step: 360 Minibatch_loss_performance: 2.433 Minibatch_accuracy_micro: 0.508
Epoch: [30/30] Step: 370 Minibatch_loss_performance: 2.718 Minibatch_accuracy_micro: 0.484
Epoch: [30/30] Step: 380 Minibatch_loss_performance: 2.307 Minibatch_accuracy_micro: 0.500
Epoch: [30/30] Step: 390 Minibatch_loss_performance: 2.836 Minibatch_accuracy_micro: 0.445
Epoch: [30/30] Step: 400 Minibatch_loss_performance: 2.403 Minibatch_accuracy_micro: 0.453
Epoch: [30/30] Step: 410 Minibatch_loss_performance: 2.764 Minibatch_accuracy_micro: 0.406
Epoch: [30/30] Step: 420 Minibatch_loss_performance: 2.469 Minibatch_accuracy_micro: 0.516
Epoch: [30/30] Step: 430 Minibatch_loss_performance: 2.680 Minibatch_accuracy_micro: 0.414
Epoch: [30/30] Step: 440 Minibatch_loss_performance: 2.146 Minibatch_accuracy_micro: 0.562
Epoch: [30/30] Step: 450 Minibatch_loss_performance: 2.557 Minibatch_accuracy_micro: 0.445
Epoch: [30/30] Step: 460 Minibatch_loss_performance: 2.604 Minibatch_accuracy_micro: 0.453
Epoch: [30/30] Step: 470 Minibatch_loss_performance: 2.752 Minibatch_accuracy_micro: 0.391
Epoch: [30/30] Step: 480 Minibatch_loss_performance: 2.318 Minibatch_accuracy_micro: 0.547
Epoch: [30/30] Step: 490 Minibatch_loss_performance: 2.462 Minibatch_accuracy_micro: 0.438
Epoch: [30/30] Step: 500 Minibatch_loss_performance: 2.541 Minibatch_accuracy_micro: 0.477
Epoch: [30/30] Step: 510 Minibatch_loss_performance: 2.681 Minibatch_accuracy_micro: 0.398
Epoch: [30/30] Step: 520 Minibatch_loss_performance: 2.608 Minibatch_accuracy_micro: 0.492
Epoch: [30/30] Step: 530 Minibatch_loss_performance: 2.566 Minibatch_accuracy_micro: 0.430
Epoch: [30/30] Step: 540 Minibatch_loss_performance: 2.698 Minibatch_accuracy_micro: 0.430
Epoch: [30/30] Step: 550 Minibatch_loss_performance: 2.329 Minibatch_accuracy_micro: 0.523
Epoch: [30/30] Step: 560 Minibatch_loss_performance: 2.531 Minibatch_accuracy_micro: 0.477
Epoch: [30/30] Step: 570 Minibatch_loss_performance: 2.595 Minibatch_accuracy_micro: 0.453
Epoch: [30/30] Step: 580 Minibatch_loss_performance: 2.625 Minibatch_accuracy_micro: 0.461
Epoch: [30/30] Step: 590 Minibatch_loss_performance: 2.839 Minibatch_accuracy_micro: 0.398
Epoch: [30/30] Step: 600 Minibatch_loss_performance: 3.175 Minibatch_accuracy_micro: 0.383
Epoch: [30/30] Step: 610 Minibatch_loss_performance: 2.679 Minibatch_accuracy_micro: 0.367
Epoch: [30/30] Step: 620 Minibatch_loss_performance: 2.710 Minibatch_accuracy_micro: 0.422
Epoch: [30/30] Step: 630 Minibatch_loss_performance: 2.895 Minibatch_accuracy_micro: 0.414
Epoch: [30/30] Step: 640 Minibatch_loss_performance: 2.462 Minibatch_accuracy_micro: 0.484
Epoch: [30/30] Step: 650 Minibatch_loss_performance: 2.285 Minibatch_accuracy_micro: 0.516
Epoch: [30/30] Step: 660 Minibatch_loss_performance: 2.644 Minibatch_accuracy_micro: 0.453
Epoch: [30/30] Step: 670 Minibatch_loss_performance: 2.943 Minibatch_accuracy_micro: 0.422
Epoch: [30/30] Step: 680 Minibatch_loss_performance: 2.542 Minibatch_accuracy_micro: 0.508
Epoch: [30/30] Step: 690 Minibatch_loss_performance: 2.716 Minibatch_accuracy_micro: 0.453
Epoch: [30/30] Step: 700 Minibatch_loss_performance: 2.673 Minibatch_accuracy_micro: 0.438
Epoch: [30/30] Step: 710 Minibatch_loss_performance: 2.660 Minibatch_accuracy_micro: 0.445
Epoch: [30/30] Step: 720 Minibatch_loss_performance: 3.312 Minibatch_accuracy_micro: 0.359
Epoch: [30/30] Step: 730 Minibatch_loss_performance: 2.430 Minibatch_accuracy_micro: 0.484
Epoch: [30/30] Step: 740 Minibatch_loss_performance: 2.376 Minibatch_accuracy_micro: 0.523
Epoch: [30/30] Step: 750 Minibatch_loss_performance: 2.464 Minibatch_accuracy_micro: 0.500
Epoch: [30/30] Step: 760 Minibatch_loss_performance: 2.506 Minibatch_accuracy_micro: 0.438
Epoch: [30/30] Step: 770 Minibatch_loss_performance: 2.920 Minibatch_accuracy_micro: 0.383
Epoch: [30/30] Step: 780 Minibatch_loss_performance: 2.449 Minibatch_accuracy_micro: 0.547
Epoch: [30/30] Step: 790 Minibatch_loss_performance: 3.060 Minibatch_accuracy_micro: 0.430
Epoch: [30/30] Step: 800 Minibatch_loss_performance: 2.656 Minibatch_accuracy_micro: 0.453
Epoch: [30/30] Step: 810 Minibatch_loss_performance: 2.757 Minibatch_accuracy_micro: 0.406
Epoch: [30/30] Step: 820 Minibatch_loss_performance: 2.482 Minibatch_accuracy_micro: 0.445
Epoch: [30/30] Step: 830 Minibatch_loss_performance: 2.679 Minibatch_accuracy_micro: 0.430
Epoch: [30/30] Step: 840 Minibatch_loss_performance: 2.907 Minibatch_accuracy_micro: 0.375
Epoch: [30/30] Step: 850 Minibatch_loss_performance: 2.605 Minibatch_accuracy_micro: 0.445
Epoch: [30/30] Step: 860 Minibatch_loss_performance: 2.474 Minibatch_accuracy_micro: 0.469
Epoch: [30/30] Step: 870 Minibatch_loss_performance: 2.651 Minibatch_accuracy_micro: 0.422
Epoch: [30/30] Step: 880 Minibatch_loss_performance: 2.678 Minibatch_accuracy_micro: 0.414
Epoch: [30/30] Step: 890 Minibatch_loss_performance: 2.647 Minibatch_accuracy_micro: 0.445
Epoch: [30/30] Step: 900 Minibatch_loss_performance: 2.643 Minibatch_accuracy_micro: 0.477
Phase: val
100%|██████████| 157/157 [00:24<00:00, 6.47it/s]
Phase: val
总体准确率(Evaluation_accuracy_micro_top1): 0.259
计算F度量值(Averaged F-measure): 0.210
多样本准确率(Many_shot_accuracy_top1): 0.491 中等样本准确率(Median_shot_accuracy_top1): 0.142 少样本准确率(Low_shot_accuracy_top1): 0.007
Training Complete.
Best validation accuracy is 0.259 at epoch 29
Done
ALL COMPLETED.
Process finished with exit code 0
/home/zy/anaconda3/envs/py/bin/python /home/zy/pycharm/project/OpenLongTailRecognition-OLTR/main.py
Loading dataset from: /home/zy/pycharm/project/ImageNet2012
{'criterions': {'FeatureLoss': {'def_file': './loss/DiscCentroidsLoss.py',
'loss_params': {'feat_dim': 512,
'num_classes': 1000},
'optim_params': {'lr': 0.01,
'momentum': 0.9,
'weight_decay': 0.0005},
'weight': 0.01},
'PerformanceLoss': {'def_file': './loss/SoftmaxLoss.py',
'loss_params': {},
'optim_params': None,
'weight': 1.0}},
'memory': {'centroids': True, 'init_centroids': True},
'networks': {'classifier': {'def_file': './models/MetaEmbeddingClassifier.py',
'optim_params': {'lr': 0.1,
'momentum': 0.9,
'weight_decay': 0.0005},
'params': {'dataset': 'ImageNet_LT',
'in_dim': 512,
'num_classes': 1000,
'stage1_weights': True}},
'feat_model': {'def_file': './models/ResNet10Feature.py',
'fix': False,
'optim_params': {'lr': 0.01,
'momentum': 0.9,
'weight_decay': 0.0005},
'params': {'dataset': 'ImageNet_LT',
'dropout': None,
'stage1_weights': True,
'use_fc': True,
'use_modulatedatt': True}}},
'training_opt': {'batch_size': 128,
'dataset': 'ImageNet_LT',
'display_step': 10,
'feature_dim': 512,
'log_dir': './logs/ImageNet_LT/meta_embedding',
'num_classes': 1000,
'num_epochs': 60,
'num_workers': 8,
'open_threshold': 0.1,
'sampler': {'def_file': './data/ClassAwareSampler.py',
'num_samples_cls': 4,
'type': 'ClassAwareSampler'},
'scheduler_params': {'gamma': 0.1, 'step_size': 20}}}
Loading data from ./data/ImageNet_LT/ImageNet_LT_train.txt
Use data transformation: Compose(
RandomResizedCrop(size=(224, 224), scale=(0.08, 1.0), ratio=(0.75, 1.3333), interpolation=bilinear)
RandomHorizontalFlip(p=0.5)
ColorJitter(brightness=[0.6, 1.4], contrast=[0.6, 1.4], saturation=[0.6, 1.4], hue=None)
ToTensor()
Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
)
Using sampler.
Sample 4 samples per-class.
Loading data from ./data/ImageNet_LT/ImageNet_LT_val.txt
Use data transformation: Compose(
Resize(size=256, interpolation=bilinear)
CenterCrop(size=(224, 224))
ToTensor()
Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
)
No sampler.
Shuffle is True.
Loading data from ./data/ImageNet_LT/ImageNet_LT_train.txt
Use data transformation: Compose(
Resize(size=256, interpolation=bilinear)
CenterCrop(size=(224, 224))
ToTensor()
Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
)
No sampler.
Shuffle is True.
Using 1 GPUs.
Loading Scratch ResNet 10 Feature Model.
Using fc.
Using self attention.
Loading ImageNet_LT Stage 1 ResNet 10 Weights.
Pretrained feature model weights path: ./logs/ImageNet_LT/stage1/final_model_checkpoint.pth
Loading Meta Embedding Classifier.
Loading ImageNet_LT Stage 1 Classifier Weights.
Pretrained classifier weights path: ./logs/ImageNet_LT/stage1/final_model_checkpoint.pth
------Enter run_networks train----
Using steps for training.
Initializing model optimizer.
Loading Softmax Loss.
Loading Discriminative Centroids Loss.
Initializing criterion optimizer.
Calculating centroids.
100%|██████████| 906/906 [02:21<00:00, 6.40it/s]
Phase: train
./loss/DiscCentroidsLoss.py:39: UserWarning: This overload of addmm_ is deprecated:
addmm_(Number beta, Number alpha, Tensor mat1, Tensor mat2)
Consider using one of the following signatures instead:
addmm_(Tensor mat1, Tensor mat2, *, Number beta, Number alpha) (Triggered internally at /pytorch/torch/csrc/utils/python_arg_parser.cpp:1005.)
distmat.addmm_(1, -2, feat.clone(), self.centroids.clone().t())
Epoch: [1/60] Step: 0 Minibatch_loss_feature: 0.109 Minibatch_loss_performance: 7.195 Minibatch_accuracy_micro: 0.000
Epoch: [1/60] Step: 900 Minibatch_loss_feature: 0.302 Minibatch_loss_performance: 5.015 Minibatch_accuracy_micro: 0.078
Phase: val
100%|██████████| 157/157 [00:25<00:00, 6.11it/s]
Phase: val
总体准确率(Evaluation_accuracy_micro_top1): 0.163
计算F度量值(Averaged F-measure): 0.142
多样本准确率(Many_shot_accuracy_top1): 0.189 中等样本准确率(Median_shot_accuracy_top1): 0.155 少样本准确率(Low_shot_accuracy_top1): 0.122
Epoch: [2/60] Step: 0 Minibatch_loss_feature: 0.291 Minibatch_loss_performance: 4.765 Minibatch_accuracy_micro: 0.141
Epoch: [2/60] Step: 900 Minibatch_loss_feature: 0.220 Minibatch_loss_performance: 4.135 Minibatch_accuracy_micro: 0.211
Phase: val
100%|██████████| 157/157 [00:25<00:00, 6.12it/s]
Phase: val
总体准确率(Evaluation_accuracy_micro_top1): 0.199
计算F度量值(Averaged F-measure): 0.179
多样本准确率(Many_shot_accuracy_top1): 0.235 中等样本准确率(Median_shot_accuracy_top1): 0.189 少样本准确率(Low_shot_accuracy_top1): 0.132
Epoch: [3/60] Step: 0 Minibatch_loss_feature: 0.213 Minibatch_loss_performance: 4.383 Minibatch_accuracy_micro: 0.172
Phase: val
100%|██████████| 157/157 [00:25<00:00, 6.28it/s]
Phase: val
总体准确率(Evaluation_accuracy_micro_top1): 0.399
计算F度量值(Averaged F-measure): 0.382
多样本准确率(Many_shot_accuracy_top1): 0.484 中等样本准确率(Median_shot_accuracy_top1): 0.391 少样本准确率(Low_shot_accuracy_top1): 0.197
Epoch: [60/60] Step: 0 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.730 Minibatch_accuracy_micro: 0.727
Epoch: [60/60] Step: 10 Minibatch_loss_feature: 0.010 Minibatch_loss_performance: 2.181 Minibatch_accuracy_micro: 0.547
Epoch: [60/60] Step: 20 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.547 Minibatch_accuracy_micro: 0.703
Epoch: [60/60] Step: 30 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.771 Minibatch_accuracy_micro: 0.648
Epoch: [60/60] Step: 40 Minibatch_loss_feature: 0.012 Minibatch_loss_performance: 1.961 Minibatch_accuracy_micro: 0.617
Epoch: [60/60] Step: 50 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.664 Minibatch_accuracy_micro: 0.656
Epoch: [60/60] Step: 60 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.954 Minibatch_accuracy_micro: 0.641
Epoch: [60/60] Step: 70 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.115 Minibatch_accuracy_micro: 0.602
Epoch: [60/60] Step: 80 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.132 Minibatch_accuracy_micro: 0.555
Epoch: [60/60] Step: 90 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.798 Minibatch_accuracy_micro: 0.656
Epoch: [60/60] Step: 100 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.915 Minibatch_accuracy_micro: 0.609
Epoch: [60/60] Step: 110 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.024 Minibatch_accuracy_micro: 0.609
Epoch: [60/60] Step: 120 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.863 Minibatch_accuracy_micro: 0.641
Epoch: [60/60] Step: 130 Minibatch_loss_feature: 0.010 Minibatch_loss_performance: 1.613 Minibatch_accuracy_micro: 0.703
Epoch: [60/60] Step: 140 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.846 Minibatch_accuracy_micro: 0.609
Epoch: [60/60] Step: 150 Minibatch_loss_feature: 0.010 Minibatch_loss_performance: 1.774 Minibatch_accuracy_micro: 0.617
Epoch: [60/60] Step: 160 Minibatch_loss_feature: 0.010 Minibatch_loss_performance: 1.796 Minibatch_accuracy_micro: 0.703
Epoch: [60/60] Step: 170 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.594 Minibatch_accuracy_micro: 0.750
Epoch: [60/60] Step: 180 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.685 Minibatch_accuracy_micro: 0.672
Epoch: [60/60] Step: 190 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.944 Minibatch_accuracy_micro: 0.648
Epoch: [60/60] Step: 200 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.977 Minibatch_accuracy_micro: 0.617
Epoch: [60/60] Step: 210 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.672 Minibatch_accuracy_micro: 0.672
Epoch: [60/60] Step: 220 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.535 Minibatch_accuracy_micro: 0.789
Epoch: [60/60] Step: 230 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.938 Minibatch_accuracy_micro: 0.633
Epoch: [60/60] Step: 240 Minibatch_loss_feature: 0.012 Minibatch_loss_performance: 1.961 Minibatch_accuracy_micro: 0.641
Epoch: [60/60] Step: 250 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.791 Minibatch_accuracy_micro: 0.688
Epoch: [60/60] Step: 260 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.663 Minibatch_accuracy_micro: 0.688
Epoch: [60/60] Step: 270 Minibatch_loss_feature: 0.012 Minibatch_loss_performance: 1.754 Minibatch_accuracy_micro: 0.656
Epoch: [60/60] Step: 280 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.112 Minibatch_accuracy_micro: 0.641
Epoch: [60/60] Step: 290 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.025 Minibatch_accuracy_micro: 0.594
Epoch: [60/60] Step: 300 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.946 Minibatch_accuracy_micro: 0.656
Epoch: [60/60] Step: 310 Minibatch_loss_feature: 0.010 Minibatch_loss_performance: 2.333 Minibatch_accuracy_micro: 0.547
Epoch: [60/60] Step: 320 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.724 Minibatch_accuracy_micro: 0.672
Epoch: [60/60] Step: 330 Minibatch_loss_feature: 0.010 Minibatch_loss_performance: 2.139 Minibatch_accuracy_micro: 0.602
Epoch: [60/60] Step: 340 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.079 Minibatch_accuracy_micro: 0.648
Epoch: [60/60] Step: 350 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.184 Minibatch_accuracy_micro: 0.617
Epoch: [60/60] Step: 360 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.165 Minibatch_accuracy_micro: 0.602
Epoch: [60/60] Step: 370 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.021 Minibatch_accuracy_micro: 0.578
Epoch: [60/60] Step: 380 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.073 Minibatch_accuracy_micro: 0.594
Epoch: [60/60] Step: 390 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.257 Minibatch_accuracy_micro: 0.531
Epoch: [60/60] Step: 400 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.067 Minibatch_accuracy_micro: 0.570
Epoch: [60/60] Step: 410 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.824 Minibatch_accuracy_micro: 0.648
Epoch: [60/60] Step: 420 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.715 Minibatch_accuracy_micro: 0.711
Epoch: [60/60] Step: 430 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.050 Minibatch_accuracy_micro: 0.578
Epoch: [60/60] Step: 440 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.672 Minibatch_accuracy_micro: 0.664
Epoch: [60/60] Step: 450 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.062 Minibatch_accuracy_micro: 0.625
Epoch: [60/60] Step: 460 Minibatch_loss_feature: 0.012 Minibatch_loss_performance: 1.850 Minibatch_accuracy_micro: 0.688
Epoch: [60/60] Step: 470 Minibatch_loss_feature: 0.012 Minibatch_loss_performance: 1.812 Minibatch_accuracy_micro: 0.625
Epoch: [60/60] Step: 480 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.685 Minibatch_accuracy_micro: 0.695
Epoch: [60/60] Step: 490 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.878 Minibatch_accuracy_micro: 0.641
Epoch: [60/60] Step: 500 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.120 Minibatch_accuracy_micro: 0.547
Epoch: [60/60] Step: 510 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.913 Minibatch_accuracy_micro: 0.625
Epoch: [60/60] Step: 520 Minibatch_loss_feature: 0.012 Minibatch_loss_performance: 1.823 Minibatch_accuracy_micro: 0.641
Epoch: [60/60] Step: 530 Minibatch_loss_feature: 0.012 Minibatch_loss_performance: 2.112 Minibatch_accuracy_micro: 0.602
Epoch: [60/60] Step: 540 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.955 Minibatch_accuracy_micro: 0.609
Epoch: [60/60] Step: 550 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.973 Minibatch_accuracy_micro: 0.609
Epoch: [60/60] Step: 560 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.888 Minibatch_accuracy_micro: 0.625
Epoch: [60/60] Step: 570 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.763 Minibatch_accuracy_micro: 0.656
Epoch: [60/60] Step: 580 Minibatch_loss_feature: 0.012 Minibatch_loss_performance: 1.704 Minibatch_accuracy_micro: 0.672
Epoch: [60/60] Step: 590 Minibatch_loss_feature: 0.012 Minibatch_loss_performance: 1.939 Minibatch_accuracy_micro: 0.570
Epoch: [60/60] Step: 600 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.996 Minibatch_accuracy_micro: 0.594
Epoch: [60/60] Step: 610 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.449 Minibatch_accuracy_micro: 0.531
Epoch: [60/60] Step: 620 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.634 Minibatch_accuracy_micro: 0.680
Epoch: [60/60] Step: 630 Minibatch_loss_feature: 0.010 Minibatch_loss_performance: 1.785 Minibatch_accuracy_micro: 0.664
Epoch: [60/60] Step: 640 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.854 Minibatch_accuracy_micro: 0.617
Epoch: [60/60] Step: 650 Minibatch_loss_feature: 0.012 Minibatch_loss_performance: 1.745 Minibatch_accuracy_micro: 0.672
Epoch: [60/60] Step: 660 Minibatch_loss_feature: 0.012 Minibatch_loss_performance: 1.661 Minibatch_accuracy_micro: 0.656
Epoch: [60/60] Step: 670 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.806 Minibatch_accuracy_micro: 0.633
Epoch: [60/60] Step: 680 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.190 Minibatch_accuracy_micro: 0.570
Epoch: [60/60] Step: 690 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.810 Minibatch_accuracy_micro: 0.648
Epoch: [60/60] Step: 700 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.198 Minibatch_accuracy_micro: 0.617
Epoch: [60/60] Step: 710 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.770 Minibatch_accuracy_micro: 0.703
Epoch: [60/60] Step: 720 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.869 Minibatch_accuracy_micro: 0.680
Epoch: [60/60] Step: 730 Minibatch_loss_feature: 0.012 Minibatch_loss_performance: 1.875 Minibatch_accuracy_micro: 0.594
Epoch: [60/60] Step: 740 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.889 Minibatch_accuracy_micro: 0.656
Epoch: [60/60] Step: 750 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.946 Minibatch_accuracy_micro: 0.672
Epoch: [60/60] Step: 760 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.834 Minibatch_accuracy_micro: 0.711
Epoch: [60/60] Step: 770 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.133 Minibatch_accuracy_micro: 0.586
Epoch: [60/60] Step: 780 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.007 Minibatch_accuracy_micro: 0.672
Epoch: [60/60] Step: 790 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.139 Minibatch_accuracy_micro: 0.609
Epoch: [60/60] Step: 800 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.847 Minibatch_accuracy_micro: 0.664
Epoch: [60/60] Step: 810 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.224 Minibatch_accuracy_micro: 0.602
Epoch: [60/60] Step: 820 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 2.099 Minibatch_accuracy_micro: 0.594
Epoch: [60/60] Step: 830 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.889 Minibatch_accuracy_micro: 0.672
Epoch: [60/60] Step: 840 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.989 Minibatch_accuracy_micro: 0.656
Epoch: [60/60] Step: 850 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.987 Minibatch_accuracy_micro: 0.625
Epoch: [60/60] Step: 860 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.919 Minibatch_accuracy_micro: 0.680
Epoch: [60/60] Step: 870 Minibatch_loss_feature: 0.012 Minibatch_loss_performance: 1.547 Minibatch_accuracy_micro: 0.734
Epoch: [60/60] Step: 880 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.712 Minibatch_accuracy_micro: 0.641
Epoch: [60/60] Step: 890 Minibatch_loss_feature: 0.011 Minibatch_loss_performance: 1.613 Minibatch_accuracy_micro: 0.727
Epoch: [60/60] Step: 900 Minibatch_loss_feature: 0.010 Minibatch_loss_performance: 2.221 Minibatch_accuracy_micro: 0.586
Phase: val
100%|██████████| 157/157 [00:25<00:00, 6.93it/s]
Phase: val
总体准确率(Evaluation_accuracy_micro_top1): 0.399
计算F度量值(Averaged F-measure): 0.381
多样本准确率(Many_shot_accuracy_top1): 0.485 中等样本准确率(Median_shot_accuracy_top1): 0.389 少样本准确率(Low_shot_accuracy_top1): 0.199
Training Complete.
Best validation accuracy is 0.400 at epoch 57
Done
ALL COMPLETED.
Process finished with exit code 0
标签：Loading,
data,
py,
ImageNet,
LT,
params,
print,
open
From： https://www.cnblogs.com/ZarkY/p/17891067.html