search

lukasruff / Deep-SVDD-PyTorch

A PyTorch implementation of the Deep SVDD anomaly detection method
MIT License
698 stars 197 forks source link

Why the "/log/" folder is empty? #5

Closed ZeyuanYu closed 5 years ago

ZeyuanYu commented 5 years ago

It requires "log.txt", "result.json", "model.tar" such files and so on. But how could I get these files to fill the "/log/" folder? There is no related content in "Readme.md". So now I'm really confused. Thanks.

lukasruff commented 5 years ago

Those files are the output from running an experiment.

You can specify the export path by the xp_path argument (3rd argument) of the main.py script.

Have you tried running the MNIST example?

python main.py mnist mnist_LeNet ../log/mnist_test ../data --objective one-class --lr 0.0001 --n_epochs 150 --lr_milestone 50 --batch_size 200 --weight_decay 0.5e-6 --pretrain True --ae_lr 0.0001 --ae_n_epochs 150 --ae_lr_milestone 50 --ae_batch_size 200 --ae_weight_decay 0.5e-3 --normal_class 3;

You can use the --load_config or the --load_model option to either load an experimental configuration or a model from the respective files (config.json or model.tar) afterwards. log.txt is a log-file of the console prints and result.json contains the results of an experiment.

peiseng commented 5 years ago

Create a blank folder named "mnist_test" inside the "log" folder. As when you configure the path to "../log/mnist_test", the folder named "mnist_test" does not exit leading to error.

lukasruff commented 5 years ago

I've updated the examples in README for clarity.