Skip to content
This repository has been archived by the owner on Jan 12, 2024. It is now read-only.

How can I entered mnist data to ./data folder ? #2

Closed
bemoregt opened this issue Jun 16, 2018 · 4 comments
Closed

How can I entered mnist data to ./data folder ? #2

bemoregt opened this issue Jun 16, 2018 · 4 comments

Comments

@bemoregt
Copy link

Hi, @samet-akcay

How can I entered mnist data to ./data folder ?

It needs ubyte type? or npz type?

from @bemoregt.

@samet-akcay
Copy link
Owner

Hi @bemoregt,

You don't need to do anything. When you just train the model on MNIST data, the script first checks whether you have MNIST data within the directory. If it is not there, it downloads it automatically and creates an anomaly dataset from it.

For instance, when you run sh experiments/run_mnist.sh, the dataset will be downloaded if it is not in the directory.

@bemoregt
Copy link
Author

Hi, @samet-akcay

Thanks quick reply.

But I met this error when run your code.

-------------- End ----------------
Traceback (most recent call last):
File "train.py", line 46, in
main()
File "train.py", line 35, in main
dataloader = load_data(opt)
File "/home/kerb/ganomaly/lib/data.py", line 73, in load_data
opt.anomaly_class = int(opt.anomaly_class)
ValueError: invalid literal for int() with base 10: '{0..9}'


What's wrong with me?

from @bemoregt.

@samet-akcay
Copy link
Owner

Can you please paste here the command you run in terminal?

@bemoregt
Copy link
Author

Hi, @samet-akcay

This matter is fixed when I using Pytorch 0.3.

But Another Problems happen.

That matter is on another issue.

from @bemoregt.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants