Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Time #15

Open
10TREE-YEAR opened this issue Dec 4, 2020 · 1 comment
Open

Time #15

10TREE-YEAR opened this issue Dec 4, 2020 · 1 comment
Assignees
Labels
question Further information is requested

Comments

@10TREE-YEAR
Copy link

Hello, I would like to ask you, what kind of running program do you use? I ran on the server and trained 5 times for 3 days. The speed is too slow, and you trained 1000 times, so can you tell me what you used to run the program method.

@muellerdo
Copy link
Member

Hello @10TREE-YEAR,

can you specify what you mean with running program?

The script, which we used to generate the results of this paper, is located under scripts/run_miscnn.py
As a GPU, we utilized a NVIDIA QUADRO RTX 6000.

Training time was less than 2 days per fold.
If you didn't change anything in the hyperparameters of this script, then it's probably your hardware.
I guess, either GPU or CPU bottleneck. Required RAM is less than 20GB and required GPU VRAM less than 17GB.

Cheers,
Dominik

@muellerdo muellerdo self-assigned this Dec 7, 2020
@muellerdo muellerdo added the question Further information is requested label Dec 7, 2020
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
question Further information is requested
Projects
None yet
Development

No branches or pull requests

2 participants