Skip to content

Conversation

@shalgi
Copy link

@shalgi shalgi commented Jan 8, 2020

for servers with more than a single GPU it is possible to choose on which one to run, using a cuda_id and settings an environment variable to include it.

shalgi added 3 commits January 8, 2020 19:48
configure CUDA_VISIBLE_DEVICES environment variable to include only the wanted cuda_id
add the argument of cuda_id
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant