[bug-fix] Set number of threads based on allocated CPU count in Docker containers #4471
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Contributor
|
Nice catch. I've run into this before and didn't think about it for pytorch. |
chriselion
reviewed
Sep 11, 2020
| def _read_in_integer_file(filename: str) -> int: | ||
| try: | ||
| with open(filename) as f: | ||
| return int(f.readlines()[0]) |
Contributor
There was a problem hiding this comment.
There's some code on https://bugs.python.org/issue36054 - they use the equivalent of int( f.read().rstrip() )
Contributor
Author
There was a problem hiding this comment.
I like the read().rstrip() better that what I had, changed
vincentpierre
approved these changes
Sep 11, 2020
chriselion
approved these changes
Sep 11, 2020
chriselion
reviewed
Sep 11, 2020
| return max(min(num_cpus // 2, 4), 1) if num_cpus is not None else None | ||
|
|
||
|
|
||
| def _get_num_cpus() -> Optional[int]: |
Contributor
There was a problem hiding this comment.
nit: maybe _get_num_available_cpus()?
chriselion
approved these changes
Sep 11, 2020
chriselion
pushed a commit
that referenced
this pull request
Sep 12, 2020
…r containers (#4471) * Set num threads properly for Docker * Pylint-friendly logic * Use f.read().rstrip() * Change function names
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Proposed change(s)
PyTorch by default sets number of threads to 1/2 available cores. However, in Docker containers, the built-in Python methods return the number of cores of the host machine, not the container. This PR reads the allocated CPU count from
cgroupand uses that to set the num threads.Types of change(s)
Checklist
Other comments