Investigate running Pytorch on GPU #211
Open
Conversation
This file contains 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
…nction to use within the model. The rest is probably fixing bugs
|
Hi Dennis, did you mean to close this issue before it was reviewed?
Thanks for posting your initial work.
--
*Lewis*
Dr. Lewis J. McGibbney Ph.D, B.Sc
*Skype*: lewis.john.mcgibbney
|
|
Actually @Luner if your work incorporated into this PR? |
|
@lewismc It looks like this PR is utilizing the code from issue 192. This was when we changed everything to pytorch but did not implement the column by column approach. The runtime for issue 192 was closer to 50 minutes. |
|
OK, we can rebase this once we've worked on #198 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
I think I managed to fix the problem I was running into. The approach I had before didn't seem to take into account how Pytorch handled wrapping the Model class and distributing the necessary information to each forward function to perform the operations. I believe the trick to getting Pytorch to work was creating a custom function class inheriting from torch.autograd.
What I have, if I can look into it some more and fix some bugs, should be able to run with any number of GPUs while evenly distributing the image data, scored and classified arrays across them.