M1 deployment progress#37
Draft
alexlichtenstein wants to merge 1 commit intoHumanCompatibleAI:masterfrom
Draft
Conversation
micahcarroll
requested changes
Jun 2, 2022
Member
micahcarroll
left a comment
micahcarroll
left a comment
There was a problem hiding this comment.
Looks good so far! We should wait until we have done one of the 2 things you sugget in the PR description before merging this in!
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This version of the code allows to run the application on the m1 architecture, but the correction of the ray library should be done (it cannot load the agents right now). There are two ways to do this:
currently tf 2.6.0 and ray 1.1.0 installed with the support of m1 architecture. Therefore, a code update is required to get the ray classes working again. The main problem is the class for the Recurrent Network (it has a different name, location and probably functionality). I changed the name in ppo_rllib.py, but it crashed, so further debugging is needed.
another option is to rebuild the ray package 0.8.5 with m1 architecture support. This can be done with bazel, I added the installation of bazel to the dockerfile so only the code is missing to get the source and rebuild. Be aware if bazel crashes during the build due to memory issues, try using the -j 1(2,4,8) option.