while_loop vmap init_val with mixed map/unmap dims #495
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
When a while_loop init_val is not mapped/batched in a vmap, we need to broadcast it out since the result of the body_fun will be mapped. Since the init_val can be a tuple in general, we need to handle mixed mapped/unmapped tuple elements, i.e. handle a mixture of NoneType and int bdims for init_val. This also let us delete some redundant code. fixes #489
Closed
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Outdated suggestions cannot be applied.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.
When a while_loop init_val is not mapped/batched in a vmap, we need to broadcast it out since the result of the body_fun will be mapped. Since the init_val can be a tuple in general, we need to handle mixed mapped/unmapped tuple elements, i.e. handle a mixture of NoneType and int bdims for init_val.
This also let us delete some redundant code.
This bug was another case of me not thinking in terms of tuple-input/tuple-output primitives, since those are rare (and while_loop is one of them).
fixes #489