[TVMScript][Relax] Use tir.SizeVar for shape variables #16949
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
As reported in apache#16877, shape inference performed during a Relax transformation may produce different results than shape inference performed during TVMScript parsing. While Relax transformations call `Analyzer::MarkGlobalNonNegValue` for each shape expression, this is not provided during TVMScript parsing. As a result, output shapes that are conditional on the sign of a variable may produce different results when simplifying. This commit provides a partial resolution for this issue. Where prior to this commit, the TVMScript parser generated a `tir.Var` for each symbolic variable, the TVMScript parser now provides a `tir.SizeVar` for symbolic variables that are defined in contexts that require a non-negative size.
Member
|
This is a nice PR that would be good to land it. cc @yongwww |
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Outdated suggestions cannot be applied.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.
As reported in #16877, shape inference performed during a Relax transformation may produce different results than shape inference performed during TVMScript parsing. While Relax transformations call
Analyzer::MarkGlobalNonNegValuefor each shape expression, this is not provided during TVMScript parsing. As a result, output shapes that are conditional on the sign of a variable may produce different results when simplifying.This commit provides a partial resolution for this issue. Where prior to this commit, the TVMScript parser generated a
tir.Varfor each symbolic variable, the TVMScript parser now provides atir.SizeVarfor symbolic variables that are defined in contexts that require a non-negative size.