[Runtime][Builtin] Using float32 accumulation in attention kernel #16667
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
Prior to this PR, the TIR attention kernels does not cast matmul operands to fp32 before multiplying. For models like Phi-2 which may have large Q/K/V data (at the level of a few hundreds), the fp16 multiplication exceeds the range of fp16, and lead to attention result being NAN sometimes. This PR fixes this issue.
Contributor
Author
|
@tvm-bot rerun |
yongwww
approved these changes
Mar 4, 2024
Lunderberg
pushed a commit
to Lunderberg/tvm
that referenced
this pull request
Mar 12, 2024
…ache#16667) Prior to this PR, the TIR attention kernels does not cast matmul operands to fp32 before multiplying. For models like Phi-2 which may have large Q/K/V data (at the level of a few hundreds), the fp16 multiplication exceeds the range of fp16, and lead to attention result being NAN sometimes. This PR fixes this issue.
thaisacs
pushed a commit
to thaisacs/tvm
that referenced
this pull request
Apr 3, 2024
…ache#16667) Prior to this PR, the TIR attention kernels does not cast matmul operands to fp32 before multiplying. For models like Phi-2 which may have large Q/K/V data (at the level of a few hundreds), the fp16 multiplication exceeds the range of fp16, and lead to attention result being NAN sometimes. This PR fixes this issue.
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Outdated suggestions cannot be applied.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.
Prior to this PR, the TIR attention kernels does not cast matmul operands to fp32 before multiplying.
For models like Phi-2 which may have large Q/K/V data (at the level of a few hundreds), the fp16 multiplication exceeds the range of fp16, and lead to attention result being NAN sometimes.
This PR fixes this issue.