JAX FP8 matmul fusion tutorial. #133
Merged
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
balancap
force-pushed
the
fp8-matmul-fusion-tutorial
branch
from
4a84fcf to
1bbe9e6
Compare
lyprince
force-pushed
the
fp8-matmul-fusion-tutorial
branch
2 times, most recently
from
9fef477 to
d7713cb
Compare
In this tutorial notebook, we investigate how the ML stack JAX + XLA handles the specificities of FP8 matmuls, while still generating an optimal fused kernel call including: * FP8 inputs scaling; * FP8 output scaling & clamping; * Non-linearity & bias fusing; * Abs-max output capture; Note: some open questions remain on bias or gelu fusing.
lyprince
force-pushed
the
fp8-matmul-fusion-tutorial
branch
from
d7713cb to
a203df7
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.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.
In this tutorial notebook, we investigate how the ML stack JAX + XLA handles the specificities of FP8 matmuls,
while still generating an optimal fused kernel call including:
Note: some open questions remain on bias or gelu fusing.