[CK_TILE] Support moe with up gemm #1793
Draft
+139
−65
Conversation
This file contains 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
huaiguxu
force-pushed
the
huaiguxu/dev/moe2
branch
from
bf07f14 to
b4a1c30
Compare
huaiguxu
force-pushed
the
huaiguxu/dev/moe2
branch
from
b4a1c30 to
2880f7a
Compare
huaiguxu
force-pushed
the
huaiguxu/dev/moe2
branch
from
78e7aae to
af02d65
Compare
|
Please note that 2 function call to the |
carlushuang
reviewed
Jan 11, 2025
| typename Ts_::IndexDataType, | ||
| // ck_tile::element_wise::FastGeluAsm, // | ||
| // TODO: hardcoded | ||
| ck_tile::element_wise::Silu, |
There was a problem hiding this comment.
@huaiguxu
hardcode original activation to silu. If you need add new function into it, you really need to modify the template interface
cc @coderfeli
carlushuang
reviewed
Jan 12, 2025
| @@ -73,7 +73,7 @@ struct FusedMoeGemmPipeline_FlatmmUk | |||
| constexpr index_t smem_0 = Policy::template GetUK_0<Problem>().GetSmemSize(); | |||
| constexpr index_t smem_1 = Policy::template GetUK_1<Problem>().GetSmemSize(); | |||
| constexpr index_t smem_bridge = | |||
| BlockShape::Block_M0 * BlockShape::Block_N0 * sizeof(YDataType); | |||
| BlockShape::Block_M0 * BlockShape::Block_N0 * (IsGateOnly ? 1 : 2); | |||
There was a problem hiding this comment.
this is not correct, not multiply sizeop(dtype)
added 2 commits
January 13, 2025 02:44
This reverts commit e674fb7.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
Proposed changes
Support fused MoE with up gemm.
Checklist
Please put an
xinto the boxes that apply. You can also fill these out after creating the PR. If you're not sure, please don't hesitate to ask.clang-formaton all changed filesDiscussion
If this is a relatively large or complex change, feel free to start a discussion by explaining why you chose the solution you did and what alternatives you considered