Draft
Conversation
…edGemmKernel Support up to 4 B tensors with configurable L sizes via b_tensor_l_sizes parameter. Each B tensor has shape (n, k, l_i) where sum of all l_i equals total L. Signed-off-by: Zongfei Jing <20381269+zongfeijing@users.noreply.github.com>
Signed-off-by: Zongfei Jing <20381269+zongfeijing@users.noreply.github.com>
Signed-off-by: Zongfei Jing <20381269+zongfeijing@users.noreply.github.com>
Signed-off-by: Zongfei Jing <20381269+zongfeijing@users.noreply.github.com>
Extend the finalize fusion kernel and its benchmark driver to accept multiple B weights and scale factors so multi-B workloads can be exercised consistently. Signed-off-by: Zongfei Jing <20381269+zongfeijing@users.noreply.github.com>
Require list-based weight/scale/alpha inputs for finalize fusion and update the MoE path and unit test to pass list values consistently. Signed-off-by: Zongfei Jing <20381269+zongfeijing@users.noreply.github.com>
Extend finalize and gather FC1 tests to exercise multi-B input lists and verify outputs against the existing reference paths. Signed-off-by: Zongfei Jing <20381269+zongfeijing@users.noreply.github.com>
# Conflicts: # tensorrt_llm/_torch/custom_ops/cute_dsl_custom_ops.py # tensorrt_llm/_torch/modules/fused_moe/fused_moe_cutlass.py # tensorrt_llm/_torch/pyexecutor/_util.py # tensorrt_llm/_torch/pyexecutor/model_loader.py # tensorrt_llm/_torch/pyexecutor/py_executor.py
- Add cute_dsl_nvfp4_gather_grouped_gemm_swiglu_blackwell_multi_b op with List[tensor] interface for weight/weight_scale/alpha (multi-B / DWDP). - Restore cute_dsl_nvfp4_gather_grouped_gemm_swiglu_blackwell to single tensor interface; implementation delegates to _multi_b with [w],[ws],[a]. - Add register_fake for both ops for torch.compile/export shape inference. - Update fused_moe_cute_dsl and test_cute_dsl_moe to call _multi_b op.
# Conflicts: # tensorrt_llm/_torch/modules/fused_moe/fused_moe_cute_dsl.py # tensorrt_llm/_torch/pyexecutor/py_executor.py
svc-trtllm-gh-bot
added
the
Community want to contribute
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 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.
@coderabbitai summary
Description
Test Coverage
PR Checklist
Please review the following before submitting your PR:
PR description clearly explains what and why. If using CodeRabbit's summary, please make sure it makes sense.
PR Follows TRT-LLM CODING GUIDELINES to the best of your knowledge.
Test cases are provided for new code paths (see test instructions)
Any new dependencies have been scanned for license and vulnerabilities
CODEOWNERS updated if ownership changes
Documentation updated as needed
Update tava architecture diagram if there is a significant design change in PR.
The reviewers assigned automatically/manually are appropriate for the PR.
Please check this after reviewing the above items as appropriate for this PR.
GitHub Bot Help
To see a list of available CI bot commands, please comment
/bot help.