Add DCU backend and related support#197
Open
bjzhjing wants to merge 6 commits into
Open
Conversation
Fix invalid ClassVar annotation for Python 3.10. Correct moe_gating_gemm to read dst_dtype properly. Minor cleanup in launch.py/server.py initialization order (no behavior change intended). Signed-off-by: Zhang Jing <[email protected]>
Introduce BackendDCU and DCU backend entrypoint under src/ following existing backend structure. Signed-off-by: Zhang Jing <[email protected]>
Signed-off-by: Zhang Jing <[email protected]> Co-authored-by: Zhu Fuzhu <[email protected]>
Add DCU vendor provider packages (lightop/custom_ops/lmslim/flash_attn/torch) and their op implementations. Fix vendor ops to follow upstream lifecycle (vendor_parser/vendor_impl/vendor_impl_run) and ensure correct tensor mapping/output allocation where required. Signed-off-by: Zhang Jing <[email protected]> Co-authored-by: Zhu Fuzhu <[email protected]> Co-authored-by: Wang Sen <[email protected]> Co-authored-by: Zhuang Luo <[email protected]>
Add DCU-specific deploy configs for seed-oss-36b, qwen3-32b, and qwen3-235b-a22b tp-ep. Signed-off-by: Zhang Jing <[email protected]> Co-authored-by: Zhu Fuzhu <[email protected]> Co-authored-by: Wang Sen <[email protected]>
|
|
Signed-off-by: Zhang Jing <[email protected]> Co-authored-by: Zhu Fuzhu <[email protected]> Co-authored-by: Wang Sen <[email protected]>
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This PR adds DCU integration for micro_perf (DCU backend), extends llm_sim with DCU deploy configs, adds DCU vendor op implementations and vendor_test_DCU workloads, and includes small Python typing / micro_perf CLI fixes.