dylanyunlong/project_6
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

under test, not sure no errors

Browse Source
This commit is contained in:
DP Migration
2026-09-02 01:00:37 +00:00
parent 89d522223f
commit 27878c8689
1 changed files with 12 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
qwen3_6_scripts/ex_engine/build_xllm_kernels.sh
View File

@@ -148,6 +148,18 @@ echo "============================================================"
build_so "xllm_moe" \ build_so "xllm_moe" \
"ex_engine/xllm_kernels/cuda/moe/moe_fused_topk.cu ex_engine/xllm_kernels/cuda/moe/moe_compute_index.cu ex_engine/xllm_kernels/cuda/moe/moe_combine.cu ex_engine/xllm_kernels/cuda/bindings/xllm_moe_bind.cpp" "ex_engine/xllm_kernels/cuda/moe/moe_fused_topk.cu ex_engine/xllm_kernels/cuda/moe/moe_compute_index.cu ex_engine/xllm_kernels/cuda/moe/moe_combine.cu ex_engine/xllm_kernels/cuda/bindings/xllm_moe_bind.cpp"
# 7. py_attention_metadata — DP-aware attention metadata (pybind11, pure C++)
# Source: core/runtime/py_attention_metadata.cpp
# Ported from xLLM upstream PR #2258 (commit 78aa2a85)
# Exports: AttentionMetadataView with dp_token_counts / dp_is_decode fields
# Note: Pure C++ pybind11 — no CUDA, no .cu files
echo ""
echo "============================================================"
echo " 7. py_attention_metadata.so"
echo "============================================================"
build_so "py_attention_metadata" \
"core/runtime/py_attention_metadata.cpp"
echo "" echo ""
echo "============================================================" echo "============================================================"
echo " Build complete. Output:" echo " Build complete. Output:"