1ffd46b0f8
根因：[ -z "$VLLM_ROOT" ] && VLLM_ROOT=fallback 当VLLM_ROOT非空时（正常情况）, [ -z ] 返回1, && 链传播非零退出码, Docker RUN 步骤判定为失败，整个build中断。 修复：改用 if/fi 结构，不影响退出码： if [ -z "$VLLM_ROOT" ]; then VLLM_ROOT=fallback; fi 同时第一行用 ; 而不是 && 连接，避免python3命令的stderr输出 影响退出码传播。
52 lines
2.6 KiB
Docker
52 lines
2.6 KiB
Docker
FROM git.modelhub.org.cn:9443/enginex-iluvatar/bi100-3.2.3-x86-ubuntu20.04-py3.10-poc-llm-infer:v1.2.3
|
|
|
|
RUN mkdir -p /workspace
|
|
WORKDIR /workspace/
|
|
|
|
# Copy all sources
|
|
COPY ./qwen3_6_scripts /workspace/qwen3_6_scripts
|
|
COPY ./computility-run.yaml /workspace/computility-run.yaml
|
|
COPY ./ex_engine /workspace/ex_engine
|
|
|
|
# Step 1: Build EX Engine .so libraries (tolerant)
|
|
RUN chmod +x /workspace/ex_engine/build.sh && \
|
|
(bash /workspace/ex_engine/build.sh --corex 2>&1 || true) | tee /workspace/ex_build.log ; \
|
|
echo "[Dockerfile] ex_engine build done"
|
|
|
|
# Step 2: Precompile MoE CUDA kernels (tolerant)
|
|
RUN (python3 /workspace/ex_engine/precompile_moe_topk.py 2>&1 || true) | tee -a /workspace/ex_build.log ; \
|
|
echo "[Dockerfile] moe_topk precompile done"
|
|
|
|
# Step 3: Precompile vllm v0.5.5 MoE kernels (tolerant)
|
|
RUN (python3 /workspace/ex_engine/precompile_moe_kernels.py 2>&1 || true) | tee -a /workspace/ex_build.log ; \
|
|
echo "[Dockerfile] moe_v055 precompile done"
|
|
|
|
# Step 4: Deploy patches (serving + engine fixes) — tolerant
|
|
RUN chmod +x /workspace/qwen3_6_scripts/patch_ops.sh && \
|
|
(cd /workspace/qwen3_6_scripts && bash ./patch_ops.sh 2>&1 || true) | tee /workspace/patch_ops.log ; \
|
|
echo "[Dockerfile] patch_ops done"
|
|
|
|
# Step 5: Build ix_unified_bridge.so (tolerant — ixformer symbols resolved at runtime)
|
|
RUN if [ -f /workspace/ex_engine/build_unified_bridge.sh ]; then \
|
|
chmod +x /workspace/ex_engine/build_unified_bridge.sh && \
|
|
(bash /workspace/ex_engine/build_unified_bridge.sh 2>&1 || true) | tee -a /workspace/ex_build.log ; \
|
|
fi ; \
|
|
echo "[Dockerfile] bridge build done"
|
|
|
|
# Step 6: Deploy ix_unified Python modules to vllm path (tolerant)
|
|
RUN VLLM_ROOT="$(python3 -c 'import vllm; print(vllm.__path__[0])' 2>/dev/null | grep -v '^INFO\|^WARNING\|^DEBUG' | tail -1)" ; \
|
|
if [ -z "$VLLM_ROOT" ]; then VLLM_ROOT="/usr/local/corex/lib64/python3/dist-packages/vllm"; fi ; \
|
|
echo "[Dockerfile] VLLM_ROOT=${VLLM_ROOT}" && \
|
|
for f in ix_unified.py corex_so_loader.py moe_fused_dispatch.py corex_moe.py; do \
|
|
[ -f "/workspace/ex_engine/python/$f" ] && cp "/workspace/ex_engine/python/$f" "${VLLM_ROOT}/$f" 2>/dev/null || true ; \
|
|
done ; \
|
|
for _so in /workspace/ex_engine/build/ix_unified_bridge*.so; do \
|
|
[ -f "$_so" ] && cp "$_so" "${VLLM_ROOT}/" 2>/dev/null || true ; \
|
|
done ; \
|
|
echo "[Dockerfile] bridge deploy done"
|
|
|
|
# Step 7: Precompile GDN kernel (tolerant)
|
|
RUN (python3 /workspace/qwen3_6_scripts/precompile_gdn.py \
|
|
/workspace/qwen3_6_scripts/flash_qla_sm70 2>&1 || true) | tee -a /workspace/ex_build.log ; \
|
|
echo "[Dockerfile] gdn precompile done"
|