hailin
/
vllm
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

.

This commit is contained in:
hailin 2025-11-16 12:41:39 +08:00
parent 66db2696fa
commit a1b1522766
2 changed files with 3 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
supervisord.conf
View File

@ -5,7 +5,7 @@ logfile_maxbytes=0
loglevel=info
[program:vllm]
command=python3 -m vllm.entrypoints.openai.api_server --model /root/.cradle/Alibaba/Qwen3-30B-A3B --port 30000 --api-key token-abc123
command=python3 -m vllm.entrypoints.openai.api_server --model /root/.cradle/Alibaba/Qwen3-4B --port 30000 --api-key token-abc123
autostart=true
autorestart=true
stdout_logfile=/dev/stdout

7
vllm_v0.10.0/Dockerfile
View File

@ -120,9 +120,6 @@ RUN --mount=type=cache,target=/root/.cache/uv \
--extra-index-url ${PYTORCH_CUDA_INDEX_BASE_URL}/cu$(echo $CUDA_VERSION | cut -d. -f1,2 | tr -d '.')
COPY ./vllm_v0.10.0 /workspace
# ARG GIT_REPO_CHECK=0
# RUN --mount=type=bind,source=.git,target=.git \
# if [ "$GIT_REPO_CHECK" != "0" ]; then bash tools/check_repo.sh ; fi
# max jobs used by Ninja to build extensions
ARG max_jobs=96
@ -357,8 +354,8 @@ RUN apt-get update && apt-get install -y supervisor && mkdir -p /etc/supervisor/
# ---- Copy UI and supervisor config ----
COPY ./meta_ui.py /app/meta_ui.py
COPY ./supervisord.conf /etc/supervisor/supervisord.conf
COPY ./Alibaba/Qwen3-30B-A3B /root/.cradle/Alibaba/Qwen3-30B-A3B
# COPY ./Alibaba/Qwen3-30B-A3B /root/.cradle/Alibaba/Qwen3-30B-A3B
COPY ./Alibaba/Qwen3-30B-A3B /root/.cradle/Alibaba/Qwen3-4B
FROM vllm-openai-base AS vllm-openai
CMD ["/usr/bin/supervisord", "-c", "/etc/supervisor/supervisord.conf"]