From 1c40564a4e75deaf3481a18145d28465602be929 Mon Sep 17 00:00:00 2001 From: zhihang Date: Tue, 10 Jun 2025 13:50:11 +0800 Subject: [PATCH 1/2] add vllm-ascend 0.9.0rc1 --- .../22.03-lts/Dockerfile | 41 +++++++++++++++++++ AI/vllm-ascend/README.md | 1 + AI/vllm/README.md | 1 + 3 files changed, 43 insertions(+) create mode 100644 AI/vllm-ascend/0.9.0rc1-torch_npu2.5.1-cann8.1.rc1-python3.10/22.03-lts/Dockerfile diff --git a/AI/vllm-ascend/0.9.0rc1-torch_npu2.5.1-cann8.1.rc1-python3.10/22.03-lts/Dockerfile b/AI/vllm-ascend/0.9.0rc1-torch_npu2.5.1-cann8.1.rc1-python3.10/22.03-lts/Dockerfile new file mode 100644 index 0000000..8248b98 --- /dev/null +++ b/AI/vllm-ascend/0.9.0rc1-torch_npu2.5.1-cann8.1.rc1-python3.10/22.03-lts/Dockerfile @@ -0,0 +1,41 @@ +FROM quay.io/ascend/cann:8.1.rc1-910b-openeuler22.03-py3.10 + +ARG PIP_INDEX_URL="https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple" +ARG COMPILE_CUSTOM_KERNELS=1 + +ENV COMPILE_CUSTOM_KERNELS=${COMPILE_CUSTOM_KERNELS} + +RUN yum update -y && \ + yum install -y python3-pip git vim wget net-tools make gcc gcc-c++ && \ + rm -rf /var/cache/yum &&\ + rm -rf /tmp/* + +RUN pip config set global.index-url ${PIP_INDEX_URL} + +WORKDIR /vllm-workspace + +# Install vLLM +ARG VLLM_REPO=https://github.com/vllm-project/vllm.git +ARG VLLM_ASCEND_REPO=https://github.com/vllm-project/vllm-ascend.git +ARG VLLM_TAG=v0.9.0 +ARG VLLM_ASCEND_TAG=v0.9.0rc1 + +RUN git clone $VLLM_REPO --branch $VLLM_TAG /vllm-workspace/vllm +# In x86, triton will be installed by vllm. But in Ascend, triton doesn't work correctly. we need to uninstall it. +RUN VLLM_TARGET_DEVICE="empty" python3 -m pip install -e /vllm-workspace/vllm/ --extra-index https://download.pytorch.org/whl/cpu/ && \ + python3 -m pip uninstall -y triton && \ + python3 -m pip cache purge + + +RUN git clone $VLLM_ASCEND_REPO --branch $VLLM_ASCEND_TAG /vllm-workspace/vllm-ascend +# Install vllm-ascend +RUN source /usr/local/Ascend/ascend-toolkit/set_env.sh && \ + source /usr/local/Ascend/nnal/atb/set_env.sh && \ + export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/Ascend/ascend-toolkit/latest/`uname -i`-linux/devlib && \ + python3 -m pip install -e /vllm-workspace/vllm-ascend/ --extra-index https://download.pytorch.org/whl/cpu/ && \ + python3 -m pip cache purge + +# Install modelscope (for fast download) and ray (for multinode) +RUN python3 -m pip install modelscope ray + +CMD ["/bin/bash"] \ No newline at end of file diff --git a/AI/vllm-ascend/README.md b/AI/vllm-ascend/README.md index a7b0bbe..8a0db90 100644 --- a/AI/vllm-ascend/README.md +++ b/AI/vllm-ascend/README.md @@ -28,6 +28,7 @@ The tag of each vLLM Ascend docker image is consist of the version of vLLM Ascen |[0.7.3-torch_npu2.5.1-cann8.1.rc1-python3.10-oe2203lts](https://gitee.com/openeuler/openeuler-docker-images/blob/master/AI/vllm-ascend/0.7.3-torch_npu2.5.1-cann8.1.rc1-python3.10/22.03-lts/Dockerfile)| vLLM Ascend 0.7.3 on openEuler 22.03-LTS | amd64, arm64 | |[0.8.4rc1-torch_npu2.5.1-cann8.0.0-python3.10-oe2203lts](https://gitee.com/openeuler/openeuler-docker-images/blob/master/AI/vllm-ascend/0.8.4rc1-torch_npu2.5.1-cann8.0.0-python3.10/22.03-lts/Dockerfile)| vLLM Ascend 0.8.4rc1 on openEuler 22.03-LTS | amd64, arm64 | |[0.8.5rc1-torch_npu2.5.1-cann8.1.rc1-python3.10-oe2203lts](https://gitee.com/openeuler/openeuler-docker-images/blob/master/AI/vllm-ascend/0.8.5rc1-torch_npu2.5.1-cann8.1.rc1-python3.10/22.03-lts/Dockerfile)| vLLM Ascend 0.8.5rc1 on openEuler 22.03-LTS | amd64, arm64 | +|[0.9.0rc1-torch_npu2.5.1-cann8.1.rc1-python3.10-oe2203lts](https://gitee.com/openeuler/openeuler-docker-images/blob/master/AI/vllm-ascend/0.9.0rc1-torch_npu2.5.1-cann8.1.rc1-python3.10/22.03-lts/Dockerfile)| vLLM Ascend 0.9.0rc1 on openEuler 22.03-LTS | amd64, arm64 | # Usage diff --git a/AI/vllm/README.md b/AI/vllm/README.md index 6747905..b04a423 100644 --- a/AI/vllm/README.md +++ b/AI/vllm/README.md @@ -38,6 +38,7 @@ The tag of each vLLM docker image is consist of the version of vLLM and the vers |[0.8.4-oe2403lts](https://gitee.com/openeuler/openeuler-docker-images/blob/master/AI/vllm/0.8.4/24.03-lts/Dockerfile)| vLLM 0.8.4 on openEuler 24.03-LTS | amd64 | |[0.8.5-oe2403lts](https://gitee.com/openeuler/openeuler-docker-images/blob/master/AI/vllm/0.8.5/24.03-lts/Dockerfile)| vLLM 0.8.5 on openEuler 24.03-LTS | amd64, arm64 | |[0.9.0-oe2403lts](https://gitee.com/openeuler/openeuler-docker-images/blob/master/AI/vllm/0.9.0/24.03-lts/Dockerfile)| vLLM 0.9.0 on openEuler 24.03-LTS | amd64, arm64 | + # Usage ## Quick start 1: supported devices -- Gitee From ec34caea8dc6a76227d87661f01bd2ef22de39d8 Mon Sep 17 00:00:00 2001 From: zhihang Date: Tue, 10 Jun 2025 14:01:34 +0800 Subject: [PATCH 2/2] add vllm 0.6.3, 0.8.3, 0.8.4, 0.8.5, 0.9.0 --- AI/image-list.yml | 2 +- AI/{vllm => vllm-cpu}/0.6.3/24.03-lts/Dockerfile | 0 AI/{vllm => vllm-cpu}/0.8.3/24.03-lts/Dockerfile | 0 AI/{vllm => vllm-cpu}/0.8.4/24.03-lts/Dockerfile | 0 AI/{vllm => vllm-cpu}/0.8.5/24.03-lts/Dockerfile | 0 AI/{vllm => vllm-cpu}/0.9.0/24.03-lts/Dockerfile | 0 AI/{vllm => vllm-cpu}/README.md | 0 AI/{vllm => vllm-cpu}/meta.yml | 0 8 files changed, 1 insertion(+), 1 deletion(-) rename AI/{vllm => vllm-cpu}/0.6.3/24.03-lts/Dockerfile (100%) rename AI/{vllm => vllm-cpu}/0.8.3/24.03-lts/Dockerfile (100%) rename AI/{vllm => vllm-cpu}/0.8.4/24.03-lts/Dockerfile (100%) rename AI/{vllm => vllm-cpu}/0.8.5/24.03-lts/Dockerfile (100%) rename AI/{vllm => vllm-cpu}/0.9.0/24.03-lts/Dockerfile (100%) rename AI/{vllm => vllm-cpu}/README.md (100%) rename AI/{vllm => vllm-cpu}/meta.yml (100%) diff --git a/AI/image-list.yml b/AI/image-list.yml index 6e214bb..ac2fc0d 100644 --- a/AI/image-list.yml +++ b/AI/image-list.yml @@ -51,7 +51,7 @@ images: pytorch: pytorch text-embeddings-inference-cpu: text-embeddings-inference-cpu text-generation-inference-cpu: text-generation-inference-cpu - vllm: vllm + vllm-cpu: vllm-cpu vllm-ascend: vllm-ascend langchain: langchain open-webui: open-webui diff --git a/AI/vllm/0.6.3/24.03-lts/Dockerfile b/AI/vllm-cpu/0.6.3/24.03-lts/Dockerfile similarity index 100% rename from AI/vllm/0.6.3/24.03-lts/Dockerfile rename to AI/vllm-cpu/0.6.3/24.03-lts/Dockerfile diff --git a/AI/vllm/0.8.3/24.03-lts/Dockerfile b/AI/vllm-cpu/0.8.3/24.03-lts/Dockerfile similarity index 100% rename from AI/vllm/0.8.3/24.03-lts/Dockerfile rename to AI/vllm-cpu/0.8.3/24.03-lts/Dockerfile diff --git a/AI/vllm/0.8.4/24.03-lts/Dockerfile b/AI/vllm-cpu/0.8.4/24.03-lts/Dockerfile similarity index 100% rename from AI/vllm/0.8.4/24.03-lts/Dockerfile rename to AI/vllm-cpu/0.8.4/24.03-lts/Dockerfile diff --git a/AI/vllm/0.8.5/24.03-lts/Dockerfile b/AI/vllm-cpu/0.8.5/24.03-lts/Dockerfile similarity index 100% rename from AI/vllm/0.8.5/24.03-lts/Dockerfile rename to AI/vllm-cpu/0.8.5/24.03-lts/Dockerfile diff --git a/AI/vllm/0.9.0/24.03-lts/Dockerfile b/AI/vllm-cpu/0.9.0/24.03-lts/Dockerfile similarity index 100% rename from AI/vllm/0.9.0/24.03-lts/Dockerfile rename to AI/vllm-cpu/0.9.0/24.03-lts/Dockerfile diff --git a/AI/vllm/README.md b/AI/vllm-cpu/README.md similarity index 100% rename from AI/vllm/README.md rename to AI/vllm-cpu/README.md diff --git a/AI/vllm/meta.yml b/AI/vllm-cpu/meta.yml similarity index 100% rename from AI/vllm/meta.yml rename to AI/vllm-cpu/meta.yml -- Gitee