diff --git a/AI/vllm-ascend/0.7.3-torch_npu2.5.1-cann8.1.rc1-python3.10/22.03-lts/Dockerfile b/AI/vllm-ascend/0.7.3-torch_npu2.5.1-cann8.1.rc1-python3.10/22.03-lts/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..2bf2e7077ec4b4dfced15a9de9fd79a110986a1f --- /dev/null +++ b/AI/vllm-ascend/0.7.3-torch_npu2.5.1-cann8.1.rc1-python3.10/22.03-lts/Dockerfile @@ -0,0 +1,32 @@ +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" + +RUN yum update -y && \ + yum install -y python3-pip git vim wget net-tools && \ + rm -rf /var/cache/yum + +WORKDIR /workspace + +RUN pip config set global.index-url ${PIP_INDEX_URL} + +# Install vLLM +ARG VLLM_REPO=https://github.com/vllm-project/vllm.git +ARG VLLM_TAG=v0.7.3 +RUN git clone --depth 1 $VLLM_REPO --branch $VLLM_TAG /vllm-workspace/vllm +RUN VLLM_TARGET_DEVICE="empty" python3 -m pip install -e /vllm-workspace/vllm/ --extra-index https://download.pytorch.org/whl/cpu/ && \ + python3 -m pip cache purge + +# Install vllm-ascend +ARG VLLM_ASCEND_REPO=https://github.com/vllm-project/vllm-ascend.git +ARG VLLM_ASCEND_TAG=v0.7.3 +RUN git clone $VLLM_ASCEND_REPO --branch $VLLM_ASCEND_TAG /vllm-workspace/vllm-ascend +RUN python3 -m pip install -e /vllm-workspace/vllm-ascend/ --extra-index https://download.pytorch.org/whl/cpu/ && \ + python3 -m pip cache purge + +# Install modelscope (for fast download) and ray (for multinode) +# TODO(yikun): Remove "<1.23.0" after v0.7.4 which resloved by https://github.com/vllm-project/vllm/pull/13807 +RUN python3 -m pip install "modelscope<1.23.0" ray && \ + python3 -m pip cache purge + +CMD ["/bin/bash"] \ No newline at end of file diff --git a/AI/vllm-ascend/0.8.4rc1-torch_npu2.5.1-cann8.0.0-python3.10/22.03-lts/Dockerfile b/AI/vllm-ascend/0.8.4rc1-torch_npu2.5.1-cann8.0.0-python3.10/22.03-lts/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..8c322d2a4c6a20933644f2cd3ab4c71616cac89d --- /dev/null +++ b/AI/vllm-ascend/0.8.4rc1-torch_npu2.5.1-cann8.0.0-python3.10/22.03-lts/Dockerfile @@ -0,0 +1,42 @@ +FROM quay.io/ascend/cann:8.0.0-910b-openeuler22.03-py3.10 + +ARG PIP_INDEX_URL="https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple" +ARG COMPILE_CUSTOM_KERNELS=0 + +ENV COMPILE_CUSTOM_KERNELS=${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 /workspace + +# Install vLLM +ARG VLLM_REPO=https://github.com/vllm-project/vllm.git +ARG VLLM_TAG=v0.8.4 + +RUN git clone --depth 1 $VLLM_REPO --branch $VLLM_TAG /workspace/vllm +RUN VLLM_TARGET_DEVICE="empty" python3 -m pip install /workspace/vllm/ --extra-index https://download.pytorch.org/whl/cpu/ +# In x86, triton will be installed by vllm. But in Ascend, triton doesn't work correctly. we need to uninstall it. +RUN python3 -m pip uninstall -y triton + +# Install torch-npu +ARG VLLM_ASCEND_REPO=https://github.com/vllm-project/vllm-ascend.git +ARG VLLM_ASCEND_TAG=v0.8.4rc1 +RUN git clone $VLLM_ASCEND_REPO --branch $VLLM_ASCEND_TAG /workspace/vllm-ascend +RUN bash /workspace/vllm-ascend/pta_install.sh + +# Install vllm-ascend +RUN source /usr/local/Ascend/ascend-toolkit/set_env.sh && \ + source /usr/local/Ascend/nnal/atb/set_env.sh && \ + export LD_LIBRARY_PATH=/usr/local/Ascend/ascend-toolkit/latest/`uname -i`-linux/devlib:$LD_LIBRARY_PATH && \ + export LIBRARY_PATH=/usr/local/Ascend/ascend-toolkit/latest/lib64:$LIBRARY_PATH && \ + python3 -m pip install -v /workspace/vllm-ascend/ --extra-index https://download.pytorch.org/whl/cpu/ + +# Install modelscope (for fast download) and ray (for multinode) +RUN python3 -m pip install modelscope ray + +CMD ["/bin/bash"] diff --git a/AI/vllm-ascend/0.8.5rc1-torch_npu2.5.1-cann8.1.rc1-python3.10/22.03-lts/Dockerfile b/AI/vllm-ascend/0.8.5rc1-torch_npu2.5.1-cann8.1.rc1-python3.10/22.03-lts/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..1b8d179aa7e53878c8a6d3b47b147945db5ba8bc --- /dev/null +++ b/AI/vllm-ascend/0.8.5rc1-torch_npu2.5.1-cann8.1.rc1-python3.10/22.03-lts/Dockerfile @@ -0,0 +1,59 @@ +# +# Copyright (c) 2025 Huawei Technologies Co., Ltd. All Rights Reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +# This file is a part of the vllm-ascend project. +# + +FROM quay.io/ascend/cann:8.1.rc1-910b-openeuler22.03-py3.10 + +ARG PIP_INDEX_URL="https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple" +ARG COMPILE_CUSTOM_KERNELS=1 + +ENV COMPILE_CUSTOM_KERNELS=${COMPILE_CUSTOM_KERNELS} + +RUN yum update -y && \ + yum install -y python3-pip git vim wget net-tools gcc gcc-c++ make cmake numactl-devel && \ + rm -rf /var/cache/yum + +RUN pip config set global.index-url ${PIP_INDEX_URL} + +WORKDIR /vllm-workspace + +# Install torch-npu +ARG VLLM_ASCEND_REPO=https://github.com/vllm-project/vllm-ascend.git +ARG VLLM_ASCEND_TAG=v0.8.5rc1 +RUN git clone $VLLM_ASCEND_REPO --branch $VLLM_ASCEND_TAG /vllm-workspace/vllm-ascend + +# Install vLLM +ARG VLLM_REPO=https://github.com/vllm-project/vllm.git +ARG VLLM_TAG=v0.8.5.post1 + +RUN git clone --depth 1 $VLLM_REPO --branch $VLLM_TAG /vllm-workspace/vllm +# In x86, triton will be installed by vllm. But in Ascend, triton doesn't work correctly. we need to uninstall it. +RUN VLLM_TARGET_DEVICE="empty" python3 -m pip install -e /vllm-workspace/vllm/ --extra-index https://download.pytorch.org/whl/cpu/ && \ + python3 -m pip uninstall -y triton && \ + python3 -m pip cache purge + +# Install vllm-ascend +RUN source /usr/local/Ascend/ascend-toolkit/set_env.sh && \ + source /usr/local/Ascend/nnal/atb/set_env.sh && \ + export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/Ascend/ascend-toolkit/latest/`uname -i`-linux/devlib && \ + python3 -m pip install -v -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 && \ + python3 -m pip cache purge + +CMD ["/bin/bash"] diff --git a/AI/vllm-ascend/README.md b/AI/vllm-ascend/README.md index 5b99922e727f4600ac2ee4b31905d0ff7d968d0a..99d13227a43ff4c2a84c8e6d009066fc63a34be0 100644 --- a/AI/vllm-ascend/README.md +++ b/AI/vllm-ascend/README.md @@ -24,7 +24,11 @@ The tag of each vLLM Ascend docker image is consist of the version of vLLM Ascen | Tags | Currently | Architectures| |--|--|--| -|[0.7.3rc2-torch_npu2.5.1-cann8.0.0-python3.10-oe2203lts](https://gitee.com/openeuler/openeuler-docker-images/blob/master/AI/opea/chatqna/1.0/24.03-lts/Dockerfile)| vLLM Ascend 0.7.3rc2 on openEuler 22.03-LTS | amd64, arm64 | +|[0.7.3rc2-torch_npu2.5.1-cann8.0.0-python3.10-oe2203lts](https://gitee.com/openeuler/openeuler-docker-images/blob/master/AI/vllm-ascend/0.7.3rc2-torch_npu2.5.1-cann8.0.0-python3.10/22.03-lts/Dockerfile)| vLLM Ascend 0.7.3rc2 on openEuler 22.03-LTS | amd64, arm64 | + +|[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 | + +|[0.8.5rc1-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 | # Usage diff --git a/AI/vllm-ascend/meta.yml b/AI/vllm-ascend/meta.yml index 5139beac0051a40c2313b18e4b302950ff3721f4..bbde9e7caa24808b6abdfcc4c079dd67155a5765 100644 --- a/AI/vllm-ascend/meta.yml +++ b/AI/vllm-ascend/meta.yml @@ -1,2 +1,11 @@ 0.7.3rc2-torch_npu2.5.1-cann8.0.0-python3.10-oe2203lts: path: 0.7.3rc2-torch_npu2.5.1-cann8.0.0-python3.10/22.03-lts/Dockerfile + +0.7.3-torch_npu2.5.1-cann8.1.rc1-python3.10-oe2203lts: + path: 0.7.3-torch_npu2.5.1-cann8.1.rc1-python3.10/22.03-lts/Dockerfile + +0.8.4rc1-torch_npu2.5.1-cann8.0.0-python3.10-oe2203lts: + path: 0.8.4rc1-torch_npu2.5.1-cann8.0.0-python3.10/22.03-lts/Dockerfile + +0.8.5rc1-torch_npu2.5.1-cann8.1.rc1-python3.10-oe2203lts: + path: 0.8.5rc1-torch_npu2.5.1-cann8.1.rc1-python3.10/22.03-lts/Dockerfile