[feat] add ascend readme and docker release

.github/workflows/pr-test-npu.yml

@@ -27,13 +27,19 @@ jobs:
         github.event.pull_request.draft == false
     runs-on: linux-arm64-npu-1
     container:
-      image: swr.cn-southwest-2.myhuaweicloud.com/base_image/ascend-ci/cann:8.2.rc1.alpha003-910b-ubuntu22.04-py3.11
+      image: swr.cn-southwest-2.myhuaweicloud.com/base_image/ascend-ci/cann:8.2.rc1-910b-ubuntu22.04-py3.11
     steps:
       - name: Checkout code
         uses: actions/checkout@v4
 
       - name: Install dependencies
         run: |
+          # speed up by using infra cache services
+          CACHING_URL="cache-service.nginx-pypi-cache.svc.cluster.local"
+          sed -Ei "s@(ports|archive).ubuntu.com@${CACHING_URL}:8081@g" /etc/apt/sources.list
+          pip config set global.index-url http://${CACHING_URL}/pypi/simple
+          pip config set global.trusted-host ${CACHING_URL}
+
           bash scripts/ci/npu_ci_install_dependency.sh
           # copy required file from our daily cache
           cp ~/.cache/modelscope/hub/datasets/otavia/ShareGPT_Vicuna_unfiltered/ShareGPT_V3_unfiltered_cleaned_split.json /tmp
@@ -56,13 +62,19 @@ jobs:
         github.event.pull_request.draft == false
     runs-on: linux-arm64-npu-2
     container:
-      image: swr.cn-southwest-2.myhuaweicloud.com/base_image/ascend-ci/cann:8.2.rc1.alpha003-910b-ubuntu22.04-py3.11
+      image: swr.cn-southwest-2.myhuaweicloud.com/base_image/ascend-ci/cann:8.2.rc1-910b-ubuntu22.04-py3.11
     steps:
       - name: Checkout code
         uses: actions/checkout@v4
 
       - name: Install dependencies
         run: |
+          # speed up by using infra cache services
+          CACHING_URL="cache-service.nginx-pypi-cache.svc.cluster.local"
+          sed -Ei "s@(ports|archive).ubuntu.com@${CACHING_URL}:8081@g" /etc/apt/sources.list
+          pip config set global.index-url http://${CACHING_URL}/pypi/simple
+          pip config set global.trusted-host ${CACHING_URL}
+
           bash scripts/ci/npu_ci_install_dependency.sh
           # copy required file from our daily cache
           cp ~/.cache/modelscope/hub/datasets/otavia/ShareGPT_Vicuna_unfiltered/ShareGPT_V3_unfiltered_cleaned_split.json /tmp
@@ -85,13 +97,19 @@ jobs:
         github.event.pull_request.draft == false
     runs-on: linux-arm64-npu-4
     container:
-      image: swr.cn-southwest-2.myhuaweicloud.com/base_image/ascend-ci/cann:8.2.rc1.alpha003-910b-ubuntu22.04-py3.11
+      image: swr.cn-southwest-2.myhuaweicloud.com/base_image/ascend-ci/cann:8.2.rc1-910b-ubuntu22.04-py3.11
     steps:
       - name: Checkout code
         uses: actions/checkout@v4
 
       - name: Install dependencies
         run: |
+          # speed up by using infra cache services
+          CACHING_URL="cache-service.nginx-pypi-cache.svc.cluster.local"
+          sed -Ei "s@(ports|archive).ubuntu.com@${CACHING_URL}:8081@g" /etc/apt/sources.list
+          pip config set global.index-url http://${CACHING_URL}/pypi/simple
+          pip config set global.trusted-host ${CACHING_URL}
+
           bash scripts/ci/npu_ci_install_dependency.sh
           # copy required file from our daily cache
           cp ~/.cache/modelscope/hub/datasets/otavia/ShareGPT_Vicuna_unfiltered/ShareGPT_V3_unfiltered_cleaned_split.json /tmp

.github/workflows/release-docker-npu-nightly.yaml

@@ -0,0 +1,76 @@
+name: Release Docker Images Nightly (Ascend NPU)
+on:
+  pull_request:
+    branches:
+      - main
+    paths:
+      - ".github/workflows/release-docker-npu-nightly.yaml"
+  workflow_dispatch:
+  schedule:
+    - cron: "0 0 * * *"
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.sha }}
+  cancel-in-progress: true
+
+jobs:
+  build:
+    runs-on: ubuntu-22.04-arm
+    strategy:
+      matrix:
+        cann_version: ["8.2.rc1"]
+        device_type: ["a3"]
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+
+      - name: Free up disk space
+        uses: jlumbroso/free-disk-space@54081f138730dfa15788a46383842cd2f914a1be # v1.3.1
+        with:
+          tool-cache: true
+          docker-images: false
+
+      - name: Setup Docker buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Docker meta
+        id: meta
+        uses: docker/metadata-action@v5
+        with:
+          images: |
+            ${{ github.repository_owner }}/sglang
+          # push with schedule event
+          # push with workflow_dispatch event
+          tags: |
+            type=ref,event=pr
+            type=ref,event=branch
+            type=schedule,pattern=main
+          flavor: |
+            latest=false
+            suffix=-cann${{ matrix.cann_version }}-${{ matrix.device_type }},onlatest=true
+      # Login against a Docker registry except on PR
+      # https://github.com/docker/login-action
+      - name: Log into docker hub
+        uses: docker/login-action@v3
+        if: ${{ github.repository_owner == 'sgl-project/sglang' && github.event_name != 'pull_request' }}
+        with:
+          username: ${{ secrets.DOCKERHUB_USERNAME }}
+          password: ${{ secrets.DOCKERHUB_TOKEN }}
+
+      # Build and push Docker image with Buildx (don't push on PR)
+      # https://github.com/docker/build-push-action
+      - name: Build and push Docker image
+        id: build-and-push
+        uses: docker/build-push-action@v6
+        with:
+          context: docker
+          file: docker/Dockerfile.npu
+          # TODO: need add x86 platforms support when memfabric is ready
+          platforms: linux/arm64
+          labels: ${{ steps.meta.outputs.labels }}
+          tags: ${{ steps.meta.outputs.tags }}
+          push: ${{ github.repository_owner == 'sgl-project/sglang' && github.event_name != 'pull_request' }}
+          provenance: false
+          build-args: |
+            CANN_VERSION=${{ matrix.cann_version }}
+            DEVICE_TYPE=${{ matrix.device_type }}

.github/workflows/release-docker-npu.yaml

@@ -0,0 +1,77 @@
+name: Release Docker Images (Ascend NPU)
+on:
+  push:
+    tags:
+      - "*" # Trigger on all tags and filterred by pep440 later
+  workflow_dispatch:
+  pull_request:
+    branches:
+      - main
+    paths:
+      - ".github/workflows/release-docker-npu.yaml"
+
+jobs:
+  build:
+    runs-on: ubuntu-22.04-arm
+    strategy:
+      matrix:
+        cann_version: ["8.2.rc1"]
+        device_type: ["a3"]
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+
+      - name: Free up disk space
+        uses: jlumbroso/free-disk-space@54081f138730dfa15788a46383842cd2f914a1be # v1.3.1
+        with:
+          tool-cache: true
+          docker-images: false
+
+        # push with tag
+      - name: Docker meta
+        id: meta
+        uses: docker/metadata-action@v5
+        with:
+          images: |
+            ${{ github.repository_owner }}/sglang
+          tags: |
+            type=ref,event=pr
+            type=ref,event=tag,suffix=-cann${{ matrix.cann_version }}-${{ matrix.device_type }}
+          flavor: |
+            latest=false
+      - name: Setup Docker buildx
+        uses: docker/setup-buildx-action@v3
+
+      # Login against a Docker registry except on PR
+      # https://github.com/docker/login-action
+      - name: Login to Docker Hub
+        uses: docker/login-action@v2
+        if: ${{ github.repository_owner == 'sgl-project/sglang' && github.event_name != 'pull_request' }}
+        with:
+          username: ${{ secrets.DOCKERHUB_USERNAME }}
+          password: ${{ secrets.DOCKERHUB_TOKEN }}
+
+      - name: Get version
+        id: get_version
+        run: |
+          version=$(cat python/sglang/version.py | cut -d'"' -f2)
+          echo "TAG=${{ github.repository_owner }}/sglang:v$version-cann${{ matrix.cann_version }}-${{ matrix.device_type }}" >> $GITHUB_OUTPUT
+          kernel_tag=$(curl -s https://api.github.com/repos/sgl-project/sgl-kernel-npu/tags | jq -r '.[0].name')
+          echo "KERNEL_NPU_TAG=${kernel_tag}" >> $GITHUB_OUTPUT
+
+      - name: Build and push Docker image
+        id: build-and-push
+        uses: docker/build-push-action@v6
+        with:
+          context: docker
+          file: docker/Dockerfile.npu
+          # TODO: need add x86 platforms support when memfabric is ready
+          platforms: linux/arm64
+          labels: ${{ steps.meta.outputs.labels }}
+          tags: ${{ steps.meta.outputs.tags || steps.get_version.outputs.TAG }}
+          push: ${{ github.repository_owner == 'sgl-project/sglang' && github.event_name != 'pull_request' }}
+          provenance: false
+          build-args: |
+            SGLANG_KERNEL_NPU_TAG=${{ steps.get_version.outputs.KERNEL_NPU_TAG }}
+            CANN_VERSION=${{ matrix.cann_version }}
+            DEVICE_TYPE=${{ matrix.device_type }}

docker/Dockerfile.npu

@@ -0,0 +1,81 @@
+ARG CANN_VERSION=8.2.rc1
+ARG DEVICE_TYPE=a3
+ARG OS=ubuntu22.04
+ARG PYTHON_VERSION=py3.11
+
+FROM quay.io/ascend/cann:$CANN_VERSION-$DEVICE_TYPE-$OS-$PYTHON_VERSION
+
+# Update pip & apt sources
+ARG PIP_INDEX_URL="https://pypi.org/simple/"
+ARG APTMIRROR=""
+ARG MEMFABRIC_URL=https://sglang-ascend.obs.cn-east-3.myhuaweicloud.com/sglang/mf_adapter-1.0.0-cp311-cp311-linux_aarch64.whl
+ARG PYTORCH_VERSION=2.6.0
+ARG TORCHVISION_VERSION=0.21.0
+ARG PTA_URL="https://gitee.com/ascend/pytorch/releases/download/v7.1.0.1-pytorch2.6.0/torch_npu-2.6.0.post1-cp311-cp311-manylinux_2_28_aarch64.whl"
+ARG VLLM_TAG=v0.8.5
+ARG TRITON_ASCEND_URL=https://sglang-ascend.obs.cn-east-3.myhuaweicloud.com/sglang/triton_ascend-3.2.0.dev20250729-cp311-cp311-manylinux_2_27_aarch64.manylinux_2_28_aarch64.whl
+ARG SGLANG_TAG=main
+ARG ASCEND_CANN_PATH=/usr/local/Ascend/ascend-toolkit
+ARG SGLANG_KERNEL_NPU_TAG=main
+
+WORKDIR /workspace
+
+# Define environments
+ENV DEBIAN_FRONTEND=noninteractive
+
+RUN pip config set global.index-url $PIP_INDEX_URL
+RUN if [ -n "$APTMIRROR" ];then sed -i "s|.*.ubuntu.com|$APTMIRROR|g" /etc/apt/sources.list ;fi
+
+# Install development tools and utilities
+RUN apt-get update -y && apt upgrade -y && apt-get install -y \
+    build-essential \
+    cmake \
+    vim \
+    wget \
+    curl \
+    net-tools \
+    zlib1g-dev \
+    lld \
+    clang \
+    locales \
+    ccache \
+    ca-certificates \
+    && rm -rf /var/cache/apt/* \
+    && rm -rf /var/lib/apt/lists/* \
+    && update-ca-certificates \
+    && locale-gen en_US.UTF-8
+
+ENV LANG=en_US.UTF-8
+ENV LANGUAGE=en_US:en
+ENV LC_ALL=en_US.UTF-8
+
+# Install dependencies
+# TODO: install from pypi released memfabric
+# TODO: install from pypi released triton-ascend
+RUN pip install $MEMFABRIC_URL --no-cache-dir \
+    && pip install torch==$PYTORCH_VERSION torchvision==$TORCHVISION_VERSION --index-url https://download.pytorch.org/whl/cpu --no-cache-dir \
+    && wget ${PTA_URL} && pip install "./torch_npu-2.6.0.post1-cp311-cp311-manylinux_2_28_aarch64.whl" --no-cache-dir \
+    && pip install ${TRITON_ASCEND_URL} --no-cache-dir \
+    && python3 -m pip install --no-cache-dir numpy==1.26.4 pybind11
+
+# Install vLLM
+RUN git clone --depth 1 https://github.com/vllm-project/vllm.git --branch $VLLM_TAG && \
+    cd vllm && VLLM_TARGET_DEVICE="empty" pip install -v . --no-cache-dir && \
+    cd .. && rm -rf vllm
+
+# Install SGLang
+RUN git clone https://github.com/sgl-project/sglang --branch $SGLANG_TAG && \
+    cd ./sglang/python && pip install .[srt_npu] --no-cache-dir && \
+    cd .. && rm -rf ./sglang
+
+# Install Deep-ep
+RUN git clone  --branch $SGLANG_KERNEL_NPU_TAG https://github.com/sgl-project/sgl-kernel-npu.git \
+    && export LD_LIBRARY_PATH=${ASCEND_CANN_PATH}/latest/runtime/lib64/stub:$LD_LIBRARY_PATH && \
+    source ${ASCEND_CANN_PATH}/set_env.sh && \
+    cd sgl-kernel-npu && \
+    bash build.sh \
+    && pip install output/deep_ep*.whl --no-cache-dir \
+    && cd .. && rm -rf sgl-kernel-npu \
+    && cd "$(pip show deep-ep | awk '/^Location:/ {print $2}')" && ln -s deep_ep/deep_ep_cpp*.so
+
+CMD ["/bin/bash"]

docs/basic_usage/deepseek.md

@@ -24,6 +24,7 @@ To run DeepSeek V3/R1 models, the requirements are as follows:
 | **Quantized weights (int8)** | 16 x A100/800 |
 | | 32 x L40S |
 | | Xeon 6980P CPU |
+| | 2 x Atlas 800I A3 |
 
 <style>
 .md-typeset__table {
@@ -64,6 +65,7 @@ Detailed commands for reference:
 - [16 x A100 (int8)](https://github.com/sgl-project/sglang/tree/main/benchmark/deepseek_v3#example-serving-with-16-a100a800-with-int8-quantization)
 - [32 x L40S (int8)](https://github.com/sgl-project/sglang/tree/main/benchmark/deepseek_v3#example-serving-with-32-l40s-with-int8-quantization)
 - [Xeon 6980P CPU](../platforms/cpu_server.md#example-running-deepseek-r1)
+- [2 x Atlas 800I A3 (int8)](../platforms/ascend_npu.md#running-deepseek-v3)
 
 ### Download Weights
 If you encounter errors when starting the server, ensure the weights have finished downloading. It's recommended to download them beforehand or restart multiple times until all weights are downloaded. Please refer to [DeepSeek V3](https://huggingface.co/deepseek-ai/DeepSeek-V3-Base#61-inference-with-deepseek-infer-demo-example-only) official guide to download the weights.