0221更新

Author: zhiyunlee

.DS_Store

@@ -15,7 +15,7 @@ RUN sed -i 's/deb.debian.org/mirrors.aliyun.com/g' /etc/apt/sources.list && \
     apt install -y build-essential curl procps git libgl1 ffmpeg python3 python3-pip python-is-python3 && \
     mkdir -p $NVM_DIR && \
     # 安装 Node.js 环境
-    bash /opt/inference/xinference/deploy/docker/nvm_v0.39.7_install.sh && \
+    bash /opt/inference/xinference/deploy/docker/nvm-v0.39.7-install.sh && \
     . $NVM_DIR/nvm.sh && \
     nvm install $NODE_VERSION && \
     nvm alias default $NODE_VERSION && \
@@ -34,11 +34,11 @@ ENV PIP_INDEX_URL=https://mirrors.aliyun.com/pypi/simple \
 # 安装 Python 依赖
 RUN python3 -m pip install --upgrade  pip && \
     # 安装 PyTorch 的 CPU 版本（适配 ARM64）
-    pip install torch torchvision torchaudio sentence-transformers  && \    
-    # 安装 requirements_cpu.txt 中列出的依赖（仅当需要时）
-    #pip install -i "$PIP_INDEX" --upgrade-strategy only-if-needed -r /opt/inference/xinference/deploy/docker/requirements_cpu.txt && \
-    # 安装 llama-cpp-python
-    #CMAKE_ARGS="-DLLAMA_METAL=on" pip install llama-cpp-python && \
+    pip install torch torchvision torchaudio sentence-transformers  && \
+    # # 安装 llama-cpp-python
+    # CMAKE_ARGS="-DLLAMA_METAL=on" pip install llama-cpp-python && \
+    # # 安装 requirements_cpu.txt 中列出的依赖（仅当需要时）
+    # pip install -i "$PIP_INDEX" --upgrade-strategy only-if-needed -r /opt/inference/xinference/deploy/docker/requirements_cpu.txt && \
 
     #构建项目
     cd /opt/inference && \

inference-main/xinference/deploy/docker/cpu_arm64.Dockerfile

@@ -0,0 +1,23 @@
+version: '4.38.0'
+services:
+  xinference:
+    container_name: Xinference:0221
+    image: xinference:1.0-cpu-arm64
+
+    # Docker Compose 开机启动容器
+    restart: always
+    environment:
+      - XINFERENCE_MODEL_SRC=modelscope
+    ports:
+      - "9998:9997"
+    
+    # 容器开机启动 xinference服务、bge-reranker-v2-m3模型
+    command: 
+      sh -c 
+      "xinference-local -H 0.0.0.0 --log-level debug &
+        until curl -sf http://localhost:9997/status; do sleep 1; done &&
+        xinference launch --model-name bge-reranker-v2-m3 --model-type rerank &&
+        tail -f /dev/null
+      "
+      # apt-get -y install netcat &&
+      # while ! nc -z localhost 9997; do sleep 10; done &&