如何与Docker一起使用特定的LLM型号组成

Question

我有一个docker组成，其中一个虚拟python应用程序在后台使用ollama lms来完成某些任务。我希望能够以某种方式告诉Ollama哪种模型要下载并在应用程序部署上使用。现在我的Docker组成就是这样：

version: '3.8'

services:
  elasticsearch:
    image: docker.elastic.co/elasticsearch/elasticsearch:7.10.0
    container_name: elasticsearch
    environment:
      - discovery.type=single-node
      - "ES_JAVA_OPTS=-Xms512m -Xmx512m"
    ports:
      - "9200:9200"
    volumes:
      - esdata:/usr/share/elasticsearch/data
    healthcheck:
      test: ["CMD-SHELL", "curl -f http://localhost:9200 || exit 1"]
      interval: 30s
      timeout: 10s
      retries: 5

  pokemon_app:
    build: .
    container_name: pokemon_app
    ports:
      - "8000:8000"
    depends_on:
      elasticsearch:
        condition: service_healthy
      ollama:
        condition: service_started
    environment:
      LLM_MODEL_NAME: ${LLM_MODEL_NAME}
      ELASTICSEARCH_URL: ${ELASTICSEARCH_URL}
      OLLAMA_URL: ${OLLAMA_URL}
      LOG_LEVEL: DEBUG
    healthcheck:
      test: ["CMD-SHELL", "curl -f http://localhost:8000/healthcheck || exit 1"]
      interval: 30s
      timeout: 10s
      retries: 5

  ollama:
    image: ollama/ollama:latest  
    container_name: ollama
    volumes:
      - ollama:/root/.ollama
    ports:
      - "11434:11434"
    healthcheck:
      test: ["CMD-SHELL", "curl -f http://localhost:11434 || exit 1"]
      interval: 30s
      timeout: 10s
      retries: 5

volumes:
  esdata:
    driver: local
  ollama:
    driver: local

我在应用中使用的默认LLM名称具有此ENV变量：

LLM_MODEL_NAME

我如何做到这一点，以便通过更改ENV变量的价值，我控制了Ollama使用哪种模型？

Answer 1

ok，所以我在某个地方显示我们可以在服务启动后安装所需的模型。因此，我现在使用称为

ollama

的ENV变量指定模型。
我的服务开始后，我有一个入口点bash脚本，称为

LLM_MODEL_VERSION

：

ollama

所以我的新yaml是：

ollama_entrypoint.sh

现在我指定的.env文件中：

#!/bin/bash

echo "Starting Ollama..."
/bin/ollama serve &

pid=$!

echo "Waiting for Ollama to start..."
until ollama list > /dev/null 2>&1; do
  sleep 2
done

MODEL_NAME=${LLM_MODEL_VERSION}

if ollama list | grep -q "$MODEL_NAME"; then
  echo "Model $MODEL_NAME already available."
else
  echo "Retrieving model: $MODEL_NAME"
  ollama pull "$MODEL_NAME"
  echo "Model $MODEL_NAME is ready!"
fi

wait $pid

然后

version: '3.8'

services:
  elasticsearch:
    image: docker.elastic.co/elasticsearch/elasticsearch:7.10.0
    container_name: elasticsearch
    environment:
      - discovery.type=single-node
      - "ES_JAVA_OPTS=-Xms512m -Xmx512m"
    ports:
      - "9200:9200"
    volumes:
      - esdata:/usr/share/elasticsearch/data
    healthcheck:
      test: ["CMD-SHELL", "curl -f http://localhost:9200 || exit 1"]
      interval: 30s
      timeout: 10s
      retries: 5

  pokemon_app:
    build: .
    container_name: pokemon_app
    ports:
      - "8000:8000"
    depends_on:
      - elasticsearch
      - ollama
    environment:
      - LLM_MODEL_VERSION=${LLM_MODEL_VERSION}
      - ELASTICSEARCH_URL=${ELASTICSEARCH_URL}
      - OLLAMA_URL=${OLLAMA_URL}
      - LOG_LEVEL=DEBUG
    healthcheck:
      test: ["CMD-SHELL", "curl -f http://localhost:8000/healthcheck || exit 1"]
      interval: 30s
      timeout: 10s
      retries: 5

  ollama:
    image: ollama/ollama:0.5.7 
    container_name: ollama
    volumes:
      - ollama:/root/.ollama  
      - ./ollama_entrypoint.sh:/ollama_entrypoint.sh 
    ports:
      - "11434:11434"
    environment:
      - LLM_MODEL_VERSION=${LLM_MODEL_VERSION} 
    entrypoint: ["/bin/bash", "/ollama_entrypoint.sh"] 
    healthcheck:
      test: ["CMD-SHELL", "curl -f http://localhost:11434 || exit 1"]
      interval: 30s
      timeout: 10s
      retries: 5

volumes:
  esdata:
    driver: local
  ollama:
    driver: local

服务在星星之后下载，我的pokemon_app也知道在来自同一环境变量的调用中使用此型号