這是本文件的舊版!
Ollama + Open-WebUI
- 安裝環境 VM 規格 :
- vCPU : 8 ([host] i7-11700 支援 avx2)
- RAM : 8G
- SSD : 64G
- OS : Alpine 3.19 + docker compose
使用 docker compose 快速安裝測試
- 編輯 docker-compose.yml
services: ollama: image: ollama/ollama:latest ports: - 11434:11434 volumes: - .:/code - ./ollama/ollama:/root/.ollama container_name: ollama environment: - OLLAMA_NUM_PARALLEL=4 - OLLAMA_MAX_LOADED_MODELS=4 pull_policy: always tty: true restart: always networks: - ollama-docker #deploy: # resources: # reservations: # devices: # - driver: nvidia # count: 1 # capabilities: [gpu] ollama-webui: image: ghcr.io/open-webui/open-webui:main container_name: ollama-webui volumes: - ./ollama/ollama-webui:/app/backend/data depends_on: - ollama ports: - 8080:8080 environment: - '/ollama/api=http://ollama:11434/api' extra_hosts: - host.docker.internal:host-gateway restart: unless-stopped networks: - ollama-docker pipelines: image: ghcr.io/open-webui/pipelines:main container_name: pipelines ports: - "9099:9099" volumes: - ./ollama/pipelines:/app/pipelines extra_hosts: - "host.docker.internal:host-gateway" restart: always networks: - ollama-docker networks: ollama-docker: external: false
- 啟動 docker compose
docker compose up -d
- Open-WebUI 網址 - http://Server-IP:8080
- Ollama API 網址 - http://Server-IP:11434
Open WebUI 設定 Anthropic API
- 採用 pipeline 機制 - https://docs.openwebui.com/pipelines/pipes
- 設定 連線 新增 http://host.docker.internal:9099 API Key: 0p3n-w3bu! 看畫面
- 管線(pipeline) 從 GitHub URL 安裝
Model
- 下載語法 Exp. ycchen/breeze-7b-instruct-v1_0
docker exec ollama ollama pull ycchen/breeze-7b-instruct-v1_0
評估安裝環境運行 Model 的效能
- Exp. phi3
pve-ollama-221:~# docker exec ollama ollama list NAME ID SIZE MODIFIED phi3:14b 1e67dff39209 7.9 GB 5 days ago codestral:latest fcc0019dcee9 12 GB 5 days ago yabi/breeze-7b-32k-instruct-v1_0_q8_0:latest ccc26fb14a68 8.0 GB 5 days ago phi3:latest 64c1188f2485 2.4 GB 5 days ago pve-ollama-221:~# docker exec -it ollama ollama run phi3 --verbose >>> hello Hello! How can I assist you today? Whether it's answering a question, providing information, or helping with a task, feel free to ask. total duration: 3.119956759s load duration: 905.796µs prompt eval count: 5 token(s) prompt eval duration: 101.53ms prompt eval rate: 49.25 tokens/s eval count: 32 token(s) eval duration: 2.889224s eval rate: 11.08 tokens/s
VM 內直接安裝與測試 Ollama
- 安裝程序
curl -fsSL https://ollama.com/install.sh | sh
- 下載模型 Exp. llama3
ollama pull llama3
- 開始測試
ollama run llama3