[AI] ollama download & run new LLM models + openweb ui + public access with ngrok

Summary

• https://ollama.com/library

安装过程

brew install ollama

# https://github.com/ollama/ollama?tab=readme-ov-file
# service at localhost:11434
ollama serve # 如果是GUI(downloaded from ollama.com),则无需
# in another shell

# https://ollama.com/library
ollama run llama3
ollama run gemma:7b
ollama run mistral
ollama run mixtral
ollama run llama3.2:1b


ollama list
ollama rm llama2


curl http://localhost:11434/api/chat -d '{
  "model": "mistral",
  "messages": [
    { "role": "user", "content": "why is the sky blue?" }
  ]
}'




Customize a prompt
Models from the Ollama library can be customized with a prompt. For example, to customize the llama2 model:

ollama pull llama2
Create a Modelfile:

FROM llama2

# set the temperature to 1 [higher is more creative, lower is more coherent]
PARAMETER temperature 1

# set the system message
SYSTEM """
You are Mario from Super Mario Bros. Answer as Mario, the assistant, only.
"""
Next, create and run the model:

ollama create mario -f ./Modelfile
ollama run mario
>>> hi
Hello! It's your friend Mario.

use docker

• https://hub.docker.com/r/ollama/ollama

cd ~/_my/cloudsync/sh88qh_h1/SynologyDrive/__qh_tmp/data/ollama
mkdir ollama

cat > docker-compose.yml <<EOF
services:
  ollama:
    image: ollama/ollama
    container_name: ollama
    ports:
      - "11434:11434"
    volumes:
      - ./ollama:/root/.ollama
    restart: always
networks:
  default:
    name: npm_nginx_proxy_manager-network
    external: true
EOF


docker exec -it ollama ollama run llama3.1:1b
docker exec -it ollama ollama run llama3.3
docker exec -it ollama ollama run mistral

GUI - ‣

brew install --cask ollamac

GUI - ‣

docker run -p 3000:3000 ghcr.io/ivanfioravanti/chatbot-ollama:main

# in browser, go to
http://127.0.0.1:3000

GUI ‣

241214 Open WebUI ‣

cd ~/_my/cloudsync/sh88qh_h1/SynologyDrive/__qh_tmp/data/open-webui
mkdir open-webui

# seperate & independent webui
# Ollama is on your computer as native app, use this command
cat > docker-compose.yml <<EOF
services:
  open-webui:
    # main branch only has ui
    image: ghcr.io/open-webui/open-webui:main
    container_name: open-webui
    restart: always
    ports:
      - "3000:8080"
    extra_hosts:
      - "host.docker.internal:host-gateway"
    # if Ollama at different host
    #environment:
    #  - OLLAMA_BASE_URL=https://example.com
    #  - OLLAMA_BASE_URL=http://ollama:11434
    volumes:
      - ./open-webui:/app/backend/data
networks:
  default:
    name: npm_nginx_proxy_manager-network
    external: true
EOF


# bundle 2in1
# a built-in, hassle-free installation of both Open WebUI and Ollama, ensuring that you can get everything up and running swiftly
mkdir ollama
cat > docker-compose.yml <<EOF
services:
  open-webui:
    # ollama branch has both
    image: ghcr.io/open-webui/open-webui:ollama
    container_name: open-webui
    restart: always
    ports:
      - "3000:8080"
    volumes:
      - ./ollama:/root/.ollama
      - ./open-webui:/app/backend/data
networks:
  default:
    name: npm_nginx_proxy_manager-network
    external: true
EOF

public access

ngrok http 11434 --host-header="localhost:11434"