Install Librechat Locally with Ollama to Chat and RAG with Any Model

 This video shows how to locally install LibreChat and integrate it with Ollama.  You can use Llama 3 , Mistral or any model to chat privately and locally with web GUI. 

Code:

conda create -n librechat python=3.11

conda activate librechat

git clone https://github.com/danny-avila/LibreChat.git

cd LibreChat

cp .env.example .env

sudo systemctl stop ollama.service

docker compose up -d

docker compose ps --format "{{.Service}} {{.State}}"

 

http://localhost:3080/

================

librechat.yaml

    - name: "Ollama"

      apiKey: "ollama"

      # use 'host.docker.internal' instead of localhost if running LibreChat in a docker container

      baseURL: "http://host.docker.internal:11434/v1/"

      models:

        default: [

          "mistral"

          ]

      # fetching list of models is supported but the `name` field must start

      # with `ollama` (case-insensitive), as it does in this example.

        fetch: false

      titleConvo: true

      titleModel: "mistral"

      summarize: false

      summaryModel: "mistral"

      forcePrompt: false

      modelDisplayLabel: "Ollama"

==================

docker-compose-Override.yaml file

services:

# # USE LIBRECHAT CONFIG FILE

 api:

   volumes:

   - type: bind

     source: ./librechat.yaml

     target: /app/librechat.yaml

# # ADD OLLAMA

 ollama:

   image: ollama/ollama:latest

   deploy:

     resources:

       reservations:

         devices:

           - driver: nvidia

             capabilities: [compute, utility]

   ports:

     - "11434:11434"

   volumes:

     - ./ollama:/root/.ollama