axelker / local-ai-chat

Local AI Chat with Image Recognition is an advanced web application that combines real-time conversation capabilities with powerful image recognition features. Leveraging state-of-the-art technologies, this project offers a comprehensive solution for both text-based and image-based interactions in a secure and customizable local environment.

Local AI Chat

Local AI Chat with Image Recognition is an advanced web application that combines real-time conversation capabilities with powerful image recognition features. Leveraging state-of-the-art technologies, this project offers a comprehensive solution for both text-based and image-based interactions in a secure and customizable local environment.

• Using Angular 16 - Flask 2.1 - Ollama - Llama3 - LLava - LangChain.

Ollama

• Download ollama. (You can move ollama models on an external disk and update the path of env var OLLAMA_MODELS)

Start for the back

• Open terminal : ollama start && ollama run modelName (ex : llama3)
• Create local env : python -m venv envname
• Activate env : source envname/Scripts/activate
• pip install -r requirements.txt

(Deactivate env : deactivate)

Run Back

• Define .env file (Exemple with .env.sample)
• flask run

Start for the front

• npm install

Run Front

• npm start

Sample rag

This repo is an exemple to make a rag with json and pdf data. (Use wsl terminal for install requirement.txt)