README.md

---
title: Sheikh LLM Studio
emoji: 🚀
colorFrom: blue
colorTo: purple
sdk: docker
pinned: false
---

# Sheikh LLM Studio

Full-featured FastAPI application deployed on Hugging Face Spaces with multi-model chat, tooling, and model-creation workflows.

## Features
- Web UI with real-time chat experience and adjustable generation settings
- Multi-model support backed by Hugging Face gated models via `InferenceClient`
- Tool integration endpoints for search and code execution prototypes
- Model Studio workflow to queue fine-tuning jobs and monitor status
- WebSocket endpoint for streaming-style interactions

## Configuration
1. Add an `HF_TOKEN` repository secret in your Space with access to the desired gated models.
2. Optional: adjust available models in `app.py` under `Config.AVAILABLE_MODELS`.

## Development
```bash
git clone git@hf.co:spaces/RecentCoders/sheikh-llm
cd sheikh-llm
python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
uvicorn app:app --reload --port 7860
```

## Deployment
```bash
./deploy.sh
```

After pushing, monitor the build logs on your Space and test the endpoints:
- `https://recentcoders-sheikh-llm.hf.space/`
- `https://recentcoders-sheikh-llm.hf.space/chat`
- `https://recentcoders-sheikh-llm.hf.space/docs`
- `https://recentcoders-sheikh-llm.hf.space/health`