# CodeLlama Copilot (Web AI Coding Assistant)
Conversational code assistant powered by Meta's CodeLlama-Instruct.

## Features
- Write, fix & explain code interactively
- Multi-round chat (remembers context)
- Runs entirely in your cloud workspace (private!)

---

## How to deploy on Hugging Face Spaces (FOR FREE)
Because HF provides **free hosting & free CPUs** (much better with Pro account + GPU though).

### Step-by-step:
1. **Sign up** or login at [huggingface.co](https://huggingface.co)
2. Go to: https://huggingface.co/spaces
3. Click **New Space**
4. Name it, select **Gradio** as SDK
5. Set it **Public** *(private needs a paid plan)*
6. Once created, open the Space repo UI
7. Upload **app.py** and **requirements.txt** exactly as above.
8. Save and **click "Commit"**
9. Wait ~1–2 mins: it will build & launch!
     URL formed as:
     `https://huggingface.co/spaces/<your-username>/<space-name>`
10. Use it live — share link with friends.

---

## Notes
- **Free Spaces use CPU, so can be slow / timeout for big models.**
- Upgrade to **Hugging Face Pro** ~$9/mo
    → Choose A10G GPU in Space hardware settings for FULL performance.
- Absolutely **no server hassles**. Just upload files!

---

## (Optional) Run locally:
```bash
pip install -r requirements.txt
python app.py