README.md

---
license: apache-2.0
title: Model
sdk: docker
colorFrom: green
colorTo: green
pinned: true
---
# DeepCoder Docker Deployment

Complete Docker setup for deploying the DeepCoder-14B AI code generation model.

## Quick Start

1. **Setup and Deploy:**
   \`\`\`bash
   chmod +x setup.sh
   ./setup.sh
   \`\`\`

2. **Test the API:**
   \`\`\`bash
   curl -X POST http://localhost:8000/generate \
     -H 'Content-Type: application/json' \
     -d '{"prompt": "def fibonacci(n):", "max_tokens": 200}'
   \`\`\`

## Deployment Options

### Local Docker
- Run `./setup.sh` for automatic setup
- Supports both GPU and CPU deployment
- Includes Nginx reverse proxy with rate limiting

### Hugging Face Spaces
- Run `./deploy-hf.sh [space-name] [username]`
- Requires `HF_TOKEN` environment variable
- Automatically configures for HF Spaces (port 7860)

## API Endpoints

- `POST /generate` - Generate code from prompts
- `POST /chat` - Chat-style code assistance
- `GET /model/info` - Model benchmarks and info
- `GET /health` - Health check

## Requirements

- Docker & Docker Compose
- 16GB+ RAM (32GB recommended)
- NVIDIA GPU with 8GB+ VRAM (optional, falls back to CPU)
- 50GB+ disk space for model cache