| # Manual HuggingFace Space Deployment | |
| Since automated deployment requires HuggingFace CLI authentication, please deploy manually: | |
| ## Option 1: Web UI (Fastest - 5 minutes) | |
| 1. Go to https://huggingface.co/new-space | |
| 2. Configure: | |
| - **Owner**: Alextorelli | |
| - **Space name**: `code-chef-modelops-trainer` | |
| - **License**: MIT | |
| - **Select the SDK**: Gradio | |
| - **Hardware**: CPU basic (free) - upgrade to t4-small later | |
| - **Visibility**: Private | |
| 3. Click **Create Space** | |
| 4. Upload these 7 files from `deploy/huggingface-spaces/modelops-trainer/`: | |
| - `app.py` | |
| - `requirements.txt` | |
| - `README.md` | |
| - `DEPLOYMENT.md` | |
| - `QUICKSTART.md` | |
| - `client_example.py` | |
| - `.spaceignore` | |
| 5. Go to **Settings → Variables and secrets** | |
| 6. Add secret: | |
| - **Name**: `HF_TOKEN` | |
| - **Value**: Your HuggingFace token from https://huggingface.co/settings/tokens | |
| 7. Wait 2-3 minutes for build | |
| 8. Test: Visit https://alextorelli-code-chef-modelops-trainer.hf.space/health | |
| ## Option 2: CLI (After `huggingface-cli login`) | |
| ```powershell | |
| # Login first | |
| huggingface-cli login | |
| # Then run deployment script | |
| cd D:\APPS\code-chef | |
| python deploy\huggingface-spaces\modelops-trainer\deploy_space.py | |
| ``` | |
| ## After Deployment | |
| Update `config/env/.env` (create if needed): | |
| ```env | |
| MODELOPS_SPACE_URL=https://alextorelli-code-chef-modelops-trainer.hf.space | |
| HUGGINGFACE_TOKEN=your_token_here | |
| ``` | |
| ## Space URL | |
| Once deployed, your Space will be at: | |
| - **Web UI**: https://huggingface.co/spaces/Alextorelli/code-chef-modelops-trainer | |
| - **API Endpoint**: https://alextorelli-code-chef-modelops-trainer.hf.space | |
| - **Health Check**: https://alextorelli-code-chef-modelops-trainer.hf.space/health | |
| --- | |
| **Proceeding with Phase 1 implementation in parallel...** | |