Spaces:
Runtime error
Runtime error
Warbler CDA HuggingFace Deployment
This directory contains the Warbler CDA package prepared for HuggingFace deployment.
Quick Start
Local Testing
# Install dependencies
pip install -r requirements.txt
# Install package in development mode
pip install -e .
# Run Gradio demo
python app.py
Deploy to HuggingFace Space
Option 1: Manual Deployment
# Install HuggingFace CLI
pip install huggingface_hub
# Login
huggingface-cli login
# Upload to Space
huggingface-cli upload YOUR_USERNAME/warbler-cda . --repo-type=space
Option 2: GitLab CI/CD (Automated)
Set up HuggingFace token in GitLab CI/CD variables:
- Go to Settings > CI/CD > Variables
- Add variable
HF_TOKENwith your HuggingFace token - Add variable
HF_SPACE_NAMEwith your Space name (e.g.,username/warbler-cda)
Push to main branch or create a tag:
git tag v0.1.0 git push origin v0.1.0The pipeline will automatically sync to HuggingFace!
Package Structure
warbler-cda-package/
βββ warbler_cda/ # Main package
β βββ __init__.py
β βββ retrieval_api.py # Core RAG API
β βββ semantic_anchors.py # Semantic memory
β βββ stat7_rag_bridge.py # STAT7 hybrid scoring
β βββ embeddings/ # Embedding providers
β βββ api/ # FastAPI service
β βββ utils/ # Utilities
βββ app.py # Gradio demo for HF Space
βββ requirements.txt # Dependencies
βββ pyproject.toml # Package metadata
βββ README.md # Documentation
βββ LICENSE # MIT License
Features
- Semantic Search: Natural language document retrieval
- STAT7 Addressing: 7-dimensional multi-modal scoring
- Hybrid Scoring: Combines semantic + STAT7 for superior results
- Production API: FastAPI service with concurrent query support
- CLI Tools: Command-line interface for management
- HF Integration: Direct dataset ingestion
Testing
# Run tests
pytest
# Run specific experiments
python -m warbler_cda.stat7_experiments
Documentation
See README.md for full documentation.