Full-Stack RAG Starter

One-liner: FastAPI + React RAG app with PDF ingest, SentenceTransformers, FAISS/Chroma, JWT auth.

Features

• Chunking + embeddings
• JWT auth
• Docker compose

Quickstart

python -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt || pip install -e .
python scripts/demo.py  # minimal demo

# Docker (optional)
docker build -t llm-fullstack-rag-starter -f docker/Dockerfile .
docker run --gpus all -p 8000:8000 llm-fullstack-rag-starter

Benchmarks

• Hardware:
• Dataset:
• Results:
  • Throughput: -
  • Speedup vs baseline: -
  • Accuracy/F1/Metric: -

Layout

llm-fullstack-rag-starter/
  docs/
  benchmarks/
  docker/
  scripts/
  src/
  tests/
  .github/workflows/ci.yml
  requirements.txt
  LICENSE

Notes

• Reproducible env (requirements + fixed seeds)
• Logging: CSV / TensorBoard in benchmarks/
• MIT License