pip install requirements.txt
python main.py --dev
- Step 1: Reserve VRAM for embedding model
python reserve_mem.py
- Step 2: Run TGI
- Step 3: Go back to terminal at Step 1 and input 'q'
- Step 4: Start the application
uvicorn main:app --host 0.0.0.0 \
--port 8000 \
--workers 4 \
--root-path llama/haystack
or
python main.py