tsi-org/LLaVA

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py (#3)

Update recommended configurations

Remove model preloading

Load 13B model with 8-bit/4-bit quantization to support more hardwares (#2)

fix: start worker proc

docs: add notifier for gpu only inference

feat: Add LLaVA model

initial commit