ComfyUI custom node

#7
by drbaph - opened

support sgLang speed?

support sgLang speed?

The node runs Fish Speech locally inside comfyui using PyTorch directly not through a serving framework like SGLang or vLLM.
For speed, it supports flash_attention, sage_attention, and torch.compile as well as some scaled quants. average throughtput is 11 it/s on a 24gb consumer gpu

Sign up or log in to comment