> Diffusers, ComfyUI, vLLM-Omni, SGLang, and LightX2V
replyI think that's all Python (not a direct executable).
You could just do (see the "Quick Start") four `pip install` and have a dozen lines script to generate the image. But `llama.cpp` and similar do not require e.g. installing Torch (or PyTorch) - you can use `llama.cpp` on a non-specialized machine.
"just"
replyI don't think I have ever once run "pip install transformers" and had it work without three rounds of fiddling
Or trying to install the whole of CUDA on machines that do not even have a GPU (not Nvidia, not anything past the embedded)...
replyYep, that's (also) what I meant ;)
Lean, efficient... Also sensible and trouble-less.
The 'not nvidia' is no longer a deal breaker by itself.
reply