Neurone/flux.1-dev-fp8Inference app for a FP8-quantized flux1-dev model. This runs on graphic cards with 16 GB of VRAM.★ 42PythonForks 6