This is a conversion of the Shuttle Jaguar model to NF4 only 6.5GB (needs the T5_encoder, clip, an VAE)
https://huggingface.co/shuttleai/shuttle-jaguar
fine-tuned Flux schnell.
This model get good quality images very quickly with just 4 steps (ideal for fixing details on the InvokeAI Canvas)
Thanks to the original author https://civitai.com/user/shuttleai