kudzueye

/

boreal-flux-dev-v2

like 140

I think you are loading the lora as the main model rather than separate lora weights. Try something like this instead:

pipe = FluxControlPipeline.from_pretrained("black-forest-labs/FLUX.1-dev", torch_dtype=torch.bfloat16).to("cuda")
pipe.load_lora_weights("kudzueye/boreal-flux-dev-v2")

You may also need to request access to black forest labs flux dev model on their hugging face page by the way if you had not done so.

Yes, that's worked but I am using a mac and CUDA did not support OSX how can I use it without CUDA can you help with this?
sorry for bothering.

You can probably get it to work without cuda but I am not familiar with setting things up for OSX. You might want to search around other guides on how to work with OSX and diffusers and what not.

Ok, got it.
Thanks a lot