black-forest-labs

/

FLUX.1-schnell

like 2.3k

Hi,

I've find the parameters admitted by Hugging Face API with the 'Black-forest-labs/FLUX.1-schnell' model and I've those:

• Parameters:
  input: Natural language description of the image you want to generate.
  image_size: The size of the generated image, usually in the format "width x height" (e.g., "512x512").
  num_inference_steps: Determines the quality of the generated image; a higher value generally produces a higher quality image.
  seed: A number that allows for the reproducibility of the generated images.
  num_images: The number of images you want to generate.
  enable_safety_checker: Whether to enable or disable the content safety checker (True/False).

Someone have tested them? I don't have a good CPU to use DEV in my computer so I access to the SCHNELL model via API with an script I've created, but when I try to use some of those parameters I'm obtaining always the same image if I don't modify one of them (maybe due the parameters aren't correct).

Could someone test or publish if are them correct?

Thanks in advance.

Best regards,

Xavi