When I try to generate an image by this guide: Text to Image when I use negative_prompt or target_size, I have an error in output. Can anyone help or know why this error occurs?
The error message is: call() got an unexpected keyword argument ‘negative_prompt’
Thanks!
I think that black-forest-labs/FLUX.1-dev is not compatible with negative_prompt (at least via the Inference API).
It maybe work without negative_prompt.
The endpoint works fine without these 2 parameters, but I tried more than 10 models, and each time I use these 2 parameters I have an error. I am not sure that they work at all.
For example, negative_prompt should work with the SDXL and SD1.5 models. If it doesn’t work with these, it’s probably a bug.