TypeError: StableDiffusionPipeline.__init__() missing 1 required positional argument: 'safety_checker'
I'm trying to compile your models myself: "BloodborneDiffusion" and "DarkSoulsDiffusion".
I just don't want to limit them to float16.
And even thou is states at the begging:{'safety_checker'} was not found in config. Values will be initialized to default values.{'timestep_values'} was not found in config. Values will be initialized to default values.
If finally states anyway thatTypeError: StableDiffusionPipeline.__init__() missing 1 required positional argument: 'safety_checker'
This happens for both models i mentioned above. So just making sure, there were no safety_checker sub-directory with file named pytorch_model.bin when you crafted those amazing models?
I'm on pytorch1.13.
I've got some bad news, and some good news, but first thanks for the compliment, I try my best and it's appreciated :)
- I don't have the diffusers of this model anymore. I thought it was a little too trained to continue training and I needed space. So I cannot check
- I just pushed an update to the diffusers by converting the ckpt back, using this space. I see the files you wanted have been added
I pushed the same things on the Dark Souls model. (and a bigger model for more soulsborne may be coming soon, it's not easy to get all of them in one while retaining quality)