Implementing Dreambooth weights

[TijmenJansen]

Is it possible to use a trained dreambooth model into cross attention control? I trained a model in Dreambooth-Stable-Diffusion on a new car and I have an image where I want to change the car to the one I trained in Dreambooth. Changing 'model_path_diffusion' to the downloaded weights of Dreambooth does not seem to work, it does not generate the new car but something totaly different.

[bloc97]

I am not sure about other deambooth implementations, but if you trained the model using the diffusers library, it should work fine. https://github.com/huggingface/diffusers/tree/main/examples/dreambooth

You can look a bit in the code, maybe it is possible to load the weights from another implementation...

[TijmenJansen]

Thanks for your reply. I am using ShivamShrirao's Dreambooth on colab (https://github.com/ShivamShrirao/diffusers/tree/main/examples/dreambooth). When downloading the model with AUTOMATIC1111 the model is converted via 'convert_diffusers_to_original_stable_diffusion.py'. Could this be the reason that, when loading the model into Cross Attention Control, it does not generate the trained object?

Does it then help to transform this model back with convert_original_stable_diffusion_to_diffusers.py ?