diffusers

diffusers copied to clipboard

does the txt2img pipe support load a finetune .pt file to improve the output performance of checkpoint model ?

Hey @CrazyBoyM,

Sorry I don't understand the question here

sorry for my poor english. it is mean hypernetwork and text inversion, just like a option in the webui version of staable diffusion, load a single finetune file to improve the base model.

You can extend the script using the method from this notebook: https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers/stable_conceptualizer_inference.ipynb

Have a look at the "Load the newly learned embeddings into CLIP" section. text_encoder and tokenizer are properties of the pipeline object.

You can extend the script using the method from this notebook: https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers/stable_conceptualizer_inference.ipynb

Have a look at the "Load the newly learned embeddings into CLIP" section. text_encoder and tokenizer are properties of the pipeline object.

Thanks for help.I will take it a try.