How do I run the model?

#3
by FrentescuCezar - opened

As the title says, I want to run the model. I cloned the repository, have the files in my computer, and there isn't any tutorial to run the pytorch_model.bin

I'm curious as well, wanting to integrate into automatic1111

It requires use of Hugging Faces's transformers library:
https://pypi.org/project/transformers/
https://huggingface.co./docs/transformers/index

You can use/reference my version based on Gustavosta's spaces app.py here:
https://github.com/RassilonSleeps/MagicPrompt-SD

My repo includes instructions for running locally or on Google Colab. It also contains Gustavosta's model, so it can be run offline after grabbing dependencies.

@suduma If you just want it in A1111's web-ui, I happened across this thread:
https://www.reddit.com/r/stablediffusion/comments/xvjm84/_/

Recent comments indicate the "--disable-safe-unpickle" launch args are needed to allow the use of the gpt2 model. Haven't tried the script myself yet.

Thank you sir, you are awesome!

Sign up or log in to comment