Closed trouble-maker007 closed 1 year ago
Yes. As mentioned in the paper, we will release the MultiGen-20M dataset to contribute to the community very soon. Please make a star of our repo and keep an eye on it.
The data is released. Thanks for your interest in our work.
@shugerdou Would you like to introduce how to generate the training data?
@trouble-maker007 we follow https://github.com/lllyasviel/ControlNet/tree/main/annotator to extract conditions. The raw data is from https://huggingface.co/datasets/ChristophSchuhmann/improved_aesthetics_6plus
Thank you for open-sourcing such a great work. I would like to know if the MultiGen-20M dataset will be released to the public in the future