UNO

UNO copied to clipboard

Hi, thank you for your excellent work! I’m curious about the settings on DreamBench. Do all the images sampled by the comparison methods have a resolution of 1024x1024?

Ryann-Ran

You have reimplemented the flux yourself, resulting in a complex application. Can you inherit from the flux classes in diffusers to make it easier to implement?

HZIH

Hi! Thanks a lot for sharing this awesome work! I tried running the default evaluation code, but I noticed the results are a bit different from what’s shown in the...

HHHHStar

您好，请问UNO-1M 数据集是否支持 multi-subject generation 的训练？如果不支持，是否有计划在未来支持？

zrk2

In the "crystal_ball" and "figurine" demo, the details of the clothes of "figurine" cannot be maintained in the generated image. Is it possible to solve this problem by finetuning the...

KleinXin

Can the dataset be placed in modelscope?

3

too slow to download from huggingface, can you tut the dataset in modelscope? thanks!

BITHG287

Will release in-context data generation pipelines~?

4

kelisiya

Could you please release the evaluation code?

1

The evaluation scores of the baselines and UNO is different when I use my evaluation code.

arturia-Xayah

When will you release in-context data generation pipelines？

1

Thank you for your outstanding open source work，but I would be more appreciate if you release in-context data generation pipelines.

Morganismine

support GPU machine training？

1

Thank you for your impressive work. How to support training on multiple GPU machines

CccccchenJD