lmpeng12
lmpeng12
> SDXL 的生态远没有 SD 完善，对应的 AnimateDiff 模块效果很差，所以我们暂时没有基于 SDXL 的动漫风格视频渲染方案，建议您使用 SD。 好的，我这边尝试混合ip-adapter,还有animatediff的xl版本，权重设为0.2，还是画面很脏，而且尝试了HotShot-XL，加载hsxl_tenporal_layers.f16.safetensors，也会变成全是噪声图。https://github.com/hotshotco/Hotshot-XL-Automatic1111
Thank you, I am looking forward to this work.
> > Hi @JZArray, can you share the training code and sample dataset that needs for training > > Sry, I can't upload my codes and datasets... Hi,May i ask...
python3.11?env python3.10 is not exist. Or you can try run " unset LD_LIBRARY_PATH" in the terminal.