Is AutoencoderKL trained along with sync multi view diffusion?

[sungmin9939]

Hi

Thanks for sharing your great work

I just wondering whether AutoencoderKL model, which is a first stage model in your model, is trained or not.
it isn't initiated from checkpoint or isn't set to be trained in your ldm/model/diffusion/sync_dreamer.py file

I look forward your response

Thanks

[liuyuan-pal]

Hi, I do not train it~