showlab / Tune-A-Video

[ICCV 2023] Tune-A-Video: One-Shot Tuning of Image Diffusion Models for Text-to-Video Generation
https://tuneavideo.github.io
Apache License 2.0
4.15k stars 377 forks source link

AttributeError: 'AttentionBlock' object has no attribute 'to_k' #92

Open saran-eizen opened 4 months ago

saran-eizen commented 4 months ago

I have a trained stable diffusion model using the DreamBooth technique, but when I have saved those weights and trying to use them as base weights I am getting the above error.