How to generate videos based on StableDiffusion v1.5

[YBYBZhang]

Thanks for your great work! Could you provide scripts or configs to produce videos based on SD v1.5?

[kodxana]

Any model will work just make sure it's safetensors pick one config and point to your local path.

[YBYBZhang]

The weights of StableDiffusion v1.5 has been passed by --pretrained_model_path. Thus, I think the path to safetensors could be None, but got weird results :)

[actionless]

@YBYBZhang having the same issue with empty safetensors (the frames are not connected to each other)

did you figured it out?

[YBYBZhang]

@actionless Using SD v1.5 as base model still gets bad results :) I replace it with Realistic Vision.

[actionless]

@YBYBZhang not my case, as i intending to use customly-trained model

however after few experiments, i've just realized that in such setup (MotionModule+base model only) it works the best when the both are of 1.5 or both sd 1.4 version - in that case result looks adequate