You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
@CorentinJ Currently, the model natively supports first-frame conditioning.
However, it is possible to implement end-frame conditioning by using an inference-time in-painting approach, specifically conditioning on the last 8 frames. This works because the last frame in the latent space corresponds to 8 frames in the pixel space.
Looking ahead, we plan to include native support for end-frame conditioning in a future update, making this process more seamless.
_Originally posted by @Shecht-ltx in https://github.com/Lightricks/LTX-Video/issues/29#issuecomment-2527379386_
The text was updated successfully, but these errors were encountered:
Just wanted to see if anyone has figured out how to do endframes in order to create seamless loops.
Referenced here: #29
The text was updated successfully, but these errors were encountered: