Ignore i2v on ltx is training on images

This commit is contained in:
Jaret Burkett
2026-01-14 18:46:27 -07:00
parent 9848de7946
commit e40d7ac605

View File

@@ -751,7 +751,7 @@ class LTX2Model(BaseModel):
video_timestep = timestep.clone()
# i2v from first frame
if batch.dataset_config.do_i2v:
if batch.dataset_config.do_i2v and batch.dataset_config.num_frames > 1:
# check to see if we had it cached
if batch.first_frame_latents is not None:
init_latents = batch.first_frame_latents.to(