Text2Image?
#11
by
GrStant
- opened
I see that in the model card, as well as in the announcement, it was mentioned that Wan2.1 has T2I capabilities. Right now though, none of the models in the Wan2.1 collection seem to explicitly target this particular combo of modalities. Can T2I be accomplished through running inference on the T2V models with some particular settings? Or is there a later T2I-specific model version to be released?
Thanks!
PS: Congratulations on the very nifty open-weights model, especially the lower VRAM requirements. That's an awesome contribution to the community!