LTX-Video is missing

[nitinmukesh]

Thank you for sharing, the output looks amazing.

1. In the intro video you have shown LTX-Video is supported but no config is present. Please add the same for us to try.

2. Also can it be used for 1.3B models, it will help low VRAM users.

https://huggingface.co/Wan-AI/Wan2.1-T2V-1.3B-Diffusers

Wan VACE diffusers implementation
huggingface/diffusers#11582
https://huggingface.co/Wan-AI/Wan2.1-VACE-1.3B-diffusers

[choi403]

Hi @nitinmukesh, thank you so much for your interest in our work.

First off, LTX-Video code is being tidied up and will be up soon (expect it to be up in 24-48hrs).
As for the Wan 2.1 1.3B checkpoints, we haven't implemented or tested for the 1.3B VACE or T2V models yet, but I'll look into it and try to release the ALG implementations for those checkpoints (especially 1.3B VACE, as it might fit more naturally for our I2V pipeline due to similarities) as soon as possible.

We plan to frequently update our repository, so please stay tuned!

[nitinmukesh]

Thank you for the update. VACE will be useful addition.

Can't wait to try

[Mikerhinos]

LTX-Video is my favourite model as it's super fast and fits easily in 16GB VRAM.

Would it be complicated to make a ComfyUI Custom Node from the pipelines ?

[choi403]

Hi @Mikerhinos, thank you for the great suggestion.

I think ComfyUI integration would be very helpful for users to try ALG.
I will work on ComfyUI implementation in this repo, and release when it is done via all our channels (this GitHub repo, my X account, and our project page).

Community contribution (PR, etc.) is always welcome 😃

Please stay tuned!