Are there plans for GGUF versions for LTX-Video 0.9.6 Distilled and Dev?
Thank you for your great work! Thanks to your GGUF, I can achieve great results with my 4 GB VRAM, especially with LTX video.
The Safetensor checkpoint of the new Distilled version is already quite fast. I think with GGUF, if possible, the speed would be unbeatable.
Thanks again for your valuable work.
I second this!
I second this!
I have good news, because the user calcuis now provides
gguf quantized for LTX-Video version 0.9.6 dev and distilled.
If you search for it on Hugging Face, you'll find it:
calcuis/ltxv0.9.6-gguf
I would still be happy for a release from city96 🤗
Well, uploaded some in case people are still looking for it:
https://huggingface.co/city96/LTX-Video-0.9.6-dev-gguf
https://huggingface.co/city96/LTX-Video-0.9.6-distilled-gguf
I also tried to make the readme less obtuse and included an example workflow, a list of the required models, and a sample comparison between the different quants for the model. Unsure if this is useful, feedback welcome. If people think it is I can add it to other models retroactively / going forward.