Allegro Quantized: Text-to-Video Model Now Runs on 8GB VRAM!

[@taobabay]

Nice as usual, keep going friend ❤

[@StableAIHub]

Thank you. If you choose to try please do post what VRAM and processing time.

[@Crisisdarkness]

You have a new subscriber, I'm glad to know this, I would like to know, which of these two tools do you consider to be more efficient and achieves better results? Allegro or Pyramid Flow?

[@StableAIHub]

Thank you. Just wait for another day or 2, new t2V tool is working on low VRAM. I am preparing guide.

[@Crisisdarkness]

@@StableAIHub Thanks friend, I will be attentive to that, since I'm interested in these tools requiring few resources, I have an RTX with only 6GB of ram, but I have been able to do things with some of these tools, at the moment with LivePortrait I have been able to make videos (at 12 fps), but generating videos in another way I have not tried it yet

[@StableAIHub]

@@Crisisdarkness LTX-Video made changes to make it work on 6 GB too provided you have atleast 16 Gb shared available. Checkout and let me know if it works for you

[@Crisisdarkness]

@@StableAIHub There is one thing that I don't understand, when you refer to having 16 GB shared, how can I do that? In the "performance" tab, I see shared GPU memory usage of 11 GB, could I increase it in some way?

[@StableAIHub]

@@Crisisdarkness If it is working than fine. That means enough memory is available to load the models. To increase shared memory your motherboard should support it. Search in google and see if you can find.