Instructions to use SimpleTuner/ltxvideo-2.3-distilled-convrot-int8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use SimpleTuner/ltxvideo-2.3-distilled-convrot-int8 with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("SimpleTuner/ltxvideo-2.3-distilled-convrot-int8", dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
Thanks
Thanks, was looking for a convrot version to try as the fp16 version is around 45GB and kinda slow. Other convrot's I have tried have been great so far in quality so expecting the same here. Thanks again.
i created this because the only one i found before was the Eros10 weights which are way overtrained on XXX
hope this one does well for you
Unfortunately I was unable to get it working in the ComfyUI I use. Get an error. May need custom node for INT8? Something about the weights.
It's not a big deal, was just looking for a smaller sized version better then the standard fp8 to run on my website as fp16 is 45GB and kinda slow to load. I have tried a few other models convrot conversion and found the quality much better then the standard fp8 conversion.
you'll need a convrot capable loader for ComfyUI.