Instructions to use TenStrip/LTX2.3-10Eros with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use TenStrip/LTX2.3-10Eros with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline from diffusers.utils import load_image, export_to_video # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("TenStrip/LTX2.3-10Eros", dtype=torch.bfloat16, device_map="cuda") pipe.to("cuda") prompt = "A man with short gray hair plays a red electric guitar." image = load_image( "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/guitar-man.png" ) output = pipe(image=image, prompt=prompt).frames[0] export_to_video(output, "output.mp4") - Notebooks
- Google Colab
- Kaggle
What is 1.4 version?
/bump
same lol I just was getting a handle on 1.3 because at first it gave me some weird results but got it working with some finetuning
I've done a few tests just replacing the 10Eros 1.3 checkpoint in place for the 1.4 one in the DMD v5 workflow posted by 10s a few days ago.
Same initial frame, same prompt, same settings, same seeds.
So far the 1.4 is working fine in this workflow and for the most part is showing better results than the same setup with 1.3, with somewhat less artifacts and hallucinations specially on scenes with lots of motion in the background.
Still, I've got a couple of weird scenarios related to "fluids" which for some reason were working much better on the previous 1.3 than with the new 1.4, but not sure if this is just bad luck with the chosen seeds, as the number of tests are still small.
Maybe 10s will also need to update the workflow or the DMD lora with fine tuning for v1.4
It's 1.3 brought back to being more explicit capable and set up to be tuned. You wanna use less preprocessing liek 27-30 on this one probably. I'm still trying best WF practices.