Instructions to use Lightricks/LTX-2.5-22b-IC-LoRA-Layout-To-Render with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- LTX-2
How to use Lightricks/LTX-2.5-22b-IC-LoRA-Layout-To-Render with LTX-2:
# Install the LTX-2 pipelines git clone https://github.com/Lightricks/LTX-2.git cd LTX-2 uv sync --extra natten
# Download the adapter weights from this repo # (base components come from Lightricks/LTX-2.5 β see Files and versions) hf download Lightricks/LTX-2.5-22b-IC-LoRA-Layout-To-Render --local-dir models/LTX-2.5-22b-IC-LoRA-Layout-To-Render
# Video-to-video with the IC-LoRA (runs on the distilled LTX-2.5 base) uv run python -m ltx_pipelines.ic_lora \ --transformer-path path/to/distilled-transformer.safetensors \ --text-encoder-path path/to/gemma4-12b-with-proj-ltx-2.5-bf16.safetensors \ --video-vae-path path/to/video-vae.safetensors \ --audio-vae-path path/to/audio-vae.safetensors \ --spatial-upsampler-path path/to/spatial-upsampler.safetensors \ --lora models/LTX-2.5-22b-IC-LoRA-Layout-To-Render/<weights>.safetensors 1.0 \ --video-conditioning reference.mp4 1.0 \ --prompt "your prompt here" \ --output-path output.mp4 - Notebooks
- Google Colab
- Kaggle
Reference image at -1?
Hello guys,
After fiddling a bit with the supplied workflow, I couldn't make it work.
What I did is take the first frame of the playblast and transformed into the final render I wanted for the rest of the video.
I supply that frame as reference, but that frame only starts to appear near end of my sequence (120 frames long).
Checked the nodes and I can't understand why the reference image only gets applied as a ic-lora in the frame -1 (the last frame).
Is that correct? Maybe I am doing something wrong?
Thank you!
How do you add extra keyframes? This isn't explained anywhere. Do I need to have the keyframes rendered out as a video file like the layout, or is there a way to chain multiple images together?
eric-venti-seeds, got same problem.
If ref image matches the first frame of the guide footage result is really bad. If it matches the last frame of the guide footage, then the result is better, but only about half of the generated footage looks decent. Still the output is unusable at all.