Instructions to use Lightricks/LTX-2.5-22b-IC-LoRA-Pixel-Spatial-Upscaler with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusion Single File
How to use Lightricks/LTX-2.5-22b-IC-LoRA-Pixel-Spatial-Upscaler with Diffusion Single File:
# No code snippets available yet for this library. # To use this model, check the repository files and the library's documentation. # Want to help? PRs adding snippets are welcome at: # https://github.com/huggingface/huggingface.js
- LTX-2
How to use Lightricks/LTX-2.5-22b-IC-LoRA-Pixel-Spatial-Upscaler 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 — see Files and versions) hf download Lightricks/LTX-2.5-22b-IC-LoRA-Pixel-Spatial-Upscaler --local-dir models/LTX-2.5-22b-IC-LoRA-Pixel-Spatial-Upscaler
# 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-Pixel-Spatial-Upscaler/<weights>.safetensors 1.0 \ --video-conditioning reference.mp4 1.0 \ --prompt "your prompt here" \ --output-path output.mp4 - Notebooks
- Google Colab
- Kaggle
workflow example
workflow example
You can use this V2V workflow: https://github.com/Lightricks/ComfyUI-LTXVideo/blob/master/example_workflows/2.5/LTX-2.5_V2V_ICLoRA_Single_Stage_Distilled.json
Simply load the ltx-2.5-22b-ic-lora-pixel-spatial-upscaler-x2-1.0.safetensors as the LoRA and you should be good to go
Hi, I can't get this workflow to work using that V2V. I've set the reference video and the IC Lora.
- First problem is that it needs an image input even when the 'use image input' toggle on the Input Parameters node group is set to false. Adding an arbitrary image to the load image node gets past this problem.
- The workflow then fails at the Preprocess node. I assume it is something to do with the input video resolution but I've tried a bunch of different resolutions and always the same, and I don't know how to address this:
ComfyUI Error Report
Error Details
- Node ID: 9002:5012
- Node Type: LTXAddVideoICLoRAGuide
- Exception Type: einops.EinopsError
- Exception Message: einops.EinopsError: Error while processing rearrange-reduction pattern "b c (d p1) (h p2) (w p3) -> b (c p1 p2 p3) d h w".
Input tensor shape: torch.Size([1, 1024, 2, 17, 29]). Additional info: {'p1': 2, 'p2': 2, 'p3': 2}.
Shape mismatch, can't divide axis of length 17 in chunks of 2
The only way I managed to get the workflow to run is by disconnecting the width and height inputs to the Preprocess node group and manually entering double the resolution of the reference video. This seems to work but I'm not sure if this is using the IC LoRA as intended and to its full potential. The other example video above in a previous post uses a Latent Upscale method with the IC LoRA, but again I'm not sure if that is using this IC LoRA as intended.
It would be really good to have an example worfklow for this LoRA since it is one of the most useful ones out of all of those posted.