Update model card with paper, code links and correct pipeline tag
#1
by
nielsr
HF Staff
- opened
README.md
CHANGED
|
@@ -1,26 +1,31 @@
|
|
| 1 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
license: other
|
| 3 |
license_name: ltx-2-community-license
|
| 4 |
license_link: https://www.github.com/Lightricks/LTX-2/LICENSE
|
| 5 |
-
|
| 6 |
tags:
|
| 7 |
- ltx-video
|
| 8 |
- image-to-video
|
| 9 |
- text-to-video
|
| 10 |
pinned: true
|
| 11 |
-
language:
|
| 12 |
-
- en
|
| 13 |
-
pipeline_tag: text-to-video
|
| 14 |
-
datasets:
|
| 15 |
-
- Lightricks/Canny-Control-Dataset
|
| 16 |
-
base_model:
|
| 17 |
-
- Lightricks/LTX-2
|
| 18 |
---
|
| 19 |
|
| 20 |
# LTX-2 19B IC-LoRA Canny Control
|
| 21 |
|
| 22 |
This is a Canny control IC-LoRA trained on top of **LTX-2-19b**, enabling structure-preserving video generation from text and reference frames.
|
| 23 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 24 |
## What is In-Context LoRA (IC LoRA)?
|
| 25 |
|
| 26 |
IC LoRA enables conditioning video generation on reference video frames at inference time, allowing fine-grained video-to-video control on top of a text-to-video, base model.
|
|
@@ -42,11 +47,22 @@ See the **LTX-2-community-license** for full terms.
|
|
| 42 |
|
| 43 |
### 🔌 Using in ComfyUI
|
| 44 |
1. Copy the LoRA weights into `models/loras`.
|
| 45 |
-
2. Use the official IC-LoRA workflow from the LTX-2 ComfyUI repository.
|
| 46 |
|
| 47 |
## Dataset
|
| 48 |
|
| 49 |
-
https://huggingface.co/datasets/Lightricks/Canny-Control-Dataset/
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 50 |
|
| 51 |
## Acknowledgments
|
| 52 |
|
|
|
|
| 1 |
---
|
| 2 |
+
base_model:
|
| 3 |
+
- Lightricks/LTX-2
|
| 4 |
+
datasets:
|
| 5 |
+
- Lightricks/Canny-Control-Dataset
|
| 6 |
+
language:
|
| 7 |
+
- en
|
| 8 |
license: other
|
| 9 |
license_name: ltx-2-community-license
|
| 10 |
license_link: https://www.github.com/Lightricks/LTX-2/LICENSE
|
| 11 |
+
pipeline_tag: any-to-any
|
| 12 |
tags:
|
| 13 |
- ltx-video
|
| 14 |
- image-to-video
|
| 15 |
- text-to-video
|
| 16 |
pinned: true
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
---
|
| 18 |
|
| 19 |
# LTX-2 19B IC-LoRA Canny Control
|
| 20 |
|
| 21 |
This is a Canny control IC-LoRA trained on top of **LTX-2-19b**, enabling structure-preserving video generation from text and reference frames.
|
| 22 |
|
| 23 |
+
It is based on the [LTX-2](https://huggingface.co/papers/2601.03233) foundation model.
|
| 24 |
+
|
| 25 |
+
- **Paper:** [LTX-2: Efficient Joint Audio-Visual Foundation Model](https://huggingface.co/papers/2601.03233)
|
| 26 |
+
- **Code:** [GitHub Repository](https://github.com/Lightricks/LTX-2)
|
| 27 |
+
- **Project Page:** [LTX-2 Playground](https://app.ltx.studio/ltx-2-playground/i2v)
|
| 28 |
+
|
| 29 |
## What is In-Context LoRA (IC LoRA)?
|
| 30 |
|
| 31 |
IC LoRA enables conditioning video generation on reference video frames at inference time, allowing fine-grained video-to-video control on top of a text-to-video, base model.
|
|
|
|
| 47 |
|
| 48 |
### 🔌 Using in ComfyUI
|
| 49 |
1. Copy the LoRA weights into `models/loras`.
|
| 50 |
+
2. Use the official IC-LoRA workflow from the [LTX-2 ComfyUI repository](https://github.com/Lightricks/ComfyUI-LTXVideo/).
|
| 51 |
|
| 52 |
## Dataset
|
| 53 |
|
| 54 |
+
The model was trained using the [Lightricks/Canny-Control-Dataset](https://huggingface.co/datasets/Lightricks/Canny-Control-Dataset/).
|
| 55 |
+
|
| 56 |
+
## Citation
|
| 57 |
+
|
| 58 |
+
```bibtex
|
| 59 |
+
@article{hacohen2025ltx2,
|
| 60 |
+
title={LTX-2: Efficient Joint Audio-Visual Foundation Model},
|
| 61 |
+
author={HaCohen, Yoav and Brazowski, Benny and Chiprut, Nisan and Bitterman, Yaki and Kvochko, Andrew and Berkowitz, Avishai and Shalem, Daniel and Lifschitz, Daphna and Moshe, Dudu and Porat, Eitan and others},
|
| 62 |
+
journal={arXiv preprint arXiv:2601.03233},
|
| 63 |
+
year={2025}
|
| 64 |
+
}
|
| 65 |
+
```
|
| 66 |
|
| 67 |
## Acknowledgments
|
| 68 |
|