Instructions to use Nichonauta/Self-Forcing2.1-T2V-1.3B-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- SelfForcing
How to use Nichonauta/Self-Forcing2.1-T2V-1.3B-GGUF with SelfForcing:
# 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
- Notebooks
- Google Colab
- Kaggle
Update README.md
Browse files
README.md
CHANGED
|
@@ -11,6 +11,7 @@ tags:
|
|
| 11 |
base_model:
|
| 12 |
- gdhe17/Self-Forcing
|
| 13 |
---
|
|
|
|
| 14 |
|
| 15 |
# Self-Forcing2.1-T2V-1.3B-GGUF
|
| 16 |
|
|
@@ -20,66 +21,66 @@ base_model:
|
|
| 20 |
|
| 21 |
---
|
| 22 |
|
| 23 |
-
|
| 24 |
|
| 25 |
-
|
| 26 |
|
| 27 |
-
|
| 28 |
|
| 29 |
-
|
| 30 |
|
| 31 |
-
## ✨
|
| 32 |
|
| 33 |
-
- ⚡️ **
|
| 34 |
-
- 🧠 **
|
| 35 |
-
- 🖼️ **
|
| 36 |
-
- 📉 **
|
| 37 |
-
- 🧬 **
|
| 38 |
|
| 39 |
-
##
|
| 40 |
|
| 41 |
-
|
| 42 |
|
| 43 |
---
|
| 44 |
|
| 45 |
-
## 🧐
|
| 46 |
|
| 47 |
-
GGUF
|
| 48 |
|
| 49 |
-
- **
|
| 50 |
-
- **
|
| 51 |
-
- **
|
| 52 |
|
| 53 |
-
**
|
| 54 |
|
| 55 |
---
|
| 56 |
|
| 57 |
-
## 📚
|
| 58 |
|
| 59 |
-
|
| 60 |
|
| 61 |
-
###
|
| 62 |
|
| 63 |
-
|
| 64 |
|
| 65 |
-
- **
|
| 66 |
-
- **
|
| 67 |
|
| 68 |
-
###
|
| 69 |
|
| 70 |
-
|
| 71 |
|
| 72 |
-
- **
|
| 73 |
|
| 74 |
-----
|
| 75 |
|
| 76 |
-
## 🙏
|
| 77 |
|
| 78 |
-
|
| 79 |
|
| 80 |
-
## ✍️
|
| 81 |
|
| 82 |
-
|
| 83 |
|
| 84 |
```bibtex
|
| 85 |
@article{wan2.1,
|
|
|
|
| 11 |
base_model:
|
| 12 |
- gdhe17/Self-Forcing
|
| 13 |
---
|
| 14 |
+
```
|
| 15 |
|
| 16 |
# Self-Forcing2.1-T2V-1.3B-GGUF
|
| 17 |
|
|
|
|
| 21 |
|
| 22 |
---
|
| 23 |
|
| 24 |
+
Developed by <a href="https://www.youtube.com/@nichonauta">Nichonauta</a>.
|
| 25 |
|
| 26 |
+
This repository contains the quantized versions in **GGUF** format of the **Self-Forcing** video generation model.
|
| 27 |
|
| 28 |
+
The Self-Forcing model is an evolution of `Wan2.1-T2V-1.3B`, optimized with an innovative "self-forcing" technique that allows it to correct its own generation errors in real-time. This results in more coherent and higher-quality videos.
|
| 29 |
|
| 30 |
+
These GGUF files allow the model to be run efficiently on **GPU/CPU**, drastically reducing VRAM consumption and making video generation accessible without the need for high-end GPUs.
|
| 31 |
|
| 32 |
+
## ✨ Key Features
|
| 33 |
|
| 34 |
+
- ⚡️ **GPU/CPU Inference:** Thanks to the GGUF format, the model can run on a wide range of hardware with optimized performance.
|
| 35 |
+
- 🧠 **Self-Forcing Technique:** The model learns from its own predictions during generation to improve temporal consistency and visual quality of the video.
|
| 36 |
+
- 🖼️ **Image-guided Generation:** Ability to generate smooth video transitions between a start and an end image, guided by a text prompt.
|
| 37 |
+
- 📉 **Low Memory Consumption:** Quantization significantly reduces the RAM/VRAM memory footprint compared to the original models (`FP16`/`FP32`).
|
| 38 |
+
- 🧬 **Based on a Solid Architecture:** It inherits the powerful base of the `Wan2.1-T2V-1.3B` model, known for its efficiency and quality.
|
| 39 |
|
| 40 |
+
## Usage
|
| 41 |
|
| 42 |
+
The model files can be used in [ComfyUI](https://github.com/comfyanonymous/ComfyUI/) with the [ComfyUI-GGUF](https://github.com/city96/ComfyUI-GGUF) custom node.
|
| 43 |
|
| 44 |
---
|
| 45 |
|
| 46 |
+
## 🧐 What is GGUF?
|
| 47 |
|
| 48 |
+
GGUF is a file format designed to store large language models (and other architectures) for fast inference on CPUs. The key advantages are:
|
| 49 |
|
| 50 |
+
- **Fast Loading:** Does not require complex deserialization.
|
| 51 |
+
- **Quantization:** Allows model weights to be stored with reduced precision (e.g., 4 or 8 bits instead of 16 or 32), which reduces file size and RAM usage.
|
| 52 |
+
- **GPU/CPU Execution:** It is optimized to run on general-purpose processors through libraries like `llama.cpp`.
|
| 53 |
|
| 54 |
+
**Note:** Running this video model in GGUF format requires compatible software that can interpret the video diffusion transformer architecture.
|
| 55 |
|
| 56 |
---
|
| 57 |
|
| 58 |
+
## 📚 Model Details and Attribution
|
| 59 |
|
| 60 |
+
This work would not be possible without the open-source projects that precede it.
|
| 61 |
|
| 62 |
+
### Base Model: Wan2.1
|
| 63 |
|
| 64 |
+
This model is based on `Wan2.1-T2V-1.3B`, a powerful 1.3 billion parameter text-to-video model. It uses a Diffusion Transformer (DiT) architecture and a 3D VAE (Wan-VAE) optimized to preserve temporal information, making it ideal for video generation.
|
| 65 |
|
| 66 |
+
- **Original Repository:** [Wan-AI/Wan2.1-T2V-1.3B](https://huggingface.co/Wan-AI/Wan2.1-T2V-1.3B)
|
| 67 |
+
- **Architecture:** Diffusion Transformer (DiT) with a T5 text encoder.
|
| 68 |
|
| 69 |
+
### Optimization Technique: Self-Forcing
|
| 70 |
|
| 71 |
+
The `Wan2.1` model was enhanced with the **Self-Forcing** method, which trains the model to recognize and correct its own diffusion errors in a single forward pass. This improves fidelity and coherence without the need for costly additional training.
|
| 72 |
|
| 73 |
+
- **Project Page:** [self-forcing.github.io](https://self-forcing.github.io/)
|
| 74 |
|
| 75 |
-----
|
| 76 |
|
| 77 |
+
## 🙏 Acknowledgements
|
| 78 |
|
| 79 |
+
We thank the teams behind [Wan2.1](https://huggingface.co/Wan-AI/), [Self-Forcing](https://self-forcing.github.io/), [Stable Diffusion](https://huggingface.co/stabilityai/stable-diffusion-3-medium), [diffusers](https://github.com/huggingface/diffusers), and the entire [Hugging Face](https://huggingface.co) community for their contribution to the open-source ecosystem.
|
| 80 |
|
| 81 |
+
## ✍️ Citation
|
| 82 |
|
| 83 |
+
If you find our work useful, please cite the original projects:
|
| 84 |
|
| 85 |
```bibtex
|
| 86 |
@article{wan2.1,
|