yugeng-amd commited on
Commit
e27e05d
·
verified ·
1 Parent(s): dc39c21

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +5 -5
README.md CHANGED
@@ -1,14 +1,14 @@
1
  ---
2
  license: gpl-3.0
3
  ---
4
- In this work, we introduce Micro-World, an action-controlled interactive world model designed to generate high-quality, open-domain scenes. Built on top of the Wan2.1 family of models, we train both image-to-video (I2V) and text-to-video (T2V) variants to support a wide range of use cases. To foster open research and practical adoption in the community, we release the model weights, full training and inference code, as well as a curated dataset specifically tailored for controllable world modeling.
5
 
6
  For action injection, we favor adaLN for its lightweight parameter footprint, and ControlNet for its strong empirical stability during training.
7
 
8
- Note that released I2V model is trained using AdaLN architecture.
9
 
10
- More info please refer to code.
11
  <div style="margin: 0; padding: 0; text-align: center;">
12
- <img src="https://github.com/user-attachments/assets/680b87ac-0c95-4a27-b4fd-fcafb9fdf609" alt="t2v architecture" title="t2v architecture" class="t2v architecture">
13
- <img src="https://github.com/user-attachments/assets/c9cd8d9e-9555-42d3-b884-04705d1e329c" alt="t2v architecture" title="t2v architecture" class="t2v architecture">
14
  </div>
 
1
  ---
2
  license: gpl-3.0
3
  ---
4
+ In this work, we introduce Micro-World, an action-controlled interactive world model designed to generate high-quality, open-domain scenes. Built on top of the Wan2.1 family of models, we train both image-to-world (I2W) and text-to-world (T2W) variants to support a wide range of use cases. To foster open research and practical adoption in the community, we release the model weights, full training and inference code, as well as a curated dataset specifically tailored for controllable world modeling.
5
 
6
  For action injection, we favor adaLN for its lightweight parameter footprint, and ControlNet for its strong empirical stability during training.
7
 
8
+ Note that released I2W model is trained using AdaLN architecture.
9
 
10
+ More info please refer to [code](https://github.com/AMD-AGI/Micro-World).
11
  <div style="margin: 0; padding: 0; text-align: center;">
12
+ <img src="https://github.com/user-attachments/assets/680b87ac-0c95-4a27-b4fd-fcafb9fdf609" alt="model architecture" title="model architecture" class="model architecture">
13
+ <img src="https://github.com/user-attachments/assets/c9cd8d9e-9555-42d3-b884-04705d1e329c" alt="model architecture" title="model architecture" class="model architecture">
14
  </div>