Update README.md

README.md

@@ -14,8 +14,8 @@ tags:
 
 ## **Key Features**
 - **Transformer-Based**: Utilizes a powerful `Transformer3DModel` for processing high-dimensional spatiotemporal data.
-- **High Resolution**: Generates videos at **256x256** resolution with seamless transitions and realism.
-- **30 FPS Output**: Smooth frame generation for real-world video applications.
+- **High Resolution**: Generates videos at **768x512** resolution with seamless transitions and realism (24 FPS).
+- **24 FPS Output**: Smooth frame generation for real-world video applications.
 - **Advanced Latent Compression**: Leverages a `CausalVideoAutoencoder` for efficient latent representation and generation.
 
 ---
@@ -110,8 +110,8 @@ The `CausalVideoAutoencoder` (VAE) handles latent space compression and decompre
 | **Created By**             | `SVECTOR`                  |
 | **Parameter Count**        | `2 Billion`                            |
 | **Framework Version**      | `0.25.1` (Diffusers)                   |
-| **Resolution**             | `256x256`                              |
-| **Frame Rate**             | `30 FPS`                               |
+| **Resolution**             | `768x512`                              |
+| **Frame Rate**             | `24 FPS`                               |
 | **Transformer Type**       | `Transformer3DModel`                   |
 | **Encoder Type**           | `CausalVideoAutoencoder`               |