Wan2.1-Fun-1.3B-InP
| Property | Value |
|---|---|
| Model Size | 19.0 GB |
| License | Apache License 2.0 |
| Author | Alibaba PAI |
| Framework | PyTorch |
What is Wan2.1-Fun-1.3B-InP?
Wan2.1-Fun-1.3B-InP is a sophisticated text-to-video generation model developed by Alibaba PAI. It represents a lightweight variant in the Wan2.1 family, specifically designed for multi-resolution video generation with support for both start and end frame prediction.
Implementation Details
The model is trained on multiple resolutions (512, 768, 1024) and specializes in frame-to-frame video generation with precise control over the starting and ending frames. It requires approximately 60GB of disk space for deployment and can operate under various GPU memory optimization modes to accommodate different hardware configurations.
- Supports multiple resolution training (512, 768, 1024)
- Implements efficient GPU memory management options
- Enables both text-to-video and image-to-video generation
- Compatible with both Windows and Linux environments
Core Capabilities
- Text-to-video generation with precise control
- Multi-resolution video output support
- Start and end frame prediction
- Flexible deployment options with memory optimization
Frequently Asked Questions
Q: What makes this model unique?
This model stands out for its ability to generate videos at multiple resolutions while maintaining control over both start and end frames, all within a relatively compact 1.3B parameter architecture. It offers an excellent balance between performance and resource requirements.
Q: What are the recommended use cases?
The model is ideal for creative video generation tasks, particularly when specific start and end frames are required. It's well-suited for content creation, artistic projects, and scenarios where controlled video generation is needed with moderate computational resources.
