Our AI Models

Explore our collection of cutting-edge AI models, each designed to push the boundaries of what's possible.

VideoAutoEncoder

View on GitHub →

An efficient Video Autoencoder designed for graphics with limited VRAM memory. Features the new AdaptiveEfficientVideoAutoencoder that supports different qualities and durations.

Key Features:

  • Optimized for 3GB VRAM
  • Support for 480p 5s videos at 15fps
  • Improved reconstruction quality in v0.3.0
⭐ Loading... 🔄 Loading...