Key features of CM3leon include:
- State-of-the-art performance for text-to-image generation
- Trained with five times less compute than previous models
- Versatility and effectiveness of autoregressive models
- Can generate sequences of text and images conditioned on other content
- Improved performance on tasks such as image caption generation and visual question answering
- Ability to generate complex compositional objects
- Efficient training costs and inference efficiency