Bỏ qua đến nội dung chính
Back to home
AI tools-ai 1 min read

Microsoft launches Lens — A 3.8-billion-parameter text-to-image model

Microsoft has released its Lens model on Hugging Face, a text-to-image model featuring 3.8 billion parameters that supports resolutions of up to 1440x1440.

Tier 1 · sources 99% confidence Reviewed
Sources x.com

Microsoft has officially released Lens, a new text-to-image model, on the Hugging Face platform, aiming for both training efficiency and high-quality image generation.

Developments

The Lens model has 3.8 billion parameters and is designed to strike a balance between computational costs and detailed image generation capabilities. A key highlight of Lens is its ability to generate high-resolution images up to 1440x1440 pixels, while maintaining a more efficient training process compared to other models in its class.

Why it matters

The release of Lens highlights a trend toward optimizing generative AI models to run on moderate infrastructure while maintaining high quality. For AI startups in Vietnam, Lens provides another powerful open-source option to integrate into content creation or design products.