Microsoft brings out a small language model that can look at pictures
Briefly

Phi-3-vision, a 4.2 billion parameter model from Microsoft, excels in general visual reasoning tasks like analyzing images or charts, without image generation capabilities.
Microsoft's Phi-3 series includes Phi-3-mini (3.8 billion parameters), Phi-3-small (7 billion parameters), Phi-3-medium (14 billion parameters), meeting various AI needs for different applications.
Microsoft's smaller AI models like Phi-3 cater to a rising demand for cost-effective, lightweight AI solutions, enabling efficient AI features on devices while conserving memory usage.
Read at The Verge
[
add
]
[
|
|
]