Skip to content

Florence-2-base-ft

A prompt-based vision foundation model from Microsoft

Visit official site

What Florence-2-base-ft is

Florence-2 is described as an advanced vision foundation model developed by Microsoft that works through a prompt-based approach. It interprets simple text prompts to run tasks such as image captioning, object detection and segmentation.

What you can do with it

  • Turn image content into text descriptions
  • Run object detection and region segmentation
  • Follow simple text prompts for a task

Application scenarios

  • A researcher generates descriptive image captions
  • A developer runs object detection on images
  • An enterprise labels and describes product images

Getting started

  • Import libraries such as requests, PIL and transformers
  • Load the model with AutoModelForCausalLM and AutoProcessor
  • Define the task prompt to run

Pros & cons

✓ What we like

  • Simple text prompts select the task
  • Custom code support is documented

! What to watch out for

  • Tasks are driven by text prompts
  • It is distributed as a model on Hugging Face

FAQ

Who developed it?

The material names Microsoft as the developer.

What dataset is used?

The material names FLD-5B with 5.4 billion annotations across 126 million images.

Last reviewed: 2026-11-15

More AI image generation tools

View all →

How we review