Skip to content

Latest commit

 

History

History
 
 

249-oneformer-segmentation

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 
 
 

Universal segmentation with OneFormer with OpenVINO™

This tutorial explains how to convert and run inference on the OneFormer HuggingFace model with OpenVINO. Additionally, NNCF quantization is applied to improve OneFormer segmentation speed.

Notebook Contents

This tutorial demonstrates step-by-step instructions on how to run HuggingFace OneFormer with OpenVINO and quantize it with NNCF.

The tutorial consists of the following steps:

  • Install required libraries
  • Prepare the environment
  • Load OneFormer fine-tuned on COCO for universal segmentation
  • Convert the model to OpenVINO IR format
  • Select inference device
  • Choose a segmentation task
  • Inference
  • Quantization
    • Preparing calibration dataset
    • Run quantization
    • Compare model size and performance
  • Interactive demo