Skip to content

Latest commit

 

History

History
24 lines (15 loc) · 835 Bytes

README.md

File metadata and controls

24 lines (15 loc) · 835 Bytes

Embryo stage classification

Model for classifying Drosophila embryo stages

Usage:

  • install conda on your laptop, if you don't have it yet
  • install a new environment from the embryo_class.yml as follows:

conda env create -f embryo_class.yml

  • activate the environment:

conda activate embryo_class

  • use the script as follows:

python predict_from_folder.py /path/to/folder_with_images

Script options:

  • --device - cpu by default, can be set to GPU number (requires a different conda env)
  • --batch_size - 10 by default, reduce if script crashes/stalls
  • --num_workers - 10 by default, reduce if script crashes/stalls
  • --model - stage_classifier.pth by default, change when using a different model or if the model is in a different directory