Skip to content

Latest commit

 

History

History
26 lines (16 loc) · 886 Bytes

README.md

File metadata and controls

26 lines (16 loc) · 886 Bytes

This directory contains a few tools.

  • convert-pretrained-model-to-d2.py

Tool to convert ImageNet pre-trained weights for D2.

  • analyze_model.py

Tool to analyze model parameters and flops.

Usage for semantic segmentation (ADE20K only, use with caution!):

python tools/analyze_model.py --num-inputs 1 --tasks flop --use-fixed-input-size --config-file CONFIG_FILE

Note that, for semantic segmentation (ADE20K only), we use a dummy image with fixed size that equals to cfg.INPUT.CROP.SIZE[0] x cfg.INPUT.CROP.SIZE[0]. Please do not use --use-fixed-input-size for calculating FLOPs on other datasets like COCO!

Usage for panoptic and instance segmentation:

python tools/analyze_model.py --num-inputs 100 --tasks flop --config-file CONFIG_FILE

Note that, for panoptic and instance segmentation, we compute the average flops over 100 real validation images.