root
/
RE-OWOD
mirror of https://github.com/RE-OWOD/RE-OWOD
1
0
Fork 0
Code Issues Projects Releases Wiki Activity
RE-OWOD/tools
History
RE-OWOD 39a272f3d3
Add files via upload 		2022-01-04 17:26:58 +08:00
..
deploy Add files via upload 2022-01-04 17:26:58 +08:00
README.md Add files via upload 2022-01-04 17:26:58 +08:00
analyze_model.py Add files via upload 2022-01-04 17:26:58 +08:00
benchmark.py Add files via upload 2022-01-04 17:26:58 +08:00
convert-torchvision-to-d2.py Add files via upload 2022-01-04 17:26:58 +08:00
plain_train_net.py Add files via upload 2022-01-04 17:26:58 +08:00
plot_energy.py Add files via upload 2022-01-04 17:26:58 +08:00
plot_tsne.py Add files via upload 2022-01-04 17:26:58 +08:00
train_net.py Add files via upload 2022-01-04 17:26:58 +08:00
tsne.py Add files via upload 2022-01-04 17:26:58 +08:00
visualize_data.py Add files via upload 2022-01-04 17:26:58 +08:00
visualize_json_results.py Add files via upload 2022-01-04 17:26:58 +08:00

README.md

This directory contains a few scripts that use detectron2.

  • train_net.py

An example training script that's made to train builtin models of detectron2.

For usage, see GETTING_STARTED.md.

  • plain_train_net.py

Similar to train_net.py, but implements a training loop instead of using Trainer. This script includes fewer features but it may be more friendly to hackers.

  • benchmark.py

Benchmark the training speed, inference speed or data loading speed of a given config.

Usage:

python benchmark.py --config-file config.yaml --task train/eval/data [optional DDP flags]
  • visualize_json_results.py

Visualize the json instance detection/segmentation results dumped by COCOEvalutor or LVISEvaluator

Usage:

python visualize_json_results.py --input x.json --output dir/ --dataset coco_2017_val

If not using a builtin dataset, you'll need your own script or modify this script.

  • visualize_data.py

Visualize ground truth raw annotations or training data (after preprocessing/augmentations).

Usage:

python visualize_data.py --config-file config.yaml --source annotation/dataloader --output-dir dir/ [--show]

NOTE: the script does not stop by itself when using --source dataloader because a training dataloader is usually infinite.