Jupyter lab & Swin OD model Inference

jupyterlab

• 놀랄 만큼 사용이 간단하다

# venv activate
source venv/bin/activate

# install jupyter lab
pip install jupyter lab

# use config file - run jupyter lab
jupyter lab config=jupyter_notebook_config.py

• 끝!

Model Serving

Swin Transformer - Object Detection

• https://github.com/SwinTransformer/Swin-Transformer-Object-Detection
  • 위 코드를 활용하여 inference 하고싶음!
• https://github.com/open-mmlab/mmdetection/blob/master/docs/get_started.md
  • 일단 여기서 하라는 설치 다 해주고

# model file download
wget https://github.com/SwinTransformer/storage/releases/download/v1.0.3/moby_mask_rcnn_swin_tiny_patch4_window7_1x.pth

# coco annotations download
wget http://images.cocodataset.org/annotations/annotations_trainval2017.zip

# unzip
unzip annotations_trainval2017.zip

# and make data/coco/annotation dir in root of Swin-Transformer dir
mkdir data
mkdir data/coco
cd data/coco
cp -r Swin-Transformer-Object-Detection/annotations .

# test inference execution : 근데 아직 에러남
python tools/test.py configs/swin/mask_rcnn_swin_tiny_patch4_window7_mstrain_480-800_adamw_1x_coco.py mask_r_cnn_swinT_Lr1x/mask_rcnn_swin_tiny_patch4_window7_1x.pth --eval bbox segm

• 다 하고 위 inference command 수행했지만
  • KeyError: “MaskRCNN: ‘SwinTransformer is not in the models registry’”
  • 뭔가 설치가 덜 되었나..