NanoDefectNet

End-to-end deep learning app for detecting defects in semiconductor wafers

Install pytorch, torchvision via pip install torch==2.6.0 torchvision==0.21.0 --index-url https://download.pytorch.org/whl/cu118 - Issues when trying to install via environment.yaml file

Add the project root ie, Folder containing this README to PYTHONPATH whichever way you want. One way would be to create a .env and write the following in it

PYTHONPATH=\full\path\to\projectroot

And place this .env file in the project root. Works for VS Code.

Another option would be to run $env:PYTHONPATH = \full\path\to\projectroot in powershell to set the env variable and then run the scripts.

Execute pip install ipykernel==6.30.1 beforing running the jupyter notebooks. And in case of ValueError: Mime type rendering requires nbformat>=4.2.0 but it is not installed error run pip install jupyter

Sanity

Before committing changes run pre-commit run --all-files or pre-commit run --file <file1>, <file2> ...

Preprocessing

Run python .\nanodefectnet\scripts\data_preprocess.py

Data Augmentation

Run python .\nanodefectnet\scripts\augment_train_data.py

Training

For ResNet model: Run python .\nanodefectnet\run_train_test.py --path_config_file .\configs\classifier_resnet152_aug.yaml

Inference

For ResNet model: Run python .\nanodefectnet\run_infer.py --model_name=ResNet152 --image=assets/test_images/center_defect.png --path_infer_config_file=configs/inference/infer.yaml

API

Start the REST server using uvicorn nanodefectnet.server.main:app --reload

Test using windows powershell: curl.exe -X POST "http://127.0.0.1:8000/api/predict-waferdefect?model_name=ResNet152" -H "accept: application/json" -H "Content-Type: multipart/form-data" -F "file=@D:/Computer Vision/Projects/NanoDefectNet/assets/test_images/center_defect.png"

Docker (for API inference only and NOT for training/validation/test pipeline)

Build the image using docker build -t nanodefectnet-app -f deploy/Dockerfile.serve .

Run docker image using docker run --gpus all -p 8000:8000 nanodefectnet-app

Test using windows powershell: curl.exe -X POST "http://127.0.0.1:8000/api/predict-waferdefect?model_name=ResNet152" -H "accept: application/json" -H "Content-Type: multipart/form-data" -F "file=@D:/Computer Vision/Projects/NanoDefectNet/assets/test_images/center_defect.png"

Testing

Run only unit tests - pytest -m unittest

Run only integration tests - pytest -m integration

Run only tests that can be run on CI - pytest -m runonci

Run ALL tests - pytest

Note : Anytime a pytest marker is added to a pytest, ensure it is registered in pytest.ini otherwise pytest will complain