/AutoML

Enter the AutoML

Primary LanguagePython

AutoML for Image Understanding

Enter the AutoML: Papers and Resources

To devs: recommended dev practice

1. Documentation: doc's doc

  • pip install sphinx
1.1 Docstring: Google style
  • pip install sphinxcontrib-napoleon
1.2 CMD: Generate API doc files
sphinx-apidoc -o rst ../src/utils -f
1.3 CMD: Generate html files from documentation
make html

2. Unit test

3. Formatter: Google Style

  • pip install yapf

4. Static Linting: code quality check

  • pip install pylint
4.1 CMD: Generate linting score and advice
pylint example.py

5. Logging

pip install glog

6. UML graph

  • brew install graphviz

  • pip install pyreverse

6.1 CMD: Generate UML graph
pyreverse -o png -p AutoML src/

To users: Getting Started

1. Requirements

pip install PyYAML

2. Example: Train DARTS on CIFAR10

2.1 Get docker image
2.2 Run docker container
docker run --runtime=nvidia -e NVIDIA_VISIBLE_DEVICES=0,1 --shm-size=8g -it  -v /data:/data automl:latest
2.3 Train DARTS
python src/core/pipeline.py > log.txt