Image classification for everyone
Project description
Image ATM (Automated Tagging Machine)
The Image ATM system is a one-click tool that automates the workflow of a typical image classification pipeline, this includes:
- preprocessing and validating input images and labels
- starting/terminating cloud instance with GPU support
- training
- model evaluation
For more details see the Image ATM confluence page.
Set up conda environment
conda env create -f environment.yml
conda activate image-atm
Testing
Run pytest -vs src/tests
Building
Run python setup.py install
Command line client
To see all available options and commands run
image-atm --help
Each component can be run individually or a pipeline can be run, which then checks in the config file which components to run.
TODOs:
- we are currently using Keras 2.2. The plan is to use tf.keras once TF 2.0 is out. Currently tf.keras is buggy, especially with model saving/loading (https://github.com/tensorflow/tensorflow/issues/22697)
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
imageatm-0.0.1.tar.gz
(24.0 kB
view hashes)
Built Distribution
imageatm-0.0.1-py3-none-any.whl
(40.7 kB
view hashes)