This repository was created so you can quickly start using Engine ML's free experiment tracking offering. Seconds after creating an Engine ML account, you can run an image classifier on the MNIST dataset using PyTorch, TensorFlow, or Keras by following these 3 simple steps.
- Clone this repository and download the MNIST dataset by running
./vision/mnist/get-data.sh
at the root of this repository. - Install the python dependencies:
pip install -r vision/mnist/pytorch/requirements.txt
(We recommend using Python>=3.6). - Launch your first experiment:
engine run vision/mnist/pytorch/local.yaml -o repository <OWNER>/quickstart
.
If you have not set up your Engine ML account or created a new project called quickstart
, see our
documentation for a more in-depth quickstart as well as a reference manual
for our CLI
and our Python library.
Visit our website or request a demo for a complete list of features.