MissingLinkAI SDK example for Keras

Requirements

You need Python 2.7 or 3.5 on your system to run this example.

To install the dependency:

• You are strongly recommended to use virtualenv to create a sandboxed environment for individual Python projects

pip install virtualenv

• Create and activate the virtual environment

virtualenv .venv
source .venv/bin/activate

• Install dependency libraries

pip install -r requirements.txt

Run

In order to run an experiment with MissingLinkAI, you would need to first create a project and obtain the credentials on the MissingLinkAI's web dashboard.

With the owner_id and project_token, you can run this example from terminal.

python mnist.py --owner-id 'owner_id' --project-token 'project_token'

Alternatively, you can copy these credentials and set them in source files.

Examples

These examples train classification models for MNIST dataset.

• mnist.py: training with Keras's evaluate