README.md

Examples

This directory contains examples of using the notebook card. Both of these examples involve training a model and visualizing various performance metrics and diagnostics in a Jupyter Notebook as part of your Flow. The notebook is dynamically updated with the results of the Flow. There are two different example flows, one that trains a model with Tensorflow and another with a Random Forest.

Note that we are using Conda for dependency management in these examples. We understand that not everyone uses Conda, so we have also included a requirments.txt file in each directory. However, we recommend using Conda due to the complex dependencies machine learning libraries often have.

Instructions on running these examples are as follows:

Deep Learning

1. Setup the environment

   cd deep_learning
   conda env create -f environment.yml
   conda activate mf-demo-dl

2. Run the flow

   python dl_flow.py --package-suffixes=".ipynb"  run 

3. View the card

   python dl_flow.py card view nb_auto

---

Random Forest

1. Setup the environment

   cd random_forest
   conda env create -f environment.yml
   conda activate mf-demo-rf

2. Run the flow

   python flow.py --package-suffixes=".ipynb"  run 

3. View the card

   python flow.py card view evaluate