H2O.ai + CORTX

Inspiration

Machine learning typically require huge datasets, which CORTX is purpose-built for. H2O allows for state-of-the-art machine learning.

What it does

It showcases how to apply H2O server configurations to use CORTX endpoints

How we built it

We purpose-built this integration via Google Colab's Jupyter Notebook, as it is easily replicable, rather than dealing with version control problems.

Challenges we ran into

H2O uses a Java backend, which gave us quite a significant challenge due to the fact that typical ML users use Python instead of Java.

Accomplishments that we're proud of

We have a sample notebook that runs through an entire AutoML pipeline to achieve SOTA results for a tabular dataset.

Built With

• colab
• cortx
• h2o
• jupyter
• python
• seagate