Training Proof Of Work Dapp

Inspiration

The inspiration for this project came from my interest in training machine learning models but not having the required compute for bigger projects.

What it does

Allows users to contribute to a pool of compute and receive crypto payments. Also allows for users to post tasks that need trained.

How we built it

Pysyft federated learning framework allows for the decentralized nodes to train models, along with truffle and ganache for smart contract development and testing. Paired with a simple frontend and flask python backend.

Challenges we ran into

I ran into a lot of challenges with versions of python being compatible with pysyft and torch.

Accomplishments that we're proud of

I feel accomplished with my proof of concept/bare bones project, I implemented the harder backend aspects.

What we learned

I learned a lot about decentralized machine-learning/datascience, as well as smart contract development and testing.

What's next for Training Proof Of Work Dapp

Eventually there will be the ability for users to submit datasets and based on dataset quality will also get rewarded.

Built With

• flask
• pysyft
• python
• truffle