High Speed Training Using Binary Neural Networks

Group Members

1. Daniel Chen

2. Thomas Forzani

3. Daniel Maevsky

4. Sachin Mathew

5. Serena Zhang

Project Objective

Training machine learning systems is currently very slow. Recent work has shown promise by using simpler representations of numbers than the commonly used floating point ones. In this project, you will create and measure neural networks which use only binary or integer numbers for both training and inference.

Works Developing On

• GoNN Github: ​https://github.com/sausheong/gonn
• MLFPGA Github: ​https://github.com/myakubov18/MLFPGA/tree/binary
• Our Project Github: ​https://github.com/sachmathew/FixedPointNN

Week 1 Activities

​Week 1 Presentation

Week 2 Activities

​Week 2 Presentation

Week 3 Activities

​Week 3 Presentation

Week 4 Activities

​Week 4 Presentation

Week 5 Activities

​Week 5 Presentation

Week 6 Activities

​Week 6 Presentation

Week 7 Activities

​Week 7 Presentation

Week 8 Activities

​Week 8 Presentation

Final Presentation

[ Final Presentation]

Research Poster

[ Research Poster]