Winter 2025
This semester I am taking 3 technical classes:
- Natural Language Processing
- Quantum Computing
- Robot Dynamics and Simulation
My favourite of the classes was NLP. I got to learn a lot about the history of NLP from n-gram models all the way to new models on the edge of research. I learned about the foundations of these models like attention mechanisms, vector embeddings, etc. The final project in this class was an open ended problem asking us to build something related to NLP. My group decided on building a model with the goal of writing a docstring for a python function. We gathered a lot of data, processed this data, and generated a model to predict the final docstring. This approach was limited given we only had one GPU with very limited RAM to process the data. If we had more time, it would have been better to train the model in batches to be able to effectively train the model on all the data.
In Quantum Computing, I learned about the foundations of QC and the main algorithms associated with it. I came into the class with very little knowledge of the field and ended feeling comfortable explaining how different algorithms worked and how different gates could be applied to these qubits to work the way it is intended.
In Robot Dynamics, the class was much more pure math and physics. The class started by building our foundations with how reference frames and rotation matrices work. Then we covered Lagrange and Kane's method to derive equations of motion. I liked how this class went over a lot of example problems such as robotic arms to describe degrees of freedom. We also got to learn the sympy library to be able to build robot simulations for our examples.