IFT 6289 - Winter 2022

Job Description

As the Head Teaching Assistant for this course, I contributed to the following:

• Building upon the coding and theoretical assignments I developed for the 2022 course iteration, I updated two assignments on Word2Vec and pretrained language models (BERT and GPT-2) to reflect recent advancements. Additionally, I independently designed a new assignment focused on modern Large Language Models (LLMs). To the best of my knowledge, this course is among the first at major universities to include a dedicated assignment on Transformer-decoder-based LLMs.
• I developed the autograder for all coding assignments.
• I reviewed and corrected the majority of students’ theoretical assignments.
• I provided feedback on students’ course projects.
• I delivered a lecture on efficient deployment and training practices for LLMs.

Course Website

For more description of the course, please refer to IFT6289 - Natural Language Processing with Deep Learning.