Overview
This course explores the concept of BERT pruning in relation to the Lottery Ticket Hypothesis, demonstrating that even subnetworks of the model can achieve comparable performance to the full model. The course covers topics such as BERT, the Lottery Ticket Hypothesis, paper abstract, pruning BERT, experiments, and conclusions. The intended audience for this course includes individuals interested in natural language processing, machine learning, and neural networks. The teaching method involves analyzing a specific paper and discussing its implications and findings.
Syllabus
- Overview
- BERT
- Lottery Ticket Hypothesis
- Paper Abstract
- Pruning BERT
- Experiments
- Conclusion
Taught by
Yannic Kilcher