Explore the world of Neural Machine Translation (NMT) in this comprehensive 90-minute tutorial presented by Annie En-Shiun Lee, Assistant Professor at the University of Toronto's Computer Science Department. Delve into the rapid evolution of NMT and the power of Pre-trained Multilingual Sequence to Sequence (PMSS) models like mBART and mT5. Learn how these models, pre-trained on extensive general data, can be fine-tuned for impressive results in various natural language tasks. Gain insights into adapting pre-trained models for NMT, discover essential tips and tricks for training and evaluation, and understand the challenges faced when implementing these models. Whether you're approaching NMT from a research or industry perspective, this tutorial offers valuable knowledge to enhance your understanding and application of cutting-edge translation technology.
Pre-Trained Multilingual Sequence to Sequence Models for NMT - Tips, Tricks and Challenges
Toronto Machine Learning Series (TMLS) via YouTube
Overview
Syllabus
Pre-Trained Multilingual Sequence to Sequence Models for NMT Tips, Tricks and Challenges
Taught by
Toronto Machine Learning Series (TMLS)
Related Courses
-
Applying Neural Networks: A Guide to Pre-trained Models
-
Gen AI Foundational Models for NLP & Language Understanding
-
CMU Multilingual NLP - Unsupervised Translation
-
CMU Multilingual NLP - Machine Translation-Sequence-to-Sequence Models
-
The Lottery Ticket Hypothesis for Gigantic Pre-Trained Models
-
The Evolution of Domain Transfer in the Era of Pre-trained Language Models
