Duration
7 hours (usually 1 day including breaks)
Requirements
- Some programming experience is helpful
- Basic understanding of neural networks
- Experience using the command line
Audience
- Localization specialists with a technical background
- Global content managers
- Localization engineers
- Software developers in charge of implementing global content solutions
Overview
In this instructor-led, live training, participants will learn how to use Facebook NMT (Fairseq) to carry out translation of sample content.
By the end of this training, participants will have the knowledge and practice needed to implement a live Fairseq based machine translation solution.
Format of the course
- Part lecture, part discussion, heavy hands-on practice
Note
- If you wish to use specific source and target language content, please contact us to arrange.
Course Outline
Introduction
- Why Neural Machine Translation?
- Borrowing from image recognition techniques
Overview of the Torch and Caffe2 projects
Overview of a Convolutional Neural Machine Translation model
- Convolutional Sequence to Sequence Learning
- Convolutional Encoder Model for Neural Machine Translation
- Standard LSTM-based model
Overview of training approaches
- About GPUs and CPUs
- Fast beam search generation
Installation and setup
Evaluating pre-trained models
Preprocessing your data
Training the model
Translating
Converting a trained model to use CPU-only operations
Joining to the community
Closing remarks