Content/Schedule
Abschnittsübersicht
-
Block 1: Fundamentals
17.4.
L1: Languages of the world & Linguistic Universals; Course organization
24.4. L2: Language modeling, word embedding models, tokenization & vocab building 26.4 Ex1: Intro to Language Modeling & Tokenization 15.5. L3: Deep Learning for (Modern) NLP — Perceptron/MLP, Backprop, Batching, Gradient Descent, Dropout… 17.5 Ex2: Backprop & Training Models in Pytorch 22.5. L4: Transformer Almighty & Pretraining Language Models (autoregressive, masked language modeling) 24.5 (Online) Ex3: Transformer Block 2: Multilinguality
6.6. (changed, online)
L5: Multilingual Word Embedding Spaces (and CL Transfer with them)
31.5 (online)
Ex4: Project Topics & Setup
12.6. L6: Multilingual LMs and Cross-Lingual Transfer; Tasks, Benchmarks & Evaluation 19.6. L7: Curse of Multilinguality, Modularization, and Language Adaptation 26.6. L8: Transfer for Token-Level Tasks: Word Alignment & Label Projection (+ maybe translate train on sequence labeling task) 28.6 Ex5: Modularization Block 3: Advanced Topics
3.7.
L9: Neural Machine Translation (incl. decoder only MT)
5.7
Ex6: Transfer for Token-Level Tasks
10.7. L10: Multilingual Sentence Representations 12.7 Ex7: Neural Machine Translation 17.7. L11: Large Language Models, Instruction-Tuning and Generative NLP 19.7 Ex8: Multilingual Sentence Representations