title: Multilingual NLP

This page contains information and teaching materials for the “Multilingual NLP” lecture (for the year 2025-2026)

Lecture n°1 (2025-09-17)

  • no lecture this week

Lecture n°2 (2025-09-24)

  • Babel Unbound: Discovering the Wonders of Many Languages pdf
  • 1st lab: pdf) due date October 8th

Lecture n°3 (2025-10-01)

  • The Astonishing Multilingual Capacities of LLMs pdf
  • 2nd lab: Word Alignment with Multilingual BERT: From Subwords to Lexicons pdf) due date October 15th

Lecture n°4 (2025-10-08)

  • Building Bitexts Corpora pdf
  • No new lab 🥳

Lecture n°5 (2025-10-15)

  • Evaluating Machine Translation: Human Evaluation pdf
  • NLP Lab Assignment: Analysing MQM Error Profiles and Metric Correlations in MT Systems pdf due date November 5th (the lab has been updated: I have added a link to the second part of the dataset that you will need to compute automatic metrics)

Lecture n°6 (2025-11-05)

  • Evaluating Machine Translation: Automatic Metric pdf

Lecture n°7 and n°8 (2025-11-12 and 2025-11-19)

  • Machine translation models pdf
  • Lab n°4 : Decoding strategies for MT pdf
  • code for greedy decoder py and model tar.bz2

Lecture n°9 (2025-11-26)

  • Translating with LLMs pdf
  • Lab n°5 : Translating with LLMs pdf

All labs must be submitted no later than 8 a.m. on 12 January 2026 — hard deadline.