Practical training: Praktikum Natual Language Processing - Details
You are not logged into Stud.IP.
Lehrveranstaltung wird in Präsenz abgehalten.

General information

Course name Practical training: Praktikum Natual Language Processing
Course number INF-0441
Semester SS 2023
Current number of participants 13
expected number of participants 15
Home institute Embedded Intelligence for Health Care and Wellbeing
participating institutes Institut für Informatik
Courses type Practical training in category Teaching
Next date Mon., 25.09.2023 09:00 - 19:00, Room: (1005 N)
Participants The Praktikum will take place as a block course at the end of the winter term (February/March 2021).
Online/Digitale Veranstaltung Veranstaltung wird in Präsenz abgehalten.
Hauptunterrichtssprache englisch
ECTS points 5

Course location / Course dates

(1005 N) Monday. 25.09.23 - Friday. 29.09.23 09:00 - 19:00

Module assignments


It is highly recommended to complete the lecture "Deep Learning" (INF-0315) or similar courses prior to this course.
In the Natural Language Processing Praktikum, students will apply their machine learning knowledge to text data.
First, "traditional" methods to automatically process and analyse texts are introduced, giving the students some foundational intuitions on the nature of natural language data and the challenges in dealing with it. The main part of the course is devoted to recent Transformer-based Language Models in the fashion of BERT and its successors which have achieved remarkable success in virtually all areas of Natural Language Processing. The transformer architecture is investigated in depth. Afterwards, the students utilise pretrained Transformer Language Models to adress text classification and language generation tasks in a hands-on manner. Lastly, more advanced topics such as few-shot learning and textual style transfer are adressed. The focus of this course is programming, while the necessary background knowledge is provided by introductory tutorials.
NOTE: Fundamental knowledge of Deep Learning and first practical experience with either PyTorch or Tensorflow are required. PyTorch will be used as the deep learning framework in this course.
Please NOTE that you need to register via STUDIS in order to participate in the Praktikum.

Admission settings

The course is part of admission "Anmeldung gesperrt (global)".
The following rules apply for the admission:
  • The admission is locked.