Upcoming Event
The Most (Subjectively) Interesting Bottlenecks of Large Language Models Development
Ivan P. Yamshchikov of Technical University of Applied Sciences Würzburg-Schweinfurt will highlight LLM bottlenecks, from tokenization, reproducible evaluation to empathetic personalization.
Studies in Prague
High-quality affordable bachelor’s and master’s programs taught in English, at one of the oldest and most highly regarded universities in Europe and in one of the most student-friendly cities.
Hands-on R for humanities
This intensive course will teach you to tidy your data as well as perform basic analysis with descriptive statistics and scientific visualisation techniques.
No programming skills required, in a non-competitive and highly supportive environment for scholars who get easily intimidated by computers.
Congratulations
to Ondřej Bojar and Dominik Macháček
on their U.S. patent Computer-implemented method of real time speech translation and a computer system for carrying out the method No. 12,056,457 issued on August 6, 2024.
New projects funded by GAČR
ÚFAL proposals were accepted for funding:
- Jindřich Libovický: Better Tokenization for Multilingual Models and Machine Translation,
- Veronika Kolářová: Czech Non-Verbal Predicates Motivated by Nouns and Their Syntactic Behaviour,
- Dominik Macháček: Live Speech Translation with Confidence.
Charles Translator
Charles Translator translates between Czech, Ukrainian, Russian, Polish, English and French.