Date: March 13th, 2024
Time: 10:30 AM – 03:30 PM
Location: Friedrich-Hirzebruch-Allee 6, Room 0.107
Description:
This tutorial will discuss recent popular small and large language models (LLMs) and impart a practical understanding of how these state-of-the-art LLMs can be utilized across various NLP tasks. Specifically, we cover three interesting research directions: Interpretability of LLMs, Robustness of LLMs, and Bridging language on machines vs. language in the brain. Also, we focus on reviewing recent methodologies such as efficient prompting, in-context learning, instruction-tuning, and RAG, and demonstrate how these techniques can be integrated into various real-world applications.
Presenters:
Prof. Dr. Lucie Flek is a full professor at the University of Bonn and the Lamarr Institute for Machine Learning and AI, leading the Data Science and Language Technologies group at Bonn-Aachen International Center for Information Technology.
Dr. Mounika Marreddy is a postdoctoral researcher in the Data Science and Language Technologies group at the University of Bonn, Germany.
Dr. Manish Gupta is a Principal Applied Researcher at Microsoft, India. He is also an Adjunct Faculty at IIIT Hyderabad and a visiting faculty at the Indian School of Business (ISB).
Dr. Subba Reddy Oota submitted his PhD thesis at Inria Bordeaux, France. He is about to join the postdoctoral position in collaboration with TU Berlin and UC Berkeley.
Schedule:
Hands-on Hacking:
- We will see the different prompt-based examples using the Flan-T5-base instruction model
- How to work on larger language models in your local machine LMStudio Link: https://lmstudio.ai/ -> You can download the LMStudio based on your OS requirements.
- How to load the Mistral and Llama-2 model in Google Colab
Feedback on LAMARR NLP Workshop:
We would like to kindly ask you to take a 5-minute survey to help us organize the upcoming LAMARR NLP events better. Thank you!