Title of education

Application development with LLM
Introduction to the development of applications using LLM (large language models). Examples of using LLM in various business applications and tasks. Transformers principles of operation. Learning principles of large language models for text generation. Learning and fine-tuning LLMs. Evaluation of LLMs. Benchmark data for evaluation.

Content of the training:

Application development project based on LLM. Phases and activities. Techniques for application development based on large language models: Prompt engineering. Fine-tuning. Castastrophic forgetting. Parameter efficient fine-tuning-PEFT and Low-Rank Adaptation of Large Language Models- LoRA techniques. Learning from human feedback and Reinforcement learning with human feedback RLHF. Retrieval Augmented Generation (RAG). Vector bases. Search vector databases. LLM agents. Construction of LLM applications- LangChain. Hardware requirements. Quantification of the grid parameter. Risks and limitations of generative AI. Bias and hallucinations. The difference between using pre-trained and fine-tuned models. Responsible use of AI and impact on society. The safety of using generative AI.

Duration: 5h

Prof. dr. sc. Sanda Martinčić-Ipšić, Faculty of Informatics and Digital Technologies Rijeka

Location: Faculty of Informatics and Digital Technologies Rijeka, Hall 357, Radmile Matejčić 2, Rijeka; online

Term: 17.04.2024, from 2 p.m. to 7 p.m.


Edih Adria
Privacy Overview

This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognising you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.