Skip to main content

Exploiting generative artificial intelligence (4cr)

Code: KL00DX67-3002

General information


Enrollment
22.04.2025 - 08.10.2025
Registration for the implementation has begun.
Timing
20.10.2025 - 14.12.2025
The implementation has not yet started.
Number of ECTS credits allocated
4 cr
Local portion
4 cr
Mode of delivery
Contact learning
Unit
SeAMK Automation Engineering and Information Technology
Campus
SeAMK Seinäjoki, Frami
Teaching languages
Finnish
Degree programmes
Bachelor of Engineering, Information Technology
Teachers
Matti Panula
Groups
TITE24
Bachelor of Engineering, Information Technology
Course
KL00DX67

Evaluation scale

1-5

Objective

Student can explain the principles of language models (LLM) and is able to apply them in practical tasks, also programmatically. Student can install LLM locally or on a private cloud. Student can fine-tune LLM and apply Retrieval-Augmented Generation (RAG) techniques.

Content

- Using the large language model (LLM) programmatically
- Exploiting a locally installed language model
- Exploring cloud-based solutions
- Language model fine-tuning and Retrieval-Augmented Generation (RAG)

Assessment criteria, satisfactory (1)

Student can explain the principles of language models (LLM) and is able to apply them in practical tasks, also programmatically. Student can install LLM locally or on a private cloud.

Assessment criteria, good (3)

Student can explain the principles of language models (LLM) and is able to apply them in practical tasks, also programmatically. Student can install LLM locally or on a private cloud. Student can fine-tune LLM and apply Retrieval-Augmented Generation (RAG) techniques.

Assessment criteria, excellent (5)

Student can explain the principles of language models (LLM) and is able to apply them in practical tasks, also programmatically. Student can install LLM locally or on a private cloud. Student can fine-tune LLM and apply Retrieval-Augmented Generation (RAG) techniques. Student understands how fine-tuning improves the performance of the model.

Go back to top of page