HPC v praksi II: Od Podatkov do področnega modeliranja / HPC in Practice II: From Data to Topic Modeling
from
Monday, 11 May 2026 (09:00)
to
Wednesday, 13 May 2026 (12:30)
Monday, 11 May 2026
09:00
NorduGrid ARC: Connect to the grid
-
Pavle Boškoski
(
FIŠ Novo mesto
)
NorduGrid ARC: Connect to the grid
Pavle Boškoski
(
FIŠ Novo mesto
)
09:00 - 10:30
How to access a wider grid infrastructure and effectively manage data transmission over distance.
11:00
Foundations of Transformer-Based NLP
-
Biljana Mileva Boshkoska
(
FIŠ Novo mesto
)
Foundations of Transformer-Based NLP
Biljana Mileva Boshkoska
(
FIŠ Novo mesto
)
11:00 - 12:30
Explore the transition from traditional statistical methods to modern transformer architectures. Learn how to leverage pre-trained BERT models to extract deep semantic meaning from large-scale, unstructured text.
Tuesday, 12 May 2026
Wednesday, 13 May 2026
09:00
The BERTopic Pipeline:
-
Biljana Mileva Boshkoska
(
FIŠ Novo mesto
)
The BERTopic Pipeline:
Biljana Mileva Boshkoska
(
FIŠ Novo mesto
)
09:00 - 10:30
Dimensionality and Clustering A deep dive into the essential components: generating document embeddings, reducing high-dimensional data with UMAP, and mastering density-based clustering with HDBSCAN.
11:00
Apptainer Deployment, and Grid Execution
-
Pavle Boškoski
(
FIŠ Novo mesto
)
Apptainer Deployment, and Grid Execution
Pavle Boškoski
(
FIŠ Novo mesto
)
11:00 - 12:30
Finalize your thematic analysis and package your BERTopic model into an Apptainer container and use NorduGrid to deploy and execute the pipeline across HPC Trdina and other available HPC clusters. Lecturer: Pavle Boškoski