The purpose of this webinar is to demonstrate how to perform automated text classification using pre-trained, open-source large language models (LLMs). The efficiency that This webinar demonstrates how to perform automated text classification using pre-trained, open-source large language models (LLMs). The use case is classification of text according to sentiment, but the general workflow is applicable to many other text classification tasks.
In the webinar, it is demonstrated how to:
– Retrieve text data from online sources – Store and prepare the data for analysis – Set up an LLM text classifier – Perform the text classification – Use a web app for human validation of the classification result
Time stamps:
00:00: Introduction and agenda 02:04: Terminology and UCloud workflow outline 08:30: Best practices in web scraping 11:30: Setting up the UCloud workflow 18:30: Scraping and storing the data 27:06: Setting up the the classification pipeline 42:15: Classifying the data and analyzing the results 57:27: Creating a web app for result validation by humans
CVAT, Computer Vision Annotation Tool, is an interactive video and image annotation tool, designed to facilitate the annotation of video and image data and accelerate the creation of high-quality datasets for computer vision tasks. CVAT is available on the UCloud platform, in the Application Store.
The webinar will show how to use CVAT on UCloud to:
Label and annotate data with the help of AI and OpenCV tools, including:
Use of cvat-cli
Run built-in model for detection and auto-annotation
Use of GPUS with built in models for faster annotation
Adding custom models (e.g. YOLO)
Efficiently manage large visual datasets with MinIO:
Allow CVAT to directly pull images from your UCloud MinIO buckets for annotation and export annotated data back, reducing manual imports/exports and ensuring data availability.
Using UCloud allows users to create fully reproducible and secure workflows that leverage high performance computing resources. Those features are often necessary for large dataset and accurate computer vision tasks.
Target audience: Researchers across all Departments, particularly who require high-precision data labeling, AI interested.
I denne videovejledning får du en praktisk introduktion til UCloud, den nationale forskningsplatform for beregning, lagring og applikationer. Sessionen er tilrettelagt til at hjælpe nye brugere med at komme i gang med UCloud og forstå, hvordan platformen kan anvendes til forskning, undervisning og projektarbejde. Bemærk, at denne tutorial foregår på engelsk.
I optagelsen guider vi dig gennem, hvordan du:
Logger ind på UCloud og navigerer i dashboardet
Forstår centrale begreber som projekter, ressourcer og applikationer
Kører dit første job og ansøger om yderligere beregnings- og lagerressourcer
Administrerer filer og samarbejder via projektarbejdsområder
Udforsker applikationskataloget og mulighederne for jobafvikling
Bliver introduceret til nye funktioner i UCloud 4.0
Optagelsen er relevant for studerende, forskere og nye UCloud-brugere på tværs af alle fagområder.
UCloud er begyndervenlig og kræver hverken teknisk baggrund eller forudgående erfaring med cloud computing.
Tidsstempler
00:00 – 02:20: Introduktion Hvad UCloud er, hvem platformen er til, og hvad webinaret dækker.
02:20 – 03:50: Centrale begreber, du skal kende Enkle forklaringer af vigtige begreber, der bruges på tværs af platformen.
03:50 – 04:30: Supportressourcer og nyttige links Hvor du finder hjælp på interactivehpc.dk samt yderligere dokumentation.
04:30 – 06:10: Loginproces og overblik over UCloud-dashboardet Sådan logger du ind og navigerer i hoveddashboardet.
06:10 – 08:30: Kør dit første job En hurtig gennemgang af, hvordan du starter en applikation i UCloud.
08:30 – 14:00: Ansøgning om ressourcer Sådan anmoder du om beregnings- og lagerressourcer til dit projekt.
14:00 – 17:20: Filsystem og drev Hvordan fillagring fungerer, og hvordan du håndterer dine data.
17:20 – 18:00: Personligt arbejdsområde vs. projektarbejdsområde De vigtigste forskelle – og hvornår du bør bruge hvad.
18:00 – 21:30: Administration af et UCloud-projekt Medlemmer, indstillinger, tildelinger og samarbejde.
21:30 – 23:50: Ressourcesiden Offentlige links og IP-links, SSH-nøgler og relaterede indstillinger.
23:50 – 25:40: Applikationskatalog og dokumentation Sådan finder du applikationer og får adgang til relevante vejledninger.
25:40 – 28:50: Indstillinger på siden for jobafvikling Konfiguration af applikationer, før et job startes.
28:50 – 29:50: Visning af kørende job Overvågning af jobs og forståelse af jobstatus.
29:50 – 32:00: Arbejdsmappe, outputfiler og siden Runs Hvor du finder resultater, logfiler og jobhistorik.
32:00 – 35:40: Nyt i UCloud 4.0 Kommandopaletten, filtræet, Syncthing og forbrugssiden.
35:40 – 36:08: Afrunding og næste skridt Opsummering og henvisning til yderligere ressourcer.
CVAT (Computer Vision Annotation Tool) er et interaktivt værktøj til annotering af video- og billeddata, udviklet til at lette annoteringsarbejdet og accelerere opbygningen af datasæt af høj kvalitet til computer vision-opgaver. CVAT er tilgængeligt på UCloud-platformen via Application Store.
Webinaret vil vise, hvordan man anvender CVAT på UCloud til at:
mærke og annotere data ved hjælp af AI- og OpenCV-værktøjer, herunder:
brug af cvat-cli
kørsel af indbyggede modeller til detektion og automatisk annotering
anvendelse af GPU’er sammen med indbyggede modeller for hurtigere annotering
tilføjelse af brugerdefinerede modeller (f.eks. YOLO)
effektivt at håndtere store visuelle datasæt med MinIO:
give CVAT direkte adgang til at hente billeder fra dine UCloud MinIO-buckets til annotering og eksportere annoterede data tilbage igen, hvilket reducerer manuelle importer/eksporter og sikrer datatilgængelighed.
Ved at bruge UCloud kan brugere etablere fuldt reproducerbare og sikre workflows, der udnytter højtydende computerressourcer. Disse funktioner er ofte nødvendige ved arbejde med store datasæt og præcise computer vision-opgaver.
Målgruppe: Forskere på tværs af alle institutter, særligt dem med behov for højpræcis datamærkning, samt forskere med interesse for AI.
Date: 19 February 2026 Time: 13:15 – 14:30 (CET) Location: Online, via Zoom
The purpose of this webinar is to demonstrate how to perform automated text classification using pre-trained, open-source large language models (LLMs). The efficiency that LLMs bring to otherwise laboursome workflows, such as text classification, makes it possible to work with much larger text corpora across all fields of research. The use case will focus on classifying text according to sentiment, but the general workflow is applicable to many other text classification tasks.
In the webinar, we will create a complete workflow which will consist of the following parts:
Retrieving text data from online sources
Storing and preparing the data for analysis
Setting up the LLM text classifier
Performing the text classification
Displaying the results (for validation etc.)
The workflow will be set up on the UCloud platform. Using UCloud allows users to create fully reproducible and secure workflows that leverage high performance computing resources which are often necessary to run LLMs locally.
I denne webinaroptagelse får du en hands-on introduktion til Dictaphone, som er en ny app på UCloud, der gør det muligt for forskere at optage og transskribere interviews direkte fra deres egne enheder – også selvom du arbejder med fortrolige data. Bemærk, at webinaret foregår på engelsk.
I denne video vil vi guide dig igennem, hvordan du:
Optager interviews og samtaler via Dictaphone direkte fra din laptop eller smartphone. Lyden streames i realtid til den sikre UCloud-backend, så data ikke lagres lokalt på din enhed – hvilket gør Dictaphone velegnet til håndtering af fortrolige data.
Transskribere optagelser automatisk i samme arbejdsgang ved hjælp af Dictaphones indbyggede transskriptionsfunktion, så du hurtigt kan omdanne tale til tekst
Får mest muligt ud af Dictaphone, herunder tips, ekstra funktioner og eksempler på anvendelse i forskning.
Optagelsen er relevant for alle forskere på tværs af discipliner og fakulteter – samt studerende.
Dictaphone er begyndervenlig og kræver ingen teknisk baggrund.
Tidsstempler
00:00 – 06:25: Introduktion og kom godt i gang Krav, dataklassifikationer og den grundlæggende arbejdsgang.
06:25 – 32:15: Live demonstration af Dictaphone Hvordan du optager og transskriberer, og hvordan du arbejder på platformen både under og efter optagelsen.
32:15 – 33:50: Datalagring og sikkerhed rdan Dictaphone lagrer data, og hvorfor appen er egnet til håndtering af fortrolige data.
33:50 – 35:30: Relaterede ressourcer og support Andre webinars, relaterede UCloud-apps og kontaktinformation.
35:30 – 43:15: Spørgsmål og svar (Q&A) Spørgsmål fra deltagerne.
43:15 – 44:22: Afrunding og næste skridt Opsummering og henvisning til yderligere ressourcer.
In this video we will guide you through two different AI based workflows, involving ChatUI and CVAT apps.
You will learn how to:
Use advanced CVAT features including auto-annotation, algorithmic assistance, management and analytics.
Use Chat UI as a flexible interface for hosting of various LLM models, and interact via a chat or API environment.
Use ChatUI for semantic search in a knowledge base.
Use CVAT as a powerful annotation tool, including image classification, object detection, semantic and instance segmentation, and video / 3D annotations.
In this video we will guide you through the complete pipeline of transcribing audio files from speech to text and editing and classifying transcription segments.
In this session, you’ll learn how to:
Use Transcriber for transcribing audio/video files. Transcriber is based on Open AI’s Whisper language model. The app can transcribe speech audio to text in various formats and uses the WhisperX package to perform speaker recognition.
Navigate the new, simple, drag and drop Transcriber user interface to make it easier for you to use AI to transcribe audio files.
Edit and classify the transcriptions with Speech Analyzer. Speech Analyzer is an application built on top of Label Studio, specifically optimized for dialogue analysis. It enables you to label, edit, and annotate transcriptions generated using Transcriber.
Perform a comprehensive dialogue analysis on UCloud involving transcribing audio files using Transcriber, followed by transcription analysis with Speech Analyzer.
All workflows will be executed inside a UCloud project environment with access to GPU resources.
Target audience: Researchers across all Departments, particularly Digital Humanities and Social Science, Students, AI interested.
In this video we will guide you through the complete pipeline of fine-tuning large language models (LLMs) for specialised tasks such as medical question-answering using NeMo Framework and Triton Inference Server.
Prepare and preprocess open-source datasets for fine-tuning.
Apply Parameter-Efficient Fine-Tuning (PEFT) using LoRA with NVIDIA NeMo Framework.
Deploy optimised LLMs using NVIDIATriton Inference Server and TensorRT-LLM.
Generate a synthetic Q&A dataset using Label Studio connected to a live inference backend.
Fine-tune and evaluate your customised LLM for domain-specific applications.
All workflows will be executed inside a UCloud project environment with access to GPU resources.
Target audience: Machine learning practitioners, researchers, and engineers interested in LLM customisation, domain adaptation, or scalable model deployment.
In this video we will go through the process of developing a course on UCloud, using UCloud Courses – a tool for hosting and managing university courses on UCloud. Wnt to know more about this new feature, check out our webinar recording introducing UCloud Courses.
Introduction
00:00 – Outline of the workshop agenda 03:10 – Introduction to the UCloud Courses concept 04:23 – Advantages of using UCloud Courses 05:56 – Outline of the steps involved in developing a UCloud Course 09:54 – Requesting and planning a UCloud Course 10:20 – Example of the planning and structuring of an existing UCloud Course
Developing a UCloud Course step by step
Setting up the environment 15:53 – Preliminary remarks 17:15 – Showing the existing UCloud Course that will be re-developed in the workshop 18:02 – Showing the UCloud-Courses GitHub repository 18:40 – Software requirements and -recommendations 20:22 – Short introduction to Git and GitHub 24:27 – Cloning the repository 25:15 – Creating a working branch 26:12 – What should go in the UCloud-Courses repository and what shouldn’t 27:24 – Installing the required dependencies
Creating a new UCloud Course and modifying the templates 28:00 – Creating a UCloud Course using a prepared script 30:31 – Walk-through of the different auto-generated files/folders in the course folder 33:35 – Short introduction to Docker 38:15 – Modifying the Dockerfile 42:32 – What should go in the UCloud-Courses repository and what shouldn’t 45:28 – Building the Docker image locally using a prepared script 47:54 – Modifying the starting script 01:00:50 – What can, shouldn’t, and mustn’t be changed while the course is running
Building and testing the course locally 01:05:00 – Re-building the Docker image locally using a prepared script 01:05:25 – Running a Docker container locally using a prepared script 01:10:58 – Opening the JupyterLab interface on localhost
Finalising the course app 01:12:48 – Opening a pull request and requesting code review 01:19:44 – Testing the course app on UCloud before it’s deployed
Closing remarks
01:25:44 – Reusing/updating an existing UCloud course 01:28:19 – The financial model in brief 01:29:38 – Useful links and resources
