Aktiviere Job-Benachrichtigungen per E-Mail!

Master Thesis - Development and Validation of Small Language Models in Finance

Siemens Energy

Mülheim an der Ruhr

Hybrid

EUR 30.000 - 50.000

Vollzeit

Vor 30+ Tagen

Erhöhe deine Chancen auf ein Interview

Erstelle einen auf die Position zugeschnittenen Lebenslauf, um deine Erfolgsquote zu erhöhen.

Zusammenfassung

An innovative opportunity awaits you with this forward-thinking company, where you will delve into the development and validation of small language models tailored for the finance sector. This master thesis role offers a unique chance to explore cutting-edge technologies, conduct impactful research, and lay the foundation for your career in a global energy leader. Collaborate with experts, gain insights into the industry, and contribute to sustainable energy solutions while enhancing your skills in Python, machine learning, and language models. Join a diverse team dedicated to making a difference in the world.

Leistungen

Gain insights into an international company
Career foundation opportunities

Qualifikationen

  • Master's studies in Computer Science or Informatics required.
  • Strong practical experience with Python and knowledge of Machine Learning.

Aufgaben

  • Conduct literature research on Small Language Models and training methods.
  • Implement and execute a pipeline to create a small language model.

Kenntnisse

Python
Machine Learning
Language Models
LaTeX
MS Office
English
German
Snowflake

Ausbildung

Master's in Computer Science

Tools

MS Office
LaTeX
Snowflake

Jobbeschreibung

Master Thesis - Development and Validation of Small Language Models in Finance

About the Role
  • Country / Region: Germany
  • State / Province / County: Land Berlin
  • City: Berlin
  • State / Province / County: Bayern
  • City: Erlangen

Remote vs. Office Hybrid (Remote / Office) | Company: Siemens Energy Global GmbH & Co. KG | Organization: SE CFO Business Unit Gas Services | Full / Part time: Full-time | Experience Level: Student (Not Yet Graduated) | Location: MLH R / BLN H / ERL S | Mode of Employment: Full-time / Fixed Term

A Snapshot of Your Day

Large Language Models (LLMs) have gained prominence, especially after OpenAI’s GPT series. While these models possess broad knowledge, they often lack domain-specific expertise. Small language models are emerging as a promising solution for domain-specific tasks at lower costs. This master thesis aims to determine and implement the necessary steps to create and validate a small language model tailored for Finance.

How You’ll Make an Impact
  • Conduct literature research on Small Language Models, efficient training methods, open-source libraries, and datasets.
  • Identify all steps necessary to train and validate small language models.
  • Select and prepare the most suitable internal training dataset.
  • Implement and execute an efficient pipeline to create a small language model.
  • Validate the results using established standards and best practices.
  • Document and present your work through a master thesis.
What You Bring
  • Master’s studies in Computer Science or Informatics.
  • Fluent in English; German is a plus.
  • Strong practical experience with Python.
  • Solid knowledge of Machine Learning and Language Models.
  • Experience with LaTeX & MS Office.
  • Experience with Snowflake is a plus.
About Siemens Energy

Siemens Energy is dedicated to meeting global energy demands sustainably. With over 94,000 employees across more than 90 countries, we generate electricity for over 16% of the world's population and are committed to decarbonization and energy transformation. We value diversity, inclusion, and innovation, fostering a culture that supports character and character-driven talent.

Rewards / Benefits
  • Gain insights into an international company.
  • Lay the foundation for your career with us.

We welcome applications from individuals with disabilities and value equal opportunities.

Hol dir deinen kostenlosen, vertraulichen Lebenslauf-Check.
eine PDF-, DOC-, DOCX-, ODT- oder PAGES-Datei bis zu 5 MB per Drag & Drop ablegen.