Your tasks include:

 

  • research and assessment of open source LLMs such as Mistral, LLaMA 2 or Falcon, as well as relevant technologies and frameworks
  • Conducting experiments to evaluate the suitability of various models for EWDE-specific applications
  • Exploration of fine-tuning techniques to adapt LLMs to the specific requirements and data sets of the EWDE
  • Development and testing of prototypes for specific in-house applications
  • Documentation of results, processes, and decision-making criteria as well as preparation of decision templates for internal strategy development
  • Reporting on progress, obstacles, and insights regarding feasible applications and required internal resources for scaling

You meet the following requirements:

 

  • current enrollment in a relevant field of study (e.g., Computer Science, Data Science, Cognitive Sciences, etc.)
  • basic understanding and initial experience with open source LLMs and machine learning
  • practical experience with the Python programming language and common AI libraries/tools (Hugging Face, vector databases, etc.)
  • Knowledge of fine-tuning techniques of LLMs is desirable
  • high self-initiative and the ability to develop solutions
  • very good knowledge of either German or English