Closed LWaetzig closed 9 months ago
Using language model via API calls is obviously much simpler but less precise -> is currently implemented as a MVP Finetuning is the best method to fulfill our objectives, probably a we need to finetune a german language model on summarization for the best results - prototype could be an english summarization model
The scope and target of the finetuning has to be discussed and can afterwards be tackled & implemented
Objective
Research how to access llm (using huggingfaceAPI or download model to run it locally)
Key features