Deciphering Poorly Formatted Text using Private LLM
Process unstructured text data from document files efficiently, prioritizing data privacy with the implementation of locally hosted private large language models. Employed a PEFT model (LoRA) to optimize finetuning for the summarizer and classifier, yielding improved performance metrics with the help of prompt engineering and LLM reasoning.
Apr 25, 2023
