Chat with your data repository for KaggleX project in cohort3
This project developed an application that will allow legal professionals such as lawyers, notaries, etc., to be more efficient by being able to ask questions to the application about very extensive documents that they would have to work on, and that would otherwise have to be read in full to be able to extract the information.This application would allow them to ask questions about what they are interested in knowing about the document without having to read it, saving time and money.
Explore options with LLM and NLP, such as fine-tuning and prompting, to “converse” with electronic documents and extract information of our interest, for personal or enterprise usage. The concepts explored / applied were: Generative AI applied to language, Natural Language Processing, Large Language Models
- Different types of LLM available to play with.
- What implies fine-tune a LLM model in terms of data and resources.
- How techniques like Prompting Engineering / Retrieval Augmented Generation can help.
This file has an application which displays an application with strealite, this applications allows user to input their API KEYs, to upload the data which they want to "converse" with (pdf files) and to make questions over that documents.
List libraries and versions requiered to be used with this application
Applications requiered to launch this application using streamlite