This work in progress repository contains a Streamlit application that allows users to ask questions about authors of the former GDR / Eastern Germany.
The first version (currently on main branch) uses a pre-trained model via Groq's API to find relevant excerpts from lexicon articles and generate responses to the user's questions.
The second version (currently on dev branch) takes text as input and extracts relevant information based on a set of pre-defined prompts to return structured output.
-
Question-Answering System: Users can ask questions about authors, and the application will generate responses based on relevant excerpts from lexicon articles.
-
Customization: Users can provide additional context for the model and choose from a list of pre-trained models.
-
Similarity Search: The application uses a vector store to find the most relevant excerpts from lexicon articles based on the user's question.
- Information Extraction: Users can paste lexicon articles and choose from a set of pre-defined prompts to extract relevant information.