Project Configurations: In “config.py”, set your Openai api key as LLM used in the project is "gpt-4o-mini" Prompt used to generate LLM response is in “prompts.py”
Embedding model used : Currently, the project usts "all-MiniLM-L6-v2" as embedding model, which can be replaced with any Huggingface embedding Model. Refer to helper.py
Project contain 2 entry points:
-
Data_ingestion.py: It takes PDF files “directory” as input and ingests all data into Chroma DB python3 data_ingestion.py --directory /path/to/pdf/folder
-
qa.py It takes a “question” as input and provides an answer based on ingested information. python3 qa.py --q your question