The raw dataset used in this project is sourced from the virattt/financial-qa-10K repository. This dataset has been carefully processed for fine-tuning the meta-llama/Meta-Llama-3.1-8B-Instruct model, specifically for the Retrieval-Augmented Generation (RAG) use case.
The dataset consists of five key columns:
Question: The financial question posed in the dataset.
Answer: The corresponding answer to the financial question.
Context: Additional relevant context to help provide a more accurate… See the full description on the dataset page:
https://huggingface.co/datasets/Shubhu07/processed_chat_messages_financial-qa-10K_for_llama.
