The NIFTY Financial News Headlines Dataset introduces a curated collection aimed at enhancing financial forecasting using large language models (LLMs). It consists of two versions: NIFTY-LM for supervised fine-tuning, and NIFTY-RL for alignment through RLHF. Key features include:

• Comprehensive data with detailed metadata and market indices.
• Systematic filtering and ranking suitable for current LLM frameworks.

Experiments highlighted:

• Efficacy in stock price movement prediction.
• Enhancing information richness through LLM embeddings.

This resource stands as a significant advancement providing extensive data for training and testing LLMs. It supports deeper insights into LLM’s abilities in processing financial information, critical for developments in AI-driven financial analytics. Further research might explore its adaptability in other areas of economic predictions.