raeidsaqur/nifty-rl

The News-Informed Financial Trend Yield (NIFTY) Dataset.

The News-Informed Financial Trend Yield (NIFTY) Dataset. Details of the dataset, including data procurement and filtering can be found in the paper here: https://arxiv.org/abs/2405.09747.

📋 Table of Contents

• 🧩 NIFTY Dataset
  • 📋 Table of Contents
  • 📖 Usage
    • Downloading the dataset
    • Dataset structure
  • Large Language Models
  • ✍️ Contributing
  • 📝 Citing
  • 🙏 Acknowledgements

📖 Usage

Downloading and using this dataset should be straight-forward following the Huggingface datasets framework.

Downloading the dataset

The NIFTY dataset is available on huggingface here and can be downloaded with the following python snipped:

from datasets import load_dataset

# If the dataset is gated/private, make sure you have run huggingface-cli login
dataset = load_dataset("raeidsaqur/nifty-rl")

Dataset structure

The dataset is split into 3 partition, train, valid and test and each partition is a jsonl file where a single row has the following keys.

['prompt', 'chosen', 'rejected', 'chosen_label', 'chosen_value']

Currently, the dataset has 2111 examples in total, the dates randing from 2010-01-06 to 2020-09-21.

✍️ Contributing

We welcome contributions to this repository (noticed a typo? a bug?). To propose a change:

git clone https://huggingface.co/datasets/raeidsaqur/nifty-rl
cd nifty-rl
git checkout -b my-branch
pip install -r requirements.txt
pip install -e .

Once your changes are made, make sure to lint and format the code (addressing any warnings or errors):

isort .
black .
flake8 .

Then, submit your change as a pull request.

📝 Citing

If you use the NIFTY Financial dataset in your work, please consider citing our paper:

@article{raeidsaqur2024NiftyRL,
    title        = {NIFTY-RL: Financial News Headlines Dataset for LLM Alignment using Reinforcement Learning.},
    author       = {Raeid Saqur},
    year         = 2024,
    journal      = {ArXiv},
    url          = {https://arxiv.org/abs/2024.5599314}
}

🙏 Acknowledgements

The authors acknowledge and thank the generous computing provided by the Vector Institute, Toronto.