tahamajs/Bitcoin-Long-Term-Trend-and-Price-Prediction-Dataset

tags:

financial-forecasting

time-series

instruction-tuning

bitcoin

finance license: apache-2.0

Bitcoin Long-Term Trend and Price Prediction Dataset This dataset is designed for fine-tuning language models on a long-term Bitcoin forecasting task. The goal is to predict both the overall price trend (up, down, or no change) and the specific daily closing prices for the next 10 days, based on a comprehensive 60-day historical context.

The dataset provides a rich, multi-factor view for each prediction point, combining technical indicators, wider market sentiment, and randomized samples of news and social media from the entire 60-day lookback period.

Dataset Description The core task is to analyze a 60-day period of market and sentiment data to make a 10-day forecast. This dual-prediction format (trend + prices) encourages the model to learn both high-level market direction and fine-grained price movements.

The context provided in the input field is a multi-source snapshot of the market, including:

Technical Analysis: Key indicators (Price, RSI) for Bitcoin, Gold, and Crude Oil on the most recent day of the historical window.

Randomized News Sample: A random selection of 10 news articles from the entire 60-day lookback period, providing a broad overview of the recent news environment.

Randomized Social Media Sample: A random selection of 10 social media posts from the 60-day lookback period, capturing a wide range of public sentiment.

Data Sources Market Data: Daily closing prices, volume, and technical indicators for BTC-USD, Gold Futures (GC=F), and WTI Crude Oil Futures (CL=F) were fetched using the yfinance library.

News Articles: Sourced from the edaschau/bitcoin_news dataset on Hugging Face.

Social Media Posts: Sourced from the gauravduttakiit/bitcoin-tweets-16m-tweets-with-sentiment-tagged dataset on Kaggle.

How to Use The dataset is provided in JSON format and can be loaded directly using the datasets library.

from datasets import load_dataset

Load the dataset from the Hugging Face Hub

repoid = "tahamajs/bitcoin-long-term-trend-dataset" # <-- Replace with your repo ID dataset = loaddataset(repo_id)

Access the training split

train_data = dataset['train']

Print the first sample

print(train_data[0])

Data Structure The dataset is split into train.json and validation.json. Each entry is a JSON object with the following fields:

instruction: (string) - A directive for the language model, explaining the dual task of predicting both the trend and the specific prices for the next 10 days.

input: (string) - The context the model should use. This multi-line string is formatted with clear headings:

[Technical Analysis for YYYY-MM-DD]: Contains the closing price and 14-day RSI for Bitcoin, Gold, and Oil.

[Random Sample of News (last 60 days)]: A string containing 10 randomly sampled news articles from the historical window, separated by ||.

[Random Sample of Social Media Buzz (last 60 days)]: A string containing 10 randomly sampled social media posts from the historical window, separated by ||.

output: (string) - The ground truth for the prediction. This string contains two parts, separated by ||:

Trend: The overall price trend over the next 10 days, classified as up, down, or no change.

Prices: A comma-separated string of the next 10 days of Bitcoin's closing prices.

Example Data Point { "instruction": "Analyze the provided historical prices, technical indicators, news, and social media sentiment from the last 60 days. Based on this context, predict the overall trend (up, down, or no change) and the specific daily closing prices for Bitcoin for the next 10 days.", "input": "[Technical Analysis for 2021-03-01]\nBTC Price: 49631.24, BTC RSI: 52.48\nGold Price: 1723.00, Gold RSI: 35.91\nOil Price: 61.50, Oil RSI: 68.79\n\n[Random Sample of News (last 60 days)]\nMicroStrategy buys another $10M of Bitcoin: The business intelligence firm now holds over 90,000 BTC on its balance sheet. || Bitcoin's Taproot upgrade is a 'major milestone,' says analyst: The upgrade, set to activate later this year, will improve privacy and smart contract capabilities.\n\n[Random Sample of Social Media Buzz (last 60 days)]\nIs this the dip we've been waiting for? Buying more #Bitcoin now. #BTFD || The energy consumption of #Bitcoin is a real problem. We need a more sustainable solution.", "output": "Trend: up || Prices: 48541.36, 49709.08, 50538.28, 51206.08, 48870.74, 47559.24, 48378.34, 49796.86, 51220.57, 52633.53" }

Citation If you use this dataset in your research, please consider citing it:

@misc{majlesi2025bitcoinlongterm, author = {taha majlesi}, title = {Bitcoin Long-Term Trend and Price Prediction Dataset}, year = {2025}, publisher = {Hugging Face}, journal = {Hugging Face repository}, howpublished = {\url{https://huggingface.co/datasets/tahamajs/bitcoin-long-term-trend-dataset}}, }

Disclaimer: This dataset is intended for research and educational purposes only. Financial markets are highly volatile, and past performance is not indicative of future results. Do not use models trained on this data for actual financial trading.