metadata
license: apache-2.0
task_categories:
- text2text-generation
language:
- en
tags:
- finance
pretty_name: Financial Statements
size_categories:
- 10K<n<100K
Financial Statements Dataset for FinGPT Fine-Tuning
Overview
This dataset contains 3000 records, specifically designed for fine-tuning a FinGPT model to generate insights from financial statements. The dataset includes:
- 1000 Balance Sheets
- 1000 Income Statements
- 1000 Cash Flow Statements
Each record consists of:
- Input: The text of a financial statement (Balance Sheet, Income Statement, or Cash Flow Statement)
- Output: Corresponding insights, analysis, or advice based on the financial statement
Dataset Structure
Each entry in the dataset is structured as follows:
{
"input": "Type of Financial Statement and its content",
"output": "Generated insights, analysis, or advice"
}