---
configs:
- config_name: tags
  data_files:
  - split: train
    path: "xbrl_tags_train.csv"
  - split: test
    path: "xbrl_tags_test.csv"
- config_name: value
  data_files:
  - split: train
    path: "value_train.csv"
  - split: test
    path: "value_test.csv"
- config_name: formula
  data_files:
  - split: train
    path: "formula_formatted_with_tags_train.csv"
  - split: test
    path: "formula_formatted_with_tags_test.csv"
- config_name: formula_calculations
  data_files:
  - split: train
    path: "formula_calculation_train.csv"
  - split: test
    path: "formula_calculation_test.csv"
---

# XBRL Extraction Dataset

The is the official dataset introduced in the paper [FinLoRA: Benchmarking LoRA Methods for Fine-Tuning LLMs on Financial Datasets](https://arxiv.org/abs/2505.19819)

<p>
  <a href="https://huggingface.co/spaces/wangd12/xbrl_llm_demo"><img src="https://raw.githubusercontent.com/wangd12rpi/FinLoRA/main/static/demo_btn.svg"></a>
  <a href="https://huggingface.co/spaces/wangd12/xbrl_llm_demo"><img src="https://raw.githubusercontent.com/wangd12rpi/FinLoRA/main/static/models_btn.svg"></a>
  <a href="https://finlora-docs.readthedocs.io/en/latest/"><img src="https://raw.githubusercontent.com/wangd12rpi/FinLoRA/main/static/doc_btn.svg"></a>
</p>