en-fincl-eval / README.md
jiminHuang's picture
Update dataset card: task, source, license, usage, citation, access form
7a5fea4 verified
|
Raw History Blame Contribute Delete
1.42 kB
---
task_categories:
- table-question-answering
license: cc-by-4.0
library_name: datasets
dataset_info:
features:
- name: context
dtype: string
- name: category
dtype: string
- name: entity
dtype: string
- name: entity_type
dtype: string
- name: query
dtype: string
- name: answer
dtype: string
splits:
- name: test
num_bytes: 242696233
num_examples: 52572
download_size: 3986141
dataset_size: 242696233
configs:
- config_name: default
data_files:
- split: test
path: data/test-*
tags:
- finance
- benchmark
- thefinai
- fintagging
- xbrl
extra_gated_heading: Request access to FinTagging FinCL eval
extra_gated_description: FinTagging FinCL eval is released by The Fin AI for research. Access is granted automatically after you complete this short form.
extra_gated_button_content: Agree and access
extra_gated_prompt: By accessing this dataset you agree to its license and to cite the FinTagging paper in any resulting publication.
extra_gated_fields:
Full name: text
Affiliation: text
Country: country
Intended use:
type: select
options:
- Research
- Education
- Commercial evaluation
- Other
I agree to the terms above and will cite the paper: checkbox
---
This dataset is from the paper [FinTagging: An LLM-ready Benchmark for Extracting and Structuring Financial Information](https://huggingface.co/papers/2505.20650).