Datasets:

Phanhai2804
/

2_2_Table_Understanding

pretty_name: Vietnamese Financial Table Understanding
language:
  - vi
task_categories:
  - question-answering
tags:
  - vietnamese
  - finance
  - financial
  - benchmark
  - format:json
  - library:datasets
  - table-understanding
  - structured-data
configs:
  - config_name: default
    data_files:
      - split: test
        path: 2_2_Table_Understanding.json

Vietnamese Financial Table Understanding

Overview

Đánh giá khả năng đọc hiểu bảng biểu và trả lời câu hỏi dựa trên dữ liệu dạng bảng.

Part of the Viet Financial Benchmark collection.

Dataset information

Language: Vietnamese (vi)
Domain: Finance and financial services
Format: JSON
Supported library: Hugging Face Datasets
Configuration: default
Split: test

Dataset Viewer

The default configuration points to the original JSON file and exposes it as the test split.

The Hugging Face Dataset Viewer must successfully parse the JSON structure to render its table preview. Nested objects, unusual schemas, or unsupported structures may require additional configuration.

Usage

from datasets import load_dataset

dataset = load_dataset(
    "Phanhai2804/2_2_Table_Understanding",
    name="default",
    split="test",
)

print(dataset)
print(dataset[0])

Intended use

This dataset is intended for evaluation and benchmarking of language models on Vietnamese financial tasks.

Please verify the annotation quality, evaluation methodology, and data usage permissions before using the dataset.

License

No license is declared because the applicable redistribution and usage permissions have not yet been verified.