# IndicGenBench XorQA-IN: Cross-lingual Open-Retrieval Question Answering in 29 Indic Languages (Google Research)

## Facts

- Provider: Google Research (Singh et al.)
- License: MIT (https://github.com/google-research-datasets/indic-gen-bench)
- Size: 89 files
- Format: JSON (in sampled files)
- Stored at: Hugging Face, https://huggingface.co/datasets/google/IndicGenBench_xorqa_in
- Access: public, readable without an account
- Category: Natural Language Processing
- Source: https://arxiv.org/abs/2404.16816
- Page: https://scigantic.com/archive/feb70439-3ca8-5ec4-87e1-28c56ffd8aa4
- Keywords: Natural Language Processing, benchmark, JSON, Hugging Face

## About

XorQA-IN is IndicGenBench's cross-lingual question answering task, built on XOR QA (cross-lingual open-retrieval question answering): a question is posed in an Indic language, the supporting evidence passage is in English, and the model must produce the answer, testing whether a model can bridge languages between the question and the evidence. IndicGenBench is a multilingual, multi-way parallel benchmark from Google Research for measuring language generation by large language models in 29 Indic languages spanning 13 writing scripts and 4 language families, including many low-resource languages such as Assamese, Bhojpuri, Awadhi, Maithili, Santali and Konkani alongside Hindi, Bengali, Tamil, Telugu, Marathi, Gujarati, Kannada and Malayalam. Google extended existing English benchmarks by collecting human translations of their English examples into each target language. The repository holds 87 JSON files, one train, dev and test file per language, named like xorqa_bho_test.json (about 45 MB in total). Load with the Hugging Face datasets library or plain JSON readers. Distributed under the MIT License, following the XOR QA source license; cite Singh et al., IndicGenBench: A Multilingual Benchmark to Evaluate Generation Capabilities of LLMs on Indic Languages, ACL 2024.

## Load the data

```python
from huggingface_hub import hf_hub_download
path = hf_hub_download(repo_id="google/IndicGenBench_xorqa_in", filename="xorqa_as_dev.json", repo_type="dataset")
print(path)
```

## Data structure

File types in sampled files: JSON

Example file: `xorqa_as_dev.json`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/feb70439-3ca8-5ec4-87e1-28c56ffd8aa4
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/feb70439-3ca8-5ec4-87e1-28c56ffd8aa4
- Over MCP: `get_schema_card` with id `feb70439-3ca8-5ec4-87e1-28c56ffd8aa4` at https://api.scigantic.com/mcp

## Related topics

- [JSON datasets](https://scigantic.com/topics/json)
- [Benchmark datasets](https://scigantic.com/topics/benchmark)
- [Natural Language Processing datasets](https://scigantic.com/topics/natural-language-processing)
