Datasets
Dataset
BhashaBench-Multi
by bharatgenai
BhashaBench-Multi Dataset Description BhashaBench-Multi is a large-scale multilingual multiple-choice question answering (MCQ) benchmark covering four specialised Indian knowledge domains across 22 Indian languages plus English. No existing open-source benc…
CC-BY-4.0multiple-choice 566 51M < n < 10M rows
published 5 Jun 2026
Preview
First 5 rows of BBA / Assamese, from the Hugging Face dataset viewer.
| id | question | option_a | option_b | option_c | option_d | correct_answer | question_type | question_level | topic | subject_domain | llm_judge_scores | source_lang | target_lang | question_translated | option_a_translated | option_b_translated | option_c_translated | option_d_translated | question_backtranslated | option_a_backtranslated | option_b_backtranslated | option_c_backtranslated | option_d_backtranslated | translation_quality_metrics |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 6cda890c-2f56-46fa-bb5d-8cee9835c23f | Immune disorders include ....., | Hypersensitivity | auto-immune diseases | immunodeficiency | all of these | D | MCQ | Easy | Kayachikitsa | Kayachikitsa (General Medicine & Internal Medicine in Ayurveda) | {"factual_accuracy":8,"question_clarity":8,"option_quality":7,"translation_fidelity":9,"safety":10} | en | as | ইমিউনজনিত বিকাৰসমূহৰ ভিতৰত আছে ..... | অতিরিক্ত সংবেদনশীলতা | স্বয়ং-প্রতিরোধ ক্ষমতা জনিত ৰোগ | রোগ প্রতিরোধ ক্ষমতাৰ অভাৱ | এই সকলোবোৰ। | Among the immune-related disorders are... | excessive sensitivity | Autoimmune diseases. | Lack of immunity. | That’s all. | {"bleu":19.494,"chrf":40.9208,"chrf++":33.0998} |
| 5b177fe6-d7d7-486a-aeef-ab47f4da428a | Snigdha and Ruksha Udvartan are the types mentioned by Acharya……… | Sushruta | Charak | Sharangdhar | Vagbhat | B | MCQ | Easy | Panchakarma | Panchakarma & Rasayana | {"factual_accuracy":9,"question_clarity":8,"option_quality":8,"translation_fidelity":9,"safety":10} | en | as | স্নিগ্ধা আৰু ৰুক্শা উডবাৰ্তন হৈছে আচাৰ্য্যৰ দ্বাৰা উল্লেখ কৰা প্ৰকাৰ। | সুশ্রুত | চৰাক | শৰংধৰ | বাগভট | Snigdha and Ruksha are types mentioned by the teacher. | Sushruta | Birds | Sharandhar | Bagbhat | {"bleu":28.228,"chrf":61.1178,"chrf++":61.4655} |
| 692f6618-7b68-4dee-9e9e-24d4f0218559 | In Charak Samhita, total number of Pushpaasava are --------------- in number. | 2 | 4 | 9 | 10 | B | MCQ | Easy | Samhita & Basic Principles | Samhita & Siddhanta (Fundamentals) | {"factual_accuracy":8,"question_clarity":8,"option_quality":7,"translation_fidelity":6,"safety":10} | en | as | চাৰক সংহিতাত, পুস্পাসৱৰ মুঠ সংখ্যা হ’ল ---------------। | 2 | চাৰি | নয় | ১০ | In the Charaka Samhita, the total number of herbal preparations is ---------------. | 2 | four | No. | 10 | {"bleu":14.9498,"chrf":55.5542,"chrf++":50.8071} |
| f14335e7-f1df-4dde-b0c3-877c49cbcdf3 | Amla Kamji is used in ................... Sneha Basti Vyapad | Annavrutta | Kaphavrutta | Pittavrutta | Vatavrutta | D | MCQ | Medium | Panchakarma | Panchakarma & Rasayana | {"factual_accuracy":9,"question_clarity":7,"option_quality":5,"translation_fidelity":7,"safety":10} | en | as | আমলা কামজি নিম্নলিখিতবোৰত ব্যৱহাৰ কৰা হয়: স্নেহ বস্তি ব্যৱসায়। | অন্নবৃত | কফযুক্ত | পিতৃবৃত্ত | বৃত্তাকাৰ | Amala is used in the following: the lubricant business. | Annawrita | containing cough | Father's occupation. | circle | {"bleu":3.5005,"chrf":16.8052,"chrf++":20.2541} |
| a384e6d5-4166-4b88-bdfb-1ea86a2a52f0 | The main office of W. H. O. is at ……………………….. | Singapore | Dubai | Geneva- Switzerland | California | C | MCQ | Easy | Ayurvedic Pharmacology | Dravyaguna & Bhaishajya | {"factual_accuracy":10,"question_clarity":9,"option_quality":9,"translation_fidelity":9,"safety":10} | en | as | ডব্লিউ. এইচ. ও.-ৰ মূল কাৰ্যালয় ইয়াত অৱস্থিত: ……………………….. | সিঙ্গাপুৰ | দুবাই | জেনভা – সুইজাৰলেণ্ড। | কেলিফৰ্ণিয়া | The WHO headquarters is located here: ……………………….. | Singapore | Dubai | Geneva – Switzerland. | California | {"bleu":10.7643,"chrf":32.715,"chrf++":29.7057} |
Cite this
CC-BY-4.0textCite
Tags
task_categories:multiple-choicetask_categories:question-answeringtask_categories:translationsource_datasets:originallanguage:enlanguage:aslanguage:bnlanguage:brxlanguage:doilanguage:gulanguage:hilanguage:knlanguage:kslanguage:koklanguage:mailanguage:mllanguage:mnilanguage:mrlanguage:nelanguage:orlanguage:palanguage:salanguage:satlanguage:sdlanguage:talanguage:telanguage:urlicense:cc-by-4.0size_categories:1M<n<10Mformat:parquetmodality:textlibrary:datasetslibrary:pandaslibrary:polarslibrary:mlcroissantarxiv:2510.25409region:us