Datasetpublic

Grass-G/jee-main-questions

JEE Main — Question Bank A structured dataset of JEE Main examination questions with full metadata, worked solutions, and diagrams. Built for education, ML training, and question-generation use cases. Subsets: Chemistry — 738 questions from 28 papers Physics — 768 questions from 28 papers Mathematics — 801 questions from 28 papers Over 2,300 questions across the three core JEE subjects. Structure Organised into subsets by subject and splits (train / test):… See the full description on the dataset page: https://huggingface.co/datasets/Grass-G/jee-main-questions.

sourceHugging Facecc-by-4.0updated 2mo agoView on Hugging Face

0likes168downloads

Dataset Card

JEE Main — Question Bank

A structured dataset of JEE Main examination questions with full metadata, worked solutions, and diagrams. Built for education, ML training, and question-generation use cases.

Subsets:

—Chemistry — 738 questions from 28 papers
—Physics — 768 questions from 28 papers
—Mathematics — 801 questions from 28 papers

Over 2,300 questions across the three core JEE subjects.

Structure

Organised into subsets by subject and splits (train / test):

chemistry/          physics/           mathematics/
├── train.jsonl     ├── train.jsonl    ├── train.jsonl
├── test.jsonl      ├── test.jsonl     ├── test.jsonl
└── images/         └── images/        └── images/

Choose the subject from the Subset dropdown and the split from the Split dropdown in the Dataset Viewer.

Schema

Field	Type	Description
`question_id`	string	Unique ID, e.g. `CH-14-Q1`, `P-05-Q1`, `M-10-Q3`.
`question`	string	Question text. `[IMAGE]` marks where a diagram appears inline.
`question_images`	list[str]	Relative paths to images used in the question.
`option_1`…`4`	string	The four answer choices (empty for numerical questions).
`correct_option`	int/null	Index (1–4) of the correct choice for single-correct questions.
`numerical_answer`	str/null	The numeric answer for integer/numerical questions.
`solution`	string	Full worked solution. `[IMAGE]` marks inline figures.
`solution_images`	list[str]	Relative paths to images used in the solution.
`subject`	string	`Chemistry`, `Physics`, or `Mathematics`.
`topic`	string	e.g. `Physical Chemistry`, `Thermodynamics`, `Circles`.
`subtopic`	string	e.g. `Electrochemistry`, `Cyclic process`.
`difficulty`	string	e.g. `Easy` / `Moderate` / `Tough` (where provided).
`question_type`	string	`single_correct` / `numerical`.
`has_image`	bool	Whether the question or solution contains a figure.
`exam`	string	`JEE Main`.
`source_paper`	string	Original source paper reference.

Images

Original .emf / .wmf vector figures have been converted to PNG and stored under each subject's images/ folder. Image fields hold relative paths (e.g. images/image5.png).

Coverage notes

—The large majority of questions include a verified answer key. A small number of the earliest papers were supplied without an answer key and are included for their questions and solutions.
—Topic / subtopic / difficulty metadata is present for most questions where the source paper provided it.
—Questions support both single-correct (4 options) and integer/numerical types.
—Answers are captured across several source conventions (numeric, letter A–D, integer, and answers embedded in the solution line).

Loading

python

from datasets import load_dataset
chem = load_dataset("eQOURSE/jee-main-questions", "chemistry")
phys = load_dataset("eQOURSE/jee-main-questions", "physics")
math = load_dataset("eQOURSE/jee-main-questions", "mathematics")