README.md

metadata

dataset_info:
  features:
    - name: Row_ID
      dtype: int64
    - name: Topic
      dtype: string
    - name: Subtopic
      dtype: string
    - name: Question
      dtype: string
    - name: Answer_1
      dtype: string
    - name: Explanation_1
      dtype: string
    - name: Answer_2
      dtype: string
    - name: Explanation_2
      dtype: string
    - name: Answer_3
      dtype: string
    - name: Explanation_3
      dtype: string
    - name: Difficulty
      dtype: string
  splits:
    - name: train
      num_examples: 62941756
license: mit
language:
  - en
tags:
  - chemistry
size_categories:
  - 10M<n<100M

Chemistry Questions Dataset (62M Rows)

Description

A dataset of 62,941,756 chemistry questions covering Organic Chemistry (Alkenes, Nomenclature), Inorganic Chemistry (Oxidation States), and Physical Chemistry (Kinetics). Each row includes a question, 2-3 answers with explanations, and difficulty level.

Columns

• Row_ID: Unique identifier (1 to 62,941,756).
• Topic: Main chemistry category (e.g., Organic Chemistry).
• Subtopic: Specific topic (e.g., Alkenes).
• Question: The chemistry question.
• Answer_1, Answer_2, Answer_3: Possible answers (Answer_3 often N/A).
• Explanation_1, Explanation_2, Explanation_3: Explanations for each answer.
• Difficulty: Hard or Expert.

Analysis

• Unique value counts are visualized in unique_value_counts.png.
• Sample analysis (based on 100k rows): See image for approximate counts per column.

Usage

from datasets import load_dataset
dataset = load_dataset("XythicK/chemistry")

Size

• Uncompressed: ~9.06GB