QASC

Overview

QASC (Question Answering via Sentence Composition) is a question-answering dataset with a focus on multi-hop sentence composition. It consists of 9,980 8-way multiple-choice questions about grade school science, requiring models to combine multiple facts to arrive at the correct answer.

Task Description

Task Type: Multi-hop Science Question Answering (Multiple-Choice)
Input: Science question with 8 answer choices
Output: Correct answer letter
Focus: Sentence composition and multi-hop reasoning

Key Features

9,980 grade school science questions
8-way multiple-choice format
Requires composing two facts to answer
Tests multi-hop reasoning over scientific knowledge
Annotated with supporting facts for each question

Evaluation Notes

Default configuration uses 0-shot evaluation
Evaluates on validation split
Simple accuracy metric
Useful for evaluating compositional reasoning

Properties

Property	Value
Benchmark Name	`qasc`
Dataset ID	extraordinarylab/qasc
Paper	N/A
Tags	`Knowledge`, `MCQ`
Metrics	`acc`
Default Shots	0-shot
Evaluation Split	`validation`

Data Statistics

Metric	Value
Total Samples	926
Prompt Length (Mean)	362.58 chars
Prompt Length (Min/Max)	283 / 505 chars

Sample Example

Subset: default

{
  "input": [
    {
      "id": "d92a545e",
      "content": "Answer the following multiple choice question. The entire content of your response should be of the following format: 'ANSWER: [LETTER]' (without quotes) where [LETTER] is one of A,B,C,D,E,F,G,H.\n\nClimate is generally described in terms of what?\n\nA) sand\nB) occurs over a wide range\nC) forests\nD) Global warming\nE) rapid changes occur\nF) local weather conditions\nG) measure of motion\nH) city life"
    }
  ],
  "choices": [
    "sand",
    "occurs over a wide range",
    "forests",
    "Global warming",
    "rapid changes occur",
    "local weather conditions",
    "measure of motion",
    "city life"
  ],
  "target": "F",
  "id": 0,
  "group_id": 0,
  "metadata": {}
}

Prompt Template

Prompt Template:

Answer the following multiple choice question. The entire content of your response should be of the following format: 'ANSWER: [LETTER]' (without quotes) where [LETTER] is one of {letters}.

{question}

{choices}

Usage

Using CLI

evalscope eval \
    --model YOUR_MODEL \
    --api-url OPENAI_API_COMPAT_URL \
    --api-key EMPTY_TOKEN \
    --datasets qasc \
    --limit 10  # Remove this line for formal evaluation

Using Python

from evalscope import run_task
from evalscope.config import TaskConfig

task_cfg = TaskConfig(
    model='YOUR_MODEL',
    api_url='OPENAI_API_COMPAT_URL',
    api_key='EMPTY_TOKEN',
    datasets=['qasc'],
    limit=10,  # Remove this line for formal evaluation
)

run_task(task_cfg=task_cfg)

3.0 KiB Raw Permalink Blame History