sciq

References:

• Code
• Huggingface

Use the following command to load this dataset in TFDS:

ds = tfds.load('huggingface:sciq')

• Description:

The SciQ dataset contains 13,679 crowdsourced science exam questions about Physics, Chemistry and Biology, among others. The questions are in multiple-choice format with 4 answer options each. For the majority of the questions, an additional paragraph with supporting evidence for the correct answer is provided.

• License: No known license
• Version: 0.1.0
• Splits:

• Features:

{
    "question": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "distractor3": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "distractor1": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "distractor2": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "correct_answer": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "support": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}