MILU(Multi-task Indic Language Understanding Benchmark,多任务印度语言理解基准)是一个全面的评估数据集,用于衡量大语言模型(LLM)在11种印度语言上的表现。该数据集涵盖8个领域和41个主题,结合了翻译而来的通用知识问题以及具有印度文化特色的本地内容。
{
"input": [
{
"id": "84726982",
"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.\n\nBakelite is what type of polymer?\n\nA) Thermosetting polymer\nB) Thermoplastic polymer\nC) Fibre\nD) Elastomer"
}
],
"choices": [
"Thermosetting polymer",
"Thermoplastic polymer",
"Fibre",
"Elastomer"
],
"target": "A",
"id": 0,
"group_id": 0,
"metadata": {
"language": "English"
}
}
提示模板
提示模板:
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}