Skip to content

Latest commit

 

History

History
275 lines (150 loc) · 5.91 KB

multiple-choice-quiz-generator.md

File metadata and controls

275 lines (150 loc) · 5.91 KB

Multiple choice quiz generator | Start Chat

This prompt is used to generate {n} multiple choice questions based on long text information {text}. The text can be a summary of a transcript, or excerpt form a book. The output is a JSON to allow integration with code or apps.

Prompt

Given the corpus of text, \
  generate {{n}} multiple choice questions\
  based on the contents of the text. The goal of the these questions is to \
  quiz the audience after who have read the text. Make sure to randomize \
  the order of the answers for each question and evenly distribute the correct \
  answer across the options. Each question should be different and not repeated. \
  Format the questions in JSON as follows, make sure to use double quotes:\n \
  {{\
    "questions": [\
      {{\
        "question": "Who did X?",\
        "options": [\
         "A) Answer 1",\
         "B) Answer 2",\
         "C) Answer 3",\
         "D) Answer 4"
        ],\
        "correct_answer": "C) Answer 3", \
        "explanation": "Explanation of the correct answer" \
      }},\
      // More questions...\
    ]\
  }}

The text delimited in triple backticks:
  ```{{text}}``` 

Welcome Message

{

"questions": [

{

  "question": "What is the main purpose of the Think GPT library?",

  "options": [

    "A) Enhancing large language models",

    "B) Storing external knowledge",

    "C) Performing code generation",

    "D) Enabling natural language conditions"

  ],

  "correct_answer": "A) Enhancing large language models",

  "explanation": "The Think GPT library aims to overcome the limitations of large language models (LLMs) and enhance their capabilities."

},

{

  "question": "Which feature of the Think GPT library allows users to fit large amounts of knowledge into the limited context size of the LLM?",

  "options": [

    "A) Long memory",

    "B) Self-refinement",

    "C) Compression",

    "D) Natural language conditions"

  ],

  "correct_answer": "C) Compression",

  "explanation": "The Think GPT library offers two methods for compression, samurai and summarize, to fit large amounts of knowledge into the limited context size of the LLM."

},

{

  "question": "What is the purpose of the LLM library?",

  "options": [

    "A) Summarization",

    "B) Self-refinement",

    "C) Code generation",

    "D) All of the above"

  ],

  "correct_answer": "D) All of the above",

  "explanation": "The LLM library provides functionalities such as summarization, self-refinement, and generative agents."

},

{

  "question": "What parameter can users specify when using LLM for summarization?",

  "options": [

    "A) Max tokens",

    "B) Target limit",

    "C) Chunk size",

    "D) Memory store"

  ],

  "correct_answer": "A) Max tokens",

  "explanation": "Users can specify parameters such as max tokens and target limit for summarization using LLM."

},

{

  "question": "What infrastructure is needed to get started with the library for implementing prompts?",

  "options": [

    "A) Vector database",

    "B) Cluster setup",

    "C) OpenAI API key",

    "D) Memory store"

  ],

  "correct_answer": "C) OpenAI API key",

  "explanation": "Users only need an OpenAI API key and the library installed to get started with the library for implementing prompts."

}

]

}

Conversation

ASSISTANT: {

"questions": [

{

  "question": "What is the main purpose of the Think GPT library?",

  "options": [

    "A) Enhancing large language models",

    "B) Storing external knowledge",

    "C) Performing code generation",

    "D) Enabling natural language conditions"

  ],

  "correct_answer": "A) Enhancing large language models",

  "explanation": "The Think GPT library aims to overcome the limitations of large language models (LLMs) and enhance their capabilities."

},

{

  "question": "Which feature of the Think GPT library allows users to fit large amounts of knowledge into the limited context size of the LLM?",

  "options": [

    "A) Long memory",

    "B) Self-refinement",

    "C) Compression",

    "D) Natural language conditions"

  ],

  "correct_answer": "C) Compression",

  "explanation": "The Think GPT library offers two methods for compression, samurai and summarize, to fit large amounts of knowledge into the limited context size of the LLM."

},

{

  "question": "What is the purpose of the LLM library?",

  "options": [

    "A) Summarization",

    "B) Self-refinement",

    "C) Code generation",

    "D) All of the above"

  ],

  "correct_answer": "D) All of the above",

  "explanation": "The LLM library provides functionalities such as summarization, self-refinement, and generative agents."

},

{

  "question": "What parameter can users specify when using LLM for summarization?",

  "options": [

    "A) Max tokens",

    "B) Target limit",

    "C) Chunk size",

    "D) Memory store"

  ],

  "correct_answer": "A) Max tokens",

  "explanation": "Users can specify parameters such as max tokens and target limit for summarization using LLM."

},

{

  "question": "What infrastructure is needed to get started with the library for implementing prompts?",

  "options": [

    "A) Vector database",

    "B) Cluster setup",

    "C) OpenAI API key",

    "D) Memory store"

  ],

  "correct_answer": "C) OpenAI API key",

  "explanation": "Users only need an OpenAI API key and the library installed to get started with the library for implementing prompts."

}

]

}