fbellame/llama2-pdf-to-quizz-13b
TEXT GENERATIONConcurrency Cost:1Model Size:13BQuant:FP8Ctx Length:4kPublished:Aug 26, 2023Architecture:Transformer0.0K Cold

The fbellame/llama2-pdf-to-quizz-13b is a 13 billion parameter Llama 2-based causal language model, fine-tuned by fbellame using H2O LLM Studio. It is specifically optimized for generating multiple-choice questions (MCQs) with four options and a corresponding answer letter from provided document text. This model excels at creating detailed, context-specific quizzes, making it ideal for educational content generation and assessment tools.

Loading preview...

Model Overview

The fbellame/llama2-pdf-to-quizz-13b is a 13 billion parameter language model built upon the meta-llama/Llama-2-13b-hf architecture. It was fine-tuned by fbellame using H2O LLM Studio, specifically for the task of generating multiple-choice questions (MCQs) from given text documents.

Key Capabilities

  • MCQ Generation: The model is trained to produce one multiple-choice question with four options (A, B, C, D) and a single-letter answer based on an input document.
  • Contextual Questioning: Questions are designed to be detailed and derived solely from the information provided within the document, ensuring relevance and accuracy.
  • Structured Output: It follows a specific output format for questions, choices, and answers, facilitating automated parsing and integration.

Training Details

The model was trained on a dataset of 168 prompts, each designed to generate a multiple-choice question and answer from a document. This specialized training dataset, available as fbellame/pdf_to_quizz_llama_13B on Hugging Face, focuses on the precise task of quiz creation.

Good For

  • Educational Content Creation: Automating the generation of quizzes and assessment materials from textbooks, articles, or other documents.
  • Knowledge Assessment: Quickly creating questions to test comprehension of specific textual content.
  • Developers: Integrating into applications requiring automated question generation from PDFs or other text sources, particularly for educational technology platforms.