luffycodes/tutorbot-spock-bio-llama-diff
luffycodes/tutorbot-spock-bio-llama-diff is a 13 billion parameter LLaMA-based "diff model" developed by the Spock team from Rice University and OpenStax. This model is specifically fine-tuned to function as an educational tutoring chatbot, trained on synthetic student-tutorbot conversations. It is designed to be applied on top of original LLaMA weights to create the Spock model, excelling in educational dialogue generation.
Loading preview...
Spock: An Educational Tutoring Chatbot
Spock is an open-source educational tutoring chatbot developed by the Spock team, comprising members from Rice University and OpenStax. This model is a "diff model" based on the LLaMA and Vicuna architectures, meaning it must be applied to the original LLaMA weights to obtain the full Spock model. It was trained between April and May 2023.
Key Capabilities
- Specialized Tutoring Dialogue: Fine-tuned on 700 synthetic student-tutorbot conversations generated using a specialized prompt from GPT-4.
- Educational Focus: Designed to embody learning science principles for effective tutoring interactions.
- Research-Backed: The development and principles are detailed in the paper "CLASS Meet SPOCK: An Education Tutoring Chatbot based on Learning Science Principles" (arXiv:2305.13272).
Good For
- Researchers and developers interested in building or experimenting with AI-powered educational tutors.
- Applications requiring conversational agents specifically tailored for teaching and learning scenarios.
- Exploring fine-tuning techniques for creating domain-specific chatbots from foundational LLMs.