Develop a website for creating new Harry Potter sentences based on all 7 previous books in the series using an LLM model.
- Dataset Selection: Finding a proper dataset that emulates all 7 Harry Potter Books.
- Data Concatenation: Separate each sentence line by line to create a clean dataset.
- Front End: Creating a website to visualize the output.
- Server: Creating a server to connect all components together.
- Integration: Combine the LLM and front end into a single platform where users can receive LLM outputs directly from a click of a button.
- Testing and Feedback: Test the system with end-users, collect feedback, and refine the tool and dataset accordingly.
- Machine Learning and NLP: Knowledge of LLMs, and natural language understanding.
- Data Annotation: Experience in creating guidelines and annotating data for AI models.
- Software Development: Proficiency in programming language Rust, especially in data processing and visualization tool development.
- Research and Analysis: Ability to conduct systematic studies and analyze data for meaningful conclusions.
This project aims to advance AI explainability, providing resources for the research community to enhance understanding and improvement of LLM decision-making processes.