Worked on the rasbt/llms-from-scratch repository, focusing on improving the clarity and accuracy of documentation within the Jupyter notebook finetuning workflow. Addressed a documentation bug by updating terminology from 'classification finetuning' to 'instruction finetuning', ensuring that the script’s purpose is communicated correctly to users. This change, implemented through Python in a Jupyter Notebook environment, reduced user confusion and streamlined onboarding for new contributors. Emphasized documentation best practices and careful version control, maintaining alignment between code functionality and supporting materials. The work demonstrated attention to detail in terminology governance and contributed to more maintainable and accessible project documentation.