Machine Learning and LMsCustom GPTs
Using GPT-3.5 and GPT-4 in order to create custom GPTs that are trained on specific datasets in order to solve problems tailored directly to the client's needs.
Mark: A custom markdown generator that is trained on a dataset of personal documentation in a markdown format. Designed to help users automate the process of documentation for any project or repository, and to convert those markdown files into HTML files.
George: A custom GPT-4 model that is trained on a dataset of rubber ducky scripts. Since ChatGPT is designed to think Rubber Ducky script is for malicious purposes, any available GPT model takes some prompt engineering in order to get proper rubber ducky scripts. None the less,
George can handle it since it was trained on a dataset of rubber ducky scripts designed for automating tasks, thus the ducky script that is not against its terms of service.
Note: Mark is personalized and tailored to my personal work.
Languages: Python, Markdown, HTML, Ducky.
Technology: GPT-3.5, GPT-4, Github, Openai