Geo-GPT: Fine-Tuning ChatGPT for Advanced Geoscience Applications
Dates: 2024
Principal Investigator: Dr. Umair bin Waheed
Description: The proposed initiative, Geo-GPT, aims to leverage the success of domain-focused AI models, like Bloomberg's Finance-centric model and Google AI's Med-Palm2, in the geosciences domain. The project's dual objectives are to create a comprehensive database by training the Geo-GPT model with geological datasets and to enhance its algorithmic capabilities for tailored code generation in machine learning and image processing within geosciences. Geo-GPT addresses the need for nuanced, domain-specific insights lacking in existing large language models. Its strategy involves data acquisition, model retraining, and optimization. Milestones include data collection, model retraining, testing, documentation, and deployment as a user-friendly web application. By interfacing with Geo-GPT, stakeholders including geoscientists, researchers, and students can benefit from enhanced productivity and informed decision-making. In conclusion, Geo-GPT promises to usher in a new era in geosciences by providing domain-specific insights and advanced coding tools, thus enriching both academic and practical aspects of geology.