GRADIO AND HUGGING CAPABILITIES FOR DEVELOPING RESEARCH AI APPLICATIONS

Recent advancements in neural network training have contributed to significant progress in computer vision, natural language processing, and data analysis. AI modules are used in education, medicine, security, autonomous systems, etc. However, their implementation requires rapid hypothesis testing, optimal solution selection, dataset collection, expert evaluation, etc. For example, in the work [2,6], it was necessary to conduct an expert evaluation before forming conclusions and implementing models in real applications. This paper considers the issue of creating and using prototypes of AI applications for research and proposes to use Gradio and Hugging Face to speed up this process.