Intellectual Property India Publishes Patent Application for 'System And Method For Configuration-Driven Fine-Tuning And Deployment Of Machine Learning Models' Filed by Alpesh Prakash Sonar; and Divyani Alpesh Sonar

MUMBAI, India, Jan. 9 -- Intellectual Property India has published a patent application (202521108356 A) filed by Alpesh Prakash Sonar; and Divyani Alpesh Sonar, Thane, Maharashtra, on Nov. 8, 2025, for 'system and method for configuration-driven fine-tuning and deployment of machine learning models.'

Inventor(s) include Alpesh Prakash Sonar; and Divyani Alpesh Sonar.

The application for the patent was published on Dec. 12, under issue no. 50/2025.

According to the abstract released by the Intellectual Property India: "The present invention relates to a retraining system (100) and method (202-232) for configuration-driven fine-tuning of private or pre-trained machine learning models. An electronic device (102) with a device processor (104) transmits a model file in supported formats, including PyTorch, TensorFlow/Keras, ONNX, or Hugging Face transformers, through a user interface (106). A server (106) with a server processor (108) receives (204) the model, analyzes (206) architecture type, hierarchical layers, parameter counts, and dependency mappings, and converts (208) it into a unified internal representation. Schema validation (210) and automated preprocessing (212) ensure dataset compatibility for classification, regression, or generative tasks. User selections (214) for training duration, optimization parameters, and layer freezing guide selective parameter updates (218) while freezing non-selected layers (220) and applying differential learning rates (222). Fine-tuning strategies (224), metrics tracking (226), encryption (228), and audit trails (230) enable secure, non-technical retraining (232) of models efficiently across diverse domains."

Disclaimer: Curated by HT Syndication.