This repository contains my work from Google Summer of Code 2025 with OpenVINO, where I developed a modular training and deployment package to optimize machine learning workflows on consumer-grade CPUs.
The project includes:
Integration of PyTorch for training and inference acceleration.
Extended support for scikit-learn classifiers (e.g., Logistic Regression, Random Forest, SVC) with OpenVINO-based inference and export to IR.
TensorFlow integration with OpenVINO for improved performance in training and deployment.
Quantization and performance optimization to reduce latency, memory usage, and increase throughput.
The goal of this project is to simplify high-performance ML deployment and provide tools that make model optimization more accessible to developers and researchers.
👉 You can also read the official blog post about this project here: 'https://medium.com/openvino-toolkit/my-google-summer-of-code-journey-with-openvino-ai-pc-model-training-kit-7dd5a9b436b0'