Python solutions to Deep-ML practice problems — explained, implemented, and benchmarked across NumPy, PyTorch, CUDA & Tinygrad.
-
Updated
Aug 20, 2026 - Python
Python solutions to Deep-ML practice problems — explained, implemented, and benchmarked across NumPy, PyTorch, CUDA & Tinygrad.
A deep neural network built with TensorFlow/Keras achieving 98.44% test accuracy on MNIST handwritten digit classification using Batch Normalization, Dropout regularization, and adaptive training callbacks.
AI/ML career-transition journal web app with Firebase Auth, Firestore, and Gemini multi-turn coaching, deployed on Cloud Run via Google AI Studio.
A FastAPI web application for customer segmentation using Machine Learning (KMeans Clustering)
To associate your repository with the ai-ml-engineering topic, visit your repo's landing page and select "manage topics."