ML Samples

Machine learning and data science samples in Python — from classical algorithms to modern deep learning.

Setup

Requires uv and Python 3.13+.

uv sync

For specific sample areas with extra dependencies:

uv sync --extra whisper    # OpenAI Whisper transcription
uv sync --extra gcloud     # Google Cloud Vertex AI
uv sync --extra genai      # OpenAI API

Samples

Classical ML

Directory	Description
Linear_regression	1D linear regression
Logistic_regression	Logistic regression (ecommerce, image classification)
Multidimensional_regression	Polynomial fitting, L1/L2 regularization
Decision_trees	Decision trees, information gain, entropy
Naive_bayes_mnist	Naive Bayes on MNIST
Perceptron	Binary linear classifier on MNIST
Regressions_sklearn	KNN and decision tree regression (scikit-learn)

Neural Networks

Directory	Description
Feedforward_neural_network	Feedforward NN, backpropagation, softmax
Optimizations	Momentum, RMSprop, Adam optimizers
Other_samples	Gradient check, regularization, NN with 1 hidden layer

Deep Learning (TensorFlow / Keras)

Directory	Description
Tensorflow_basics	TF2 basics and sign language recognition app
Tensorflow_distributed	Distributed training with tf.distribute.MirroredStrategy
tf2	TF2 ResNet on CIFAR-10
tf_book	TensorFlow book samples
Generative_DL	Generative deep learning (CIFAR-10, Gemini, OpenAI)

Deep Learning (PyTorch)

Directory	Description
Pytorch_basics	PyTorch fundamentals, 2-layer NN on MNIST
Pytorch	PyTorch CUDA tensor operations
Convolutional_Neural_Networks	CNN samples (PyTorch, TF2 ResNet)

Tools and Integrations

Directory	Description
whisper	OpenAI Whisper audio transcription with Minio upload
gcloud	Google Cloud Vertex AI function calling
Tools	Model converter utilities
learn	Async Python examples