pchaberski Goto Github PK

cars

Car model recognition using lightweight neural net architectures intended for mobile applications. Training GhostNet on Stanford Cars Dataset with PyTorch.

gpu_tests

Two experiments (#1 GAN image generation; #2 image classification with Resnet; both in PyTorch) to test differences in training times between free Colab GPUs and mobile/desktop GPUs.

image-embeddings

Train and tune an autoencoder model and use it later to extract features (create embeddings) from images for further machine learning applications.

ink-detection

keras_and_shiny

A showcase based on the tutorial presented at ML@Enterprise Forum 2018 in Warsaw. Combining R (Shiny) and Python (Keras) to build a simple web application for image recognition.

losowanie-prezentow

quickstart-ml-blueprints

Data science project development best practices and state of the art open-source tooling forged into a set of solved ML use cases to serve as blueprints for efficient prototyping.

quickstart-ml-starter

Kedro starterts to quickly set up new projects according to QuickStart ML Blueprints practice.