raytune

Recommendation Model Implementation by using PyTorch

Research on Autonomous Driving Based on the Carla Simulation Environment

This repository implements a Bayesian Optimization workflow for hyperparameter tuning using the Ray Tune framework and ConfigSpace for configuration management.

We use the Titanic dataset to implement machine learning and deep learning. Preprocessing data, visualizing, building models, and ensembling are practiced in the ML section; PyTorch basics, PyTorchLightning framework, and RayTune hyperparameter-tuning are in the DL section.

Under this problem setup, the goal is to leverage existing configuration data to make the overall optimization process as efficient as possible.

This project is part of my Master's thesis for the MSc in Data Science with Artificial Intelligence program at the University of Exeter. It implements a multi-task deep learning model for simultaneous driver identification and transport mode classification using smartphone sensor data from the SHL preview dataset.

Python wrapper for starting distributed Ray clusters with LSF job submission

Examples scripts demonstrating raytune and comet-ml usage.

Simple example of hyperparameter optimization using Raytune (with PyTorch ignite)