Explore automated hyperparameter tuning for deep neural networks in this comprehensive 56-minute video tutorial. Learn how to optimize a PyTorch neural network using Optuna, covering the entire process from problem selection to finding optimal parameters. Dive into creating a dataset class, implementing cross-validation folds, and building the model and engine. Follow along as the instructor demonstrates Optuna's capabilities in tuning layers, dropout rates, learning rates, and other crucial parameters. Gain practical insights using a real-world dataset from a Kaggle competition. Perfect for machine learning enthusiasts looking to streamline their deep learning workflows and improve model performance through automated hyperparameter optimization.
Automated Hyperparameter Tuning for Deep Neural Networks
Overview
Syllabus
Introduction
Dataset class
Start with train.py
Cross-validation folds
Reading the data
Engine
Model
Add model and engine to training
Optuna
Start tuning with Optuna
Training, suggestions and outro
Taught by
Abhishek Thakur
