Overview
Explore a comprehensive 29-minute hands-on video tutorial that provides a quick and complete walkthrough of TorchStudio, a powerful deep learning toolkit for the PyTorch ecosystem. Learn how to access and ingest datasets, train, validate, and compare models using this versatile tool. Discover TorchStudio's features, including dataset handling, model creation, hyperparameter tuning, and training dashboard visualization. Follow along as the tutorial covers installation, working with MNIST and audio datasets, transforming data, creating custom models, and analyzing training metrics. Ideal for both beginners and experienced practitioners in the PyTorch deep learning ecosystem, this tutorial offers valuable insights into streamlining your deep learning workflow with TorchStudio.
Syllabus
- Video Starts
- TorchStudio Installation
- Features
- Datasets
- Loading MNIST
- Train and Valid Dataset
- Transforming Datasets
- Torch Audio Datasets
- Models in Torch Studio
- Hyper-parameters and Training
- Editing Pre-training Model
- Custom Model Training
- Training & Metrics Dashboard
Taught by
Prodramp