CECAM - NVIDIA Workshop: Introduction to Deep Learning, Autoencoders and PyTorchNVIDIA Workshop: Introduction to Deep Learning, Autoencoders and PyTorch

CECAM-IRL and CECAM-STFC (Daresbury Laboratory) organized NVIDIA Deep Learning Institute Workshop

The NVIDIA Deep Learning Institute (DLI) Introduction to Deep Learning course provides a hands-on overview of how deep learning works in computer vision and natural language processing. Participants learn the history and fundamentals of neural networks, how to build and train them from scratch, and how to use PyTorch. The course covers CNNs, data augmentation, and transfer learning to improve model performance and efficiency. Learners gain practical experience with real datasets, explore pre-trained models for faster deployment, and develop the confidence to design and train their own AI applications using state-of-the-art deep learning techniques and tools.

Registration is now closed.

Topics to Cover in Depth	Key Focus
What is AI, machine learning and deep learning	History of deep learning, the AI big bang and the fundamentals of how a neural network works. Introduction to training via back-propagation.
How to train a NN	How NN are trained, tuning, avoid overfitting and data augmentation techniques commonly used.
Convolutional NN and technique for stable training and transfer learning	What are convolutional NN, how to use them and combine together with other components. Batch normalization and drop-off techniques for better training and build deeper NN. Transfer learning from pretrained models.
Advanced NN	Introduction to advanced NN, like autoencoder and Generative Adversarial Networks (GANs). Applications of Autoencoders for model reduction.

More details can be found at https://learn.nvidia.com/courses/course-detail?course_id=course-v1:DLI+S-FX-01+V1 (free registration may be required).

Participation

Due to the hands-on nature of this workshop the number of places is strictly limited for both in person and online participation. Priority will be given to participants who are integrating deep learning into their research or thesis projects. A working knowledge of python coding is a prerequisite.

Only a laptop with browser will be needed, please read the following instructions to be ready: https://developer.nvidia.com/dli/getready.

NVIDIA Workshop: Introduction to Deep Learning, Autoencoders and PyTorch

Location: University College Dublin School of Physics, Beech-hill

Organisers

References