Our two-day course on high-performance machine learning provides the necessary skills to train neural networks and extract the most relevant information from datasets. During our hands-on sessions you will have the opportunity to work on our high-performance systems with different types of data, and learn how to tune your model to obtain optimal results in the most efficient way. Curious about what you’re going to learn?
Uncover the essence of machine learning theories and algorithmic insights through hands-on exploration. For example, work with a high-level machine learning API (Keras) or explore hyperparameter space to improve a neural network. Additionally, dive into fully connected networks, convolutional networks and autoencoders (time permitting).
Program day 1:
- Introduction to Deep Learning
- Using the PyTorch framework
- Fully connected networks, Convolutional networks, Autoencoders (time permitting)
Learn how to set up your (preinstalled) software environment and overcome file limitations that can slow down your training speed. Grasp the technical capabilities of modern day CPUs and GPUs and how to use multiple CPUs or GPUs in a single training. Finally, discover how to find bottlenecks in your code through creating a (PyTorch) profile.
Program day 2:
- Software installations on HPC systems
- Packed file formats for Machine Learning
- Parallel computing for deep learning
- Hardware (e.g. Tensor cores) and software features (e.g. low level libraries for deep learning) to accelerated deep learning
- Profiling TensorFlow with TensorBoard
Everyone interested in getting familiar with machine learning at scale, from the basics up to more advanced topics.
You can apply for this course if you have:
- Basic knowledge on statistics;
- Basic knowledge on linear algebra;
- Basic knowledge on Python programming;
some experience with the use of Jupyter Notebooks is desirable, but not essential;
- Basic knowledge on parallel computing is helpful, but not required.