This workshop will introduce participants to Generative Adversarial Networks (GANs). We will demonstrate the core techniques of GANs, including how to use Deep Convolutional GANs (DCGANs) to generate images using PyTorch. A basic understanding of the topics covered in previous sessions is assumed. Having working experience with Python, Jupyter Notebooks, and linear algebra will be…

Read More

# Python

## Learning Convolutional Neural Networks (2)

This workshop will be the second lecture in our introduction to convolutional neural networks (CNNs). We will continue our learning by applying data augmentation and transfer learning techniques to improve our solution for the classic Dogs-vs-Cats problem using PyTorch. A basic understanding of the topics covered in previous sessions is assumed. Having working experience with…

Read More

## Code Profilers: Measuring Performance and Resource Usage

Learn to use code profilers (software) to analyze a program’s performance and resource usage, and to use the measurement data to guide tuning methodologies and workflows, in the context of scientific computing. Any questions about this workshop can be emailed to Shao-Ching Huang at sch@ucla.edu. Presented by the Office of Advanced Research Computing (OARC). Register here:…

Read More

## Learning Convolutional Neural Networks (1)

This workshop will provide an introduction to convolutional neural networks (CNNs). We will begin by using PyTorch to perform image processing on the classic Dogs-vs-Cats problem. Basic knowledge of the topics covered in previous sessions is assumed. Having working experience with Python, Jupyter Notebooks, and linear algebra will be helpful for fully participating in the…

Read More

## Learning PyTorch

We will give a general introduction to PyTorch, a popular deep learning framework, with practical illustrations on the primary usage of tensors and automatic differentiation, and on solving a simple temperature-conversion problem using PyTorch. The knowledge of topics covered in the previous session about machine/deep learning is assumed. Working experience with Python and Jupyter Notebooks…

Read More

## Big Data on HPC

This workshop will go over using Big Data techniques on HPC resources. Big Data methods are used when data size because so large, it becomes challenging to compute. Also, when machine learning models become so complex, it can also be challenging to train. In this workshop, we will go over examples of solving Big Data…

Read More

## High Performance Machine Learning Using Scikit-Learn

As machine learning gains more and more popularity in science and technology in recent years, scikit-learn becomes one of the must-have libraries in the general machine learning toolbox. In this lecture we will discuss some advanced topic on using scikit-learn python library to make high performance machine learning, specifically the speedup modeling using multicore and…

Read More

## Learning Scikit-Learn

As machine learning gains more and more popularity in science and technology in recent years, scikit-learn becomes one of the must-have libraries in the general machine learning toolbox. In this lecture we will present an introduction about the basics of scikit-learn python library. Prerequisite knowledge for the workshop includes Python programming and basic machine learning…

Read More

## Hoffman2 Happy Hour: Using Anaconda on Hoffman2

Anaconda (https://www.anaconda.com/) is a distribution of R and Python that can be used to easily install many popular data science, biostats, and other packages. This Hoffman2 Happy Hour will go discus using Anaconda on Hoffman2. This can be applied to using Anaconda on various other HPC resources. We will go over creating conda environments and…

Read More

## Boosting Python for High Performance Data Analytics (2) DataFrame Game

While Python becomes the most popular programming language since 2019, data scientists often have a few common complaints about its slow speed and the limited capabilities of handling the big data scenarios. In this workshop series, we will present an extensive discussion on how to improve the performance of Python in data science by looking…

Read More