BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Drupal//recurring_events_ical//2.0//EN
BEGIN:VEVENT
UID:2e8dee12-9551-4bb1-88a5-940bf3469d9f@support.access-ci.org
DTSTAMP:20251009T085527Z
DTSTART:20251010T140000Z
DTEND:20251010T150000Z
SUMMARY:Intro to PyTorch & TensorFlow
DESCRIPTION:This session provides an introduction to two of the most widely
  used deep learning frameworks, PyTorch and TensorFlow. We will begin wit
 h a comparison of their core architectures and programming styles, highlig
 hting how PyTorch’s dynamic computation graph differs from TensorFlow’
 s static graph approach. Next, we will cover the fundamentals of tensors 
 as the building blocks for both frameworks and explore how PyTorch and Ten
 sorFlow handle them. We will then walk through the basics of automatic di
 fferentiation (autograd) and demonstrate how it enables gradient-based opt
 imization for training models. Finally, we will build and train simple ne
 ural networks in both frameworks, showing the step-by-step process from de
 fining layers to running training loops. The session combines conceptual 
 explanations with hands-on code examples in both a slideshow and a Jupyter
  notebook, which will also be shared with participants afterward for furth
 er practice.
URL:https://support.access-ci.org/events/8564
END:VEVENT
END:VCALENDAR