BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Drupal//recurring_events_ical//2.0//EN
BEGIN:VEVENT
UID:f0da91a6-210a-4a6c-b4d5-11a88ad34113@support.access-ci.org
DTSTAMP:20250401T185031Z
DTSTART:20250422T170000Z
DTEND:20250422T190000Z
SUMMARY:CI Pathways: Multi-GPU Training with Accelerate
DESCRIPTION:Accelerate, HuggingFace's library for efficient multi-GPU train
 ing enables practitioners to scale their AI/ML applications effectively. T
 his course will explore advanced training techniques for deep learning mod
 els, specifically targeting the ResNet50 architecture. Participants will g
 ain a comprehensive understanding of the architecture and the significance
  of residual connections in deep neural networks. Additionally, the course
  will cover scheduling strategies to optimize learning rates during traini
 ng, Weights & Biases (W&B) use for tracking metrics and experiments, and m
 ethods for saving intermittent checkpoints and resuming training.
URL:https://support.access-ci.org/events/7907
END:VEVENT
END:VCALENDAR