Deploy Fast and Accurate YOLOv8 Object Detection Models on CPUs You Already Have
As machine learning models become larger and larger, it’s difficult to deploy them to production at desired cost and performance levels. Enter Neural Magic. At Neural Magic, we've developed cutting-edge sparsification techniques to address this issue.
We’ve recently applied our techniques to YOLOv8, a popular object detection model from Ultralytics. The result is a 65% pruned and INT8-quantized model that recovers to 99% of its baseline accuracy that can run on CPUs at GPU-class speeds.
Join our webinar on June 15, 2023, to learn how you can:
- Download already-optimized YOLOv8 models from the SparseZoo
- Apply your data with a few lines of code using SparseML
- Deploy YOLOv8 on commodity CPUs that you already own at GPU speeds and better.
- Shubhra Pandit, Senior ML Research Engineer
- Mark Kurtz, Director of ML, Neural Magic
Time: 1:00 PM EST; 10:00 AM PST
Was this article helpful?