Pruning AI Models for Peak Performance - NVIDIA DRIVE Labs Ep. 31

Check out HALP (Hardware-Aware Latency Pruning), a new method designed to adapt convolutional neural networks (CNNs) and #transformer-based architectures for real-time performance. HALP optimizes pre-trained models to maximize compute utilization. In testing with NVIDIA DRIVE Orin™ on the road, it consistently outperformed alternative approaches.

00:00:00 - Introducing Hardware-Aware Latency Pruning (HALP)
00:00:29 - Common Model Optimization
00:00:59 - DNN Pruning
00:01:21 - Hardware Aware Latency Pruning
00:01:31 - Classification Tasks
00:01:37 - 3D Object Detection
00:02:04 - HALP with Transformers …

adapt ai models architectures check cnns compute convolutional neural networks drive drive labs hardware labs latency networks neural networks nvidia peak performance pre-trained models pruning real-time testing transformer