VideoRoPE: What Makes for Good Video Rotary Position Embedding? Paper • 2502.05173 • Published 4 days ago • 57
InternLM-XComposer2.5-OmniLive: A Comprehensive Multimodal System for Long-term Streaming Video and Audio Interactions Paper • 2412.09596 • Published Dec 12, 2024 • 94
All the ImageNets Collection Noteworthy instances of ImageNet on the Hub. Vetted and tested with timm train and validation scripts. • 8 items • Updated Nov 21, 2024 • 6
Addition is All You Need for Energy-efficient Language Models Paper • 2410.00907 • Published Oct 1, 2024 • 145