Native Sparse Attention: Hardware-Aligned and Natively Trainable Sparse Attention Paper • 2502.11089 • Published 12 days ago • 134
Native Sparse Attention: Hardware-Aligned and Natively Trainable Sparse Attention Paper • 2502.11089 • Published 12 days ago • 134
TripoSG: High-Fidelity 3D Shape Synthesis using Large-Scale Rectified Flow Models Paper • 2502.06608 • Published 18 days ago • 32
Running on Zero 408 408 Chat with DeepSeek-VL2-small 🌍 Generate responses using images and text input