SageAttention
[ICLR2025, ICML2025, NeurIPS2025 Spotlight] Quantized Attention achieves speedup of 2-5x compared to FlashAttention, without losing end-to-end metrics across language, image, and video models.
- ATK
- 4732
- DEF
- 3701
Common
thu-ml/SageAttention
Common · Base · Genesis[ICLR2025, ICML2025, NeurIPS2025 Spotlight] Quantized Attention achieves speedup of 2-5x compared to FlashAttention, without losing end-to-end metrics across language, image, and video models.
Attack4,732
Defense3,701
Drag the card to turn it. Use the flip button to see the back.
Open card page