ModelSigLIP2 SO400M / SigLIP
SigLIP2
A vision-language model released by Google DeepMind in 2024 that replaces the conventional softmax loss with a sigmoid loss and introduces NaFlex variable-resolution input technology, achieving strong performance on image-text contrastive learning benchmarks.
Timeline (last 90 days)
Oct 12
d1-omni-600M 包含 381M 共享主干加决策头、94M SigLIP2 视觉编码器（来自 LFM2.5-VL-450M）和 112M 17 层 FastConformer 音频编码器。
Unverified50%
All Facts (1)
Source Articles
Liquid d1-omni-600M：树莓派上的多模态决策模型实战Oct 12
DINOv2 vs SigLIP: Why the Huge Gap in k-NN Classification? A Guide to Avoiding Vision Encoder Selection PitfallsJul 10