Module mobileclip 

Mobile CLIP model, combining a lightweight vision encoder with a text encoder

A mobile-optimized CLIP implementation that uses:

• FastViT as the vision encoder
• OpenCLIP text encoder
• Projection layers to align the feature spaces

See model details at:

• FastViT
• OpenCLIP

References:

• MobileVLM
• MetaCLIP

Structs

MobileClipConfig

MobileClipModel

Functions

div_l2_norm