model_338442762_mobilevit_nano.py

Model Overview

A nano-scale implementation of the mobilevit architecture, built for matching tasks.

Architecture

• Architecture: mobilevit
• Scale: nano
• Attention: grouped query
• Fusion strategy: low rank
• Task head: matching
• Activation: gelu tanh
• Normalization: layernorm
• Initialization: kaiming

Training

• Optimizer: novograd
• LR scheduler: cosine

Files

• model_338442762_mobilevit_nano.py — main artifact of this repository

License

See the license field above.