model_355416055_mobilevit_large.py

Model Overview

A large-scale implementation of the mobilevit architecture, built for classification tasks.

Architecture

• Architecture: mobilevit
• Scale: large
• Attention: sliding window
• Fusion strategy: co attention
• Task head: classification
• Activation: approx gelu
• Normalization: scalenorm
• Initialization: orthogonal

Training

• Optimizer: sgd
• LR scheduler: linear warmup

Files

• model_355416055_mobilevit_large.py — main artifact of this repository

License

See the license field above.