model_601606437_beit_tiny.py

Model Overview

A tiny-scale implementation of the beit architecture, built for matching tasks.

Architecture

• Architecture: beit
• Scale: tiny
• Attention: sliding window
• Fusion strategy: co attention
• Task head: matching
• Activation: approx gelu
• Normalization: batchnorm
• Initialization: kaiming normal

Training

• Optimizer: adamw
• LR scheduler: constant warmup

Files

• model_601606437_beit_tiny.py — main artifact of this repository

License

See the license field above.