model_626338886_beit_huge.py

Model Overview

A huge-scale implementation of the beit architecture, built for classification tasks.

Architecture

• Architecture: beit
• Scale: huge
• Attention: grouped query
• Fusion strategy: tensor fusion
• Task head: classification
• Activation: mish
• Normalization: instancenorm
• Initialization: kaiming normal

Training

• Optimizer: rmsprop
• LR scheduler: constant warmup

Files

• model_626338886_beit_huge.py — main artifact of this repository

License

See the license field above.