model_611672312_beit_base.py
Model Overview
A base-scale implementation of the beit architecture, built for classification tasks.
Architecture
- Architecture: beit
- Scale: base
- Attention: multi query
- Fusion strategy: tucker
- Task head: classification
- Activation: gelu
- Normalization: rmsnorm
- Initialization: kaiming normal
Training
- Optimizer: novograd
- LR scheduler: onecycle
Files
model_611672312_beit_base.py — main artifact of this repository
License
See the license field above.