model.py
Model Overview
A huge-scale implementation of the deit architecture, built for multitask tasks.
Architecture
- Architecture: deit
- Scale: huge
- Attention: sliding window
- Fusion strategy: bilinear
- Task head: multitask
- Activation: swish
- Normalization: layernorm
- Initialization: xavier uniform
Training
- Optimizer: lamb
- LR scheduler: onecycle
Files
model.py — main artifact of this repository
License
See the license field above.