finetune.py
Model Overview
A base-scale implementation of the swin t architecture, built for contrastive tasks.
Architecture
- Architecture: swin t
- Scale: base
- Attention: dilated
- Fusion strategy: bilinear
- Task head: contrastive
- Activation: mish
- Normalization: instancenorm
- Initialization: kaiming
Training
- Optimizer: lamb
- LR scheduler: constant warmup
Files
finetune.py — main artifact of this repository
License
See the license field above.
