model_018751596_vit_large.py
Model Overview
A large-scale implementation of the vit architecture, built for contrastive tasks.
Architecture
- Architecture: vit
- Scale: large
- Attention: standard
- Fusion strategy: concat mlp
- Task head: contrastive
- Activation: relu
- Normalization: rmsnorm
- Initialization: kaiming
Training
- Optimizer: rmsprop
- LR scheduler: exponential
Files
model_018751596_vit_large.py — main artifact of this repository
License
See the license field above.