model_250345822_albef_xlarge.py
Model Overview
A xlarge-scale implementation of the albef architecture, built for contrastive tasks.
Architecture
- Architecture: albef
- Scale: xlarge
- Attention: dilated
- Fusion strategy: co attention
- Task head: contrastive
- Activation: mish
- Normalization: scalenorm
- Initialization: kaiming
Training
- Optimizer: adamw
- LR scheduler: onecycle
Files
model_250345822_albef_xlarge.py — main artifact of this repository
License
See the license field above.
