metadata
license: cc-by-4.0
tags:
- cross-attention
- gelu
- groupnorm
- huge
- kaiming
- mobilevit
- polynomial
- retrieval
- sgd
- sparse
finetune.py
Model Overview
A huge-scale implementation of the mobilevit architecture, built for retrieval tasks.
Architecture
- Architecture: mobilevit
- Scale: huge
- Attention: sparse
- Fusion strategy: cross attention
- Task head: retrieval
- Activation: gelu
- Normalization: groupnorm
- Initialization: kaiming
Training
- Optimizer: sgd
- LR scheduler: polynomial
Files
finetune.py— main artifact of this repository
License
See the license field above.