walker8096's picture
Upload 2 files
4426813 verified
|
Raw
History Blame Contribute Delete
779 Bytes
---
license: cc-by-4.0
tags:
- constant-warmup
- deit
- generation
- instancenorm
- kaiming
- lamb
- multi-query
- relu
- small
- tensor-fusion
---
# model_570051144_deit_small.py
## Model Overview
A **small**-scale implementation of the **deit** architecture, built for **generation** tasks.
## Architecture
- **Architecture**: deit
- **Scale**: small
- **Attention**: multi query
- **Fusion strategy**: tensor fusion
- **Task head**: generation
- **Activation**: relu
- **Normalization**: instancenorm
- **Initialization**: kaiming
## Training
- **Optimizer**: lamb
- **LR scheduler**: constant warmup
## Files
- `model_570051144_deit_small.py` — main artifact of this repository
## License
See the license field above.