| ---
|
| license: mit
|
| tags:
|
| - constant-warmup
|
| - giant
|
| - kaiming
|
| - lamb
|
| - layernorm
|
| - linear
|
| - mixer
|
| - multitask
|
| - swish
|
| - tucker
|
| ---
|
|
|
| # inference.py
|
|
|
| ## Model Overview
|
|
|
| A **giant**-scale implementation of the **mixer** architecture, built for **multitask** tasks.
|
|
|
| ## Architecture
|
|
|
| - **Architecture**: mixer
|
| - **Scale**: giant
|
| - **Attention**: linear
|
| - **Fusion strategy**: tucker
|
| - **Task head**: multitask
|
| - **Activation**: swish
|
| - **Normalization**: layernorm
|
| - **Initialization**: kaiming
|
|
|
| ## Training
|
|
|
| - **Optimizer**: lamb
|
| - **LR scheduler**: constant warmup
|
|
|
| ## Files
|
|
|
| - `inference.py` — main artifact of this repository
|
|
|
| ## License
|
|
|
| See the license field above.
|
|
|
