--- license: cc-by-4.0 tags: - adamw - batchnorm - exponential - flash - generation - large - mish - perceiver - tucker - xavier --- # model_094972195_perceiver_large.py ## Model Overview A **large**-scale implementation of the **perceiver** architecture, built for **generation** tasks. ## Architecture - **Architecture**: perceiver - **Scale**: large - **Attention**: flash - **Fusion strategy**: tucker - **Task head**: generation - **Activation**: mish - **Normalization**: batchnorm - **Initialization**: xavier ## Training - **Optimizer**: adamw - **LR scheduler**: exponential ## Files - `model_094972195_perceiver_large.py` — main artifact of this repository ## License See the license field above.