model_110449132_perceiver_giant.py

Model Overview

A giant-scale implementation of the perceiver architecture, built for multitask tasks.

Architecture

• Architecture: perceiver
• Scale: giant
• Attention: grouped query
• Fusion strategy: co attention
• Task head: multitask
• Activation: gelu
• Normalization: groupnorm
• Initialization: xavier

Training

• Optimizer: rmsprop
• LR scheduler: exponential

Files

• model_110449132_perceiver_giant.py — main artifact of this repository

License

See the license field above.