model_291826171_coca_huge.py

Model Overview

A huge-scale implementation of the coca architecture, built for multitask tasks.

Architecture

• Architecture: coca
• Scale: huge
• Attention: grouped query
• Fusion strategy: tucker
• Task head: multitask
• Activation: approx gelu
• Normalization: rmsnorm
• Initialization: trunc normal

Training

• Optimizer: lamb
• LR scheduler: cosine

Files

• model_291826171_coca_huge.py — main artifact of this repository

License

See the license field above.