model_283765557_coca_xlarge.py

Model Overview

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

Architecture

• Architecture: coca
• Scale: xlarge
• Attention: standard
• Fusion strategy: bilinear
• Task head: multitask
• Activation: swish
• Normalization: batchnorm
• Initialization: xavier uniform

Training

• Optimizer: adamw
• LR scheduler: constant warmup

Files

• model_283765557_coca_xlarge.py — main artifact of this repository

License

See the license field above.