model_278235102_coca_base.py

Model Overview

A base-scale implementation of the coca architecture, built for retrieval tasks.

Architecture

• Architecture: coca
• Scale: base
• Attention: dilated
• Fusion strategy: co attention
• Task head: retrieval
• Activation: relu
• Normalization: groupnorm
• Initialization: kaiming

Training

• Optimizer: rmsprop
• LR scheduler: constant warmup

Files

• model_278235102_coca_base.py — main artifact of this repository

License

See the license field above.