model_243472682_albef_large.py

Model Overview

A large-scale implementation of the albef architecture, built for retrieval tasks.

Architecture

• Architecture: albef
• Scale: large
• Attention: flash
• Fusion strategy: concat mlp
• Task head: retrieval
• Activation: swish
• Normalization: layernorm
• Initialization: xavier

Training

• Optimizer: lion
• LR scheduler: constant warmup

Files

• model_243472682_albef_large.py — main artifact of this repository

License

See the license field above.