model_230090610_albef_tiny.py

Model Overview

A tiny-scale implementation of the albef architecture, built for contrastive tasks.

Architecture

• Architecture: albef
• Scale: tiny
• Attention: sliding window
• Fusion strategy: concat mlp
• Task head: contrastive
• Activation: gelu
• Normalization: batchnorm
• Initialization: xavier

Training

• Optimizer: rmsprop
• LR scheduler: linear warmup

Files

• model_230090610_albef_tiny.py — main artifact of this repository

License

See the license field above.