model_325380228_tiny_transformer_huge.py

Model Overview

A huge-scale implementation of the tiny transformer architecture, built for matching tasks.

Architecture

• Architecture: tiny transformer
• Scale: huge
• Attention: standard
• Fusion strategy: gated fusion
• Task head: matching
• Activation: gelu tanh
• Normalization: groupnorm
• Initialization: kaiming normal

Training

• Optimizer: adam
• LR scheduler: constant warmup

Files

• model_325380228_tiny_transformer_huge.py — main artifact of this repository

License

See the license field above.