model_629789311_beit_giant.py

Model Overview

A giant-scale implementation of the beit architecture, built for generation tasks.

Architecture

• Architecture: beit
• Scale: giant
• Attention: sliding window
• Fusion strategy: bilinear
• Task head: generation
• Activation: approx gelu
• Normalization: scalenorm
• Initialization: kaiming

Training

• Optimizer: lamb
• LR scheduler: polynomial

Files

• model_629789311_beit_giant.py — main artifact of this repository

License

See the license field above.