model_331570886_tiny_transformer_giant.py

Model Overview

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

Architecture

• Architecture: tiny transformer
• Scale: giant
• Attention: flash
• Fusion strategy: cross attention
• Task head: generation
• Activation: mish
• Normalization: layernorm
• Initialization: xavier uniform

Training

• Optimizer: lamb
• LR scheduler: exponential

Files

• model_331570886_tiny_transformer_giant.py — main artifact of this repository

License

See the license field above.