model_161107712_blip_small.py

Model Overview

A small-scale implementation of the blip architecture, built for multitask tasks.

Architecture

• Architecture: blip
• Scale: small
• Attention: multi query
• Fusion strategy: gated fusion
• Task head: multitask
• Activation: gelu
• Normalization: rmsnorm
• Initialization: orthogonal

Training

• Optimizer: lion
• LR scheduler: exponential

Files

• model_161107712_blip_small.py — main artifact of this repository

License

See the license field above.