model_188733365_clip_nano.py

Model Overview

A nano-scale implementation of the clip architecture, built for contrastive tasks.

Architecture

• Architecture: clip
• Scale: nano
• Attention: multi query
• Fusion strategy: cross attention
• Task head: contrastive
• Activation: approx gelu
• Normalization: batchnorm
• Initialization: xavier uniform

Training

• Optimizer: lamb
• LR scheduler: step

Files

• model_188733365_clip_nano.py — main artifact of this repository

License

See the license field above.