yemiadeleke's picture
Upload 2 files
f649c79 verified
|
Raw
History Blame Contribute Delete
755 Bytes
metadata
license: mit
tags:
  - adafactor
  - classification
  - clip
  - layernorm
  - linear
  - low-rank
  - onecycle
  - small
  - swish
  - xavier

model_196177330_clip_small.py

Model Overview

A small-scale implementation of the clip architecture, built for classification tasks.

Architecture

  • Architecture: clip
  • Scale: small
  • Attention: linear
  • Fusion strategy: low rank
  • Task head: classification
  • Activation: swish
  • Normalization: layernorm
  • Initialization: xavier

Training

  • Optimizer: adafactor
  • LR scheduler: onecycle

Files

  • model_196177330_clip_small.py — main artifact of this repository

License

See the license field above.