---
license: apache-2.0
new_version: nightknocker/contrastive-u-dae
---

# Contrastive Booru-Image Embedding Model

The model was trained on [masked](https://huggingface.co/nightknocker/vit-base-patch16-224-masked) pixiv images, tagged by [CL tagger](https://huggingface.co/cella110n/cl_tagger).
The caption (prompt) was tokenized by [ModernBERT](https://huggingface.co/answerdotai/ModernBERT-base).


## Objectives

*Image-Caption Similarity Scoring*: The model assigns a score to each image based on how closely it is related to its associated caption.

*Image Labeling*: The image embedding is used to label images by character name.

## Updates

This version contains up to 4200 characters extracted from public and private datasets.