# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("nzs234/siglip2-so400m-aesthetic-scorer-v1", dtype="auto")Quick Links
SigLIP2 Aesthetic Scorer (Local Bundle)
This repository contains a standalone local bundle for aesthetic scoring:
- Backbone:
google/siglip2-so400m-patch16-512(saved locally) - Head: custom MLP regressor (
head.safetensors) - Score range:
1..9
Files
backbone/: local SigLIP2 backbone + processor fileshead.safetensors: MLP head weightsmetadata.json: model config and score range
Output
The model outputs a continuous score, and common usage rounds it to integer score_1 ... score_9.
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-classification", model="nzs234/siglip2-so400m-aesthetic-scorer-v1") pipe("https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/hub/parrots.png")