# Load model directly
from transformers import AutoProcessor, AutoModelForZeroShotImageClassification
processor = AutoProcessor.from_pretrained("dleemiller/siglip2-math-base-patch16-256")
model = AutoModelForZeroShotImageClassification.from_pretrained("dleemiller/siglip2-math-base-patch16-256")Quick Links
SigLip2 Math
This version of siglip2 is fine tuned on shiwk24/MathCanvas-Imagen using the code_derived_captions split.
I trained for 1 epoch on 4M math images, with a random selection between the tikz code or caption using a batch size of 640.
This is not a classification model, since the loss function was pairwise contrastive loss. Use for embedding or downstream classifier training is recommended.
- Downloads last month
- 6
Model tree for dleemiller/siglip2-math-base-patch16-256
Base model
google/siglip2-base-patch16-224
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("zero-shot-image-classification", model="dleemiller/siglip2-math-base-patch16-256") pipe( "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/hub/parrots.png", candidate_labels=["animals", "humans", "landscape"], )