This model is a 48.97% smaller version of google/siglip2-base-patch16-512 optimized for Bengali language via vocabulary size reduction using the trimming method.
This trimmed model should perform similarly to the original model with only 16,384 tokens and a much smaller memory footprint. However, it may not perform well for other languages as tokens not commonly used in the selected languages were removed from the vocabulary.

Model Statistics

Metric	Original	Trimmed	Reduction
Vocabulary size	256,000 tokens	16,384 tokens	93.60%
Model size	375,823,874 params	191,798,786 params	48.97%

Mining Dataset Statistics

Number of texts used for mining: 200,000 texts
Dataset: lbourdois/fineweb-2-trimming

Usage

Transformers (zero-shot image classification)

python

from transformers import pipeline

# load pipeline
image_classifier = pipeline(model="alphaedge-ai/siglip2-base-patch16-512-ben-16384", task="zero-shot-image-classification")

# load image and candidate labels
image = "http://images.cocodataset.org/val2017/000000039769.jpg"
candidate_labels = ["Potential label 1 in Bengali", "Potential label 2 in Bengali", "Potential label 3 in Bengali", "Potential label 4 in Bengali"]

# run inference
outputs = image_classifier(image, candidate_labels)
print(outputs)

Sentence-transformers (texts-images similarity)

python

from sentence_transformers import SentenceTransformer

model = SentenceTransformer("alphaedge-ai/siglip2-base-patch16-512-ben-16384")

images = [
    "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/tasks/car.jpg",
    "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/bee.jpg",
    "https://huggingface.co/datasets/huggingface/cats-image/resolve/main/cats_image.jpeg"
]
texts = ["Text 1 in Bengali", "Text 2 in Bengali", "Text 3 in Bengali", "Text 4 in Bengali"]

image_embeddings = model.encode(images)
text_embeddings = model.encode(texts)
print(image_embeddings.shape, text_embeddings.shape)

similarities = model.similarity(image_embeddings, text_embeddings)
print(similarities)

Citations

SigLIP 2

@misc{tschannen2025siglip2multilingualvisionlanguage,
      title={SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features}, 
      author={Michael Tschannen and Alexey Gritsenko and Xiao Wang and Muhammad Ferjad Naeem and Ibrahim Alabdulmohsin and Nikhil Parthasarathy and Talfan Evans and Lucas Beyer and Ye Xia and Basil Mustafa and Olivier Hénaff and Jeremiah Harmsen and Andreas Steiner and Xiaohua Zhai},
      year={2025},
      eprint={2502.14786},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2502.14786}, 
}

Trimming blog post

@misc{hf_blogpost_trimming,
      title={Introduction to Trimming}, 
      author={Loïck BOURDOIS and Tom AARSEN and Bram VANROY and Christopher AKIKI and Woojun JUNG and Manuel ROMERO and Prithiv SAKTHI},
      year={2026},
      url={https://huggingface.co/blog/lbourdois/introduction-to-trimming}, 
}