EIDORA / VITB16_IN1k

VIT_B16_768

VIT_B16_768 is a medium Vision Transformer image embedding model for visual grouping in EIDORA.

Best For

• Transformer-based image embedding baselines.
• General image grouping and retrieval.

Not Ideal For

• Very large projects where lighter models are preferred.
• Text, video, or audio inputs.

Compute Tier

Medium: better quality or broader domain coverage with moderate runtime cost. Intended for recent laptops and desktops.

Inputs

• image : required image input from media_source .

Output

The primary output is embedding , a float32 tensor shaped [batch, 768] . Embeddings are already normalized and are intended for cosine similarity.

Usage In EIDORA

EIDORA shows this package as a medium image embedding model in the Model Zoo. Use it for discovery maps, grouping, retrieval, and related embedding workflows.

Preprocessing

• image : resize mode defaults to crop_center ; rescale uses 1/255 ; normalize inside the ONNX graph with mean [0.485, 0.456, 0.406] and std [0.229, 0.224, 0.225] .

Authorship And Citation

This ONNX package was produced by EIDORA from the original ViT-B/16 ImageNet model. EIDORA converted the model to ONNX and is not the original model creator. Please cite An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale and the original model repository when using this converted model.

Original model: https://github.com/google-research/vision_transformer

Original paper: https://arxiv.org/abs/2010.11929

Authors: Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, Neil Houlsby

@inproceedings{dosovitskiy2021image,
  title={An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale},
  author={Dosovitskiy, Alexey and others},
  booktitle={ICLR},
  year={2021}
}

Training Data And Provenance

Base model: torchvision/vit-b-16-imagenet1k-v1 . Source repository: https://pytorch.org/vision/stable/models/generated/torchvision.models.vit_b_16.html . Known training data: ImageNet-1K supervised classification data. Package payload size: 343631343 bytes.

Evaluation And Validation

The package validation checks that the ONNX graph loads with ONNX Runtime CPU execution, runs the declared fixtures, returns finite float32 embeddings with the declared shape, and matches the artifact hash recorded in config.yaml .

Limitations And Safety

This supervised ViT-B/16 model is an ImageNet feature baseline and may not match self-supervised or image-text models for semantic retrieval.

License And Attribution

This package uses license bsd-3-clause . Upstream license: BSD-3-Clause for TorchVision code; ImageNet weights are distributed by PyTorch under documented model terms. Converted to ONNX for EIDORA from TorchVision ViT-B/16 ImageNet weights.

Version

Package version: 1.0.0. ONNX opset: 17. Exporter: eidora-onnx-exporter 0.1.0.