GitHub - xetdata/onnx-models: A copy of ONNX models, datasets, and code all in one GitHub repository. Follow the README to learn more.
Nov 29, 2023 - github.com
The ONNX Model Zoo has expanded its repository to include models and data from the original ONNX repository. The repository now allows users to see links to Netron views of each Xet-managed model from within GitHub, use normal commands like `git clone`, `git add` to develop, or try a fancy lazy clone command to only download the files they're interested in. The repository also includes a wide range of models for various applications such as image classification, object detection, language modelling, machine translation, and more. Users are encouraged to contribute their models to the growing model zoo.
To use the models, every ONNX backend should support running the models out of the box. After downloading and extracting the tarball of each model, users will find a protobuf file `model.onnx` that represents the serialized ONNX model and test data. The repository also provides starter code for using the test data files. Users can also get quantized ONNX models by using the Intel® Neural Compressor. The models can be visualized using Netron. The ONNX Model Zoo is licensed under the Apache License v2.0.
Key takeaways:
The ONNX Model Zoo repository now includes models and data, enhancing the original ONNX repository.
Users can install the XetData integration for GitHub to work with this repository and can use commands like 'git clone' and 'git add' to develop.
The repository includes a wide range of models for tasks such as image classification, object detection, language modelling, machine translation, and more.
Users can contribute their own models to the ONNX Model Zoo, following the provided guidelines.
