OpenNLP support for ONNX models pull request

[Jeff Zemerick <jz...@apache.org>]

I have created a pull request to add OpenNLP implementations of
DocumentCategorizer and TokenNameFinder for ONNX models. The goal of this
PR is to enable the use of Huggingface transformers models from OpenNLP.

Look at the README.md under the opennlp-dl project for info on how to
convert a transformers model to ONNX and for download links to get the
models that are referenced by the unit tests. The models referenced by the
tests are big so I will add them to the eval test data zip file.

Thanks Bruno for the PR comments! I pushed some changes to address those
and the checkstyle issues.

https://github.com/apache/opennlp/pull/400

Thanks,
Jeff

Re: OpenNLP support for ONNX models pull request

[Jeff Zemerick <jz...@apache.org>]

Hi devs,

Just following up on this PR. If you get a few free minutes I'd love to
have feedback on it.

If/when it gets merged I'd like to begin the conversation about an OpenNLP
2.0 release, pending documentation updates for the ONNX support.

https://github.com/apache/opennlp/pull/400

Thanks,
Jeff


On Tue, Jan 4, 2022 at 10:35 AM Jeff Zemerick <jz...@apache.org> wrote:

> I have created a pull request to add OpenNLP implementations of
> DocumentCategorizer and TokenNameFinder for ONNX models. The goal of this
> PR is to enable the use of Huggingface transformers models from OpenNLP.
>
> Look at the README.md under the opennlp-dl project for info on how to
> convert a transformers model to ONNX and for download links to get the
> models that are referenced by the unit tests. The models referenced by the
> tests are big so I will add them to the eval test data zip file.
>
> Thanks Bruno for the PR comments! I pushed some changes to address those
> and the checkstyle issues.
>
> https://github.com/apache/opennlp/pull/400
>
> Thanks,
> Jeff
>