[
https://issues.apache.org/jira/browse/OPENNLP-1374?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17576992#comment-17576992
]
ASF GitHub Bot commented on OPENNLP-1374:
-----------------------------------------
jzonthemtn commented on PR #422:
URL: https://github.com/apache/opennlp/pull/422#issuecomment-1208648304
This PR adds initial support for long documents by splitting them up into
sections. (The size of the splits and size of the overlap of splits can be set
in the `InferenceOptions`.)
> Update ONNX code to allow for unlimited text length and for reusability
> ------------------------------------------------------------------------
>
> Key: OPENNLP-1374
> URL: https://issues.apache.org/jira/browse/OPENNLP-1374
> Project: OpenNLP
> Issue Type: Task
> Components: Deep Learning
> Affects Versions: 2.0.0
> Reporter: Jeff Zemerick
> Assignee: Jeff Zemerick
> Priority: Major
>
> There are a few updates needed in the ONNX code that make sense to make
> together.
> * Allow for processing text of unlimited length. Currently requires input
> text to be broken up outside of OpenNLP prior to inference.
> * Use constants for labels. Allow the labels for B_PER and I_PER to be
> customized by the user (if needed).
> * Refactor the code for better reuse looking down the road at other
> interface implementations.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
