[
https://issues.apache.org/jira/browse/OPENNLP-1982?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18123036#comment-18123036
]
ASF GitHub Bot commented on OPENNLP-1982:
-----------------------------------------
krickert opened a new issue, #1390:
URL: https://github.com/apache/opennlp/issues/1390
SentenceVectorsDL always creates default session options, so it runs on the
CPU even with opennlp-dl-gpu. The opennlp-dl-gpu README says it brings GPU
acceleration to opennlp-dl.
To reproduce:
There is no constructor that takes InferenceOptions, so setGpu(true) cannot
be passed.
Fix:
A new constructor takes InferenceOptions, like NameFinderDL and
DocumentCategorizerDL. Its GPU settings and lower case setting are used. The
existing constructors keep their behavior.
Branch with a fix and tests:
https://github.com/apache/opennlp/tree/OPENNLP-1982-sentence-vectors-inference-options
> SentenceVectorsDL cannot run on a GPU
> -------------------------------------
>
> Key: OPENNLP-1982
> URL: https://issues.apache.org/jira/browse/OPENNLP-1982
> Project: OpenNLP
> Issue Type: Improvement
> Reporter: Richard Zowalla
> Priority: Major
> Fix For: 3.0.0
>
>
> {{SentenceVectorsDL}} always creates default {{OrtSession.SessionOptions}},
> so unlike {{NameFinderDL}} and {{DocumentCategorizerDL}} it cannot use the
> CUDA execution provider through {{InferenceOptions}}. Add a constructor that
> takes {{InferenceOptions}} and, once OPENNLP\-1937 is in, expose the device
> \(e.g. {{cpu}} / {{cuda:N}}\), a maximum batch size and the intra\-op thread
> count as options of the ONNX text embedder provider.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
