[OPENNLP-1515] Default to onnxruntime instead of onnxruntime-gpu

• Type: Task
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 2.3.1
• Component/s: Deep Learning
• Assignee: Jeff Zemerick
• Reporter: Jeff Zemerick
• Created: 11/Oct/23 15:54
• Updated: 22/Nov/23 10:06
• Resolved: 15/Oct/23 11:05

Description

The onnxruntime-gpu dependency is currently being included by opennlp-dl.

    <dependency>
      <groupId>com.microsoft.onnxruntime</groupId>
      <!-- This dependency supports CPU and GPU -->
      <artifactId>onnxruntime_gpu</artifactId>
      <version>${onnxruntime.version}</version>
    </dependency>

The problem is, GPU support is only on Linux and Windows and not OSX. I think it would be best to use the onnxruntime dependency instead.

But we need to make OpenNLP able to use GPU easily.

Issue Links

• relates to: OPENNLP-1514 Skip ONNX runtime tests when on OSX ARM