Operator Verification on Network (Inference)
Overview
After a custom operator is deployed into the operator library, you can load the operator to the network for inference verification to check its execution result.
Operator verification on network involves all deliverables generated during operator development, including the implementation files, operator prototype definition files, operator information library, and operator plugins. This section describes only the verification method.
In MindStudio
Take the following steps in MindStudio to verify your operator on a network:
- Use the ATC tool to convert the original network that contains the custom operator into an offline model adapted to Ascend AI Processor.
For details about how to convert the model, see "Model conversion and tuning" in MindStudio IDE User Guide .
- Develop an AscendCL application, load the generated offline model file for model inference, and verify the inference result.
For details about how to develop an AscendCL application, see "Application development" in MindStudio IDE User Guide .
On Command Line
Take the following steps on the command line to verify your operator on a network:
- Use the ATC tool to convert the original network that contains the custom operator into an offline model adapted to Ascend AI Processor.
For details about how to convert the model, see ATC Instructions.
- Develop an AscendCL application, load the generated offline model file for model inference, and verify the inference result.
For details about how to develop an AscendCL application, see CANN AscendCL Application Software Development Guide (C&C++).