ModelInference
Function Usage
Performs model inference. You need to call Init to initialize the model first.
Prototype
1 | APP_ERROR ModelInferenceProcessor::ModelInference(std::vector<BaseTensor>& inputTensors, std::vector<BaseTensor>& outputTensors, size_t dynamicBatchSize = 0); |
1 | APP_ERROR ModelInferenceProcessor::ModelInference(std::vector<BaseTensor>& inputTensors, std::vector<BaseTensor>& outputTensors, DynamicInfo dynamicInfo); |
1 | APP_ERROR ModelInferenceProcessor::ModelInference(const std::vector<TensorBase>& inputTensors, std::vector<TensorBase>& outputTensors, DynamicInfo dynamicInfo); |
Parameters
Parameter |
Input/Output |
Description |
|---|---|---|
inputTensors |
Input |
Input tensor, which is the input data of model inference |
outputTensors |
Output |
Output tensor, which is the output data of model inference |
dynamicBatchSize |
Input |
Batch size for model inference |
dynamicInfo |
Input |
Dynamic batch information |
Response Parameters
Data Structure |
Description |
|---|---|
APP_ERROR |
For details about the returned error codes, see APP_ERROR Description. |
Parent topic: ModelInferenceProcessor