AIModelManager::Process

Processes input data with a model. When inference is performed on a single model, the buffer queue length of the process is limited to 2048 bytes. If the buffer queue length exceeds 2048 bytes, a failure message is returned.

Syntax

virtual AIStatus AIModelManager::Process(AIContext &context, const std::vector<std::shared_ptr<IAITensor>> &in_data, std::vector<std::shared_ptr<IAITensor>> &out_data, uint32_t timeout);

Parameter Description

Parameter	Description	Value Range
context	Context information, including the configurations of variable parameters when the engine is running For details about the syntax of the AIContext data type, see AIContext.	-
in_data	List of input tensors for a model For details about the syntax of the IAITensor data type, see IAITensor.	-
out_data	List of output tensors for a model For details about the syntax of the IAITensor data type, see IAITensor.	-
timeout	Calculation timeout. This parameter is reserved. The default value is 0, indicating that the configuration is invalid.	-