Python APIs

Class 1

The class name is vart.Runner. The following table lists all the functions defined in the vart.Runner class.

Table 1. Quick Function Reference
Type Name Arguments
vart.Runner create_runner xir.Subgraph subgraph

string mode
List[xir.Tensor] get_input_tensors
List[xir.Tensor] get_output_tensors
tuple[uint32, int] execute_async List[vart.TensorBuffer] inputs

List[vart.TensorBuffer] outputs

Note: vart.TensorBuffer complete with buffer protocol .
int wait tuple[uint32, int] jobID

Class 2

The class name is vart.RunnerExt. The following table lists all the functions defined in the vart.RunnerExt class.

vart.RunnerExt extends from vart.Runner.

Table 2. Quick Function Reference
Type Name Arguments
vart.RunnerExt create_runner xir.Subgraph subgraph

String mode
List[vart.TensorBuffer] get_inputs
List[vart.TensorBuffer] get_outputs

Class 3

The class name is vitis_ai_library.GraphRunner. The following table lists all the functions defined in the vitis_ai_library.GraphRunner class.

Table 3. Quick Function Reference
Type Name Arguments
vart.RunnerExt create_graph_runner xir.Graph graph

create_runner

Creates an instance of DPU runner by subgraph. This is a factory function.

Prototype

vart.Runner create_runner(xir.Subgraph subgraph, String mode)

Parameters

The following table lists the create_runner function arguments.

Table 4. create_runner Arguments
Type Name Description
xir.Subgraph subgraph XIR Subgraph
String mode 'run' - DPU runner

Returns

An instance of DPU runner.

execute_async

Executes the runner. This is a block function.

Prototype

tuple[uint32_t, int] execute_async(
      List[vart.TensorBuffer] inputs,
      List[vart.TensorBuffer] outputs)
Note: vart.TensorBuffer complete with buffer protocol.

Parameters

The following table lists the execute_async function arguments.

Table 5. execute_async Arguments
Type Name Description
List[vart.TensorBuffer] inputs A list of vart.TensorBuffer containing the input data for inference.
List[vart.TensorBuffer] outputs A list of vart.TensorBuffer which will be filled with output data.

Returns

tuple[jobID, status] status 0 for exit successfully, others for customized warnings or errors.

wait

Waits for the end of DPU processing. This is a block function.

Prototype

int wait(tuple[uint32_t, int] jobid)

Parameters

The following table lists the wait function arguments.

Table 6. wait Arguments
Type Name Description
tuple[uint32_t, int] jobid job id

Returns

Status 0 for exit successfully, others for customized warnings or errors.

get_input_tensors

Gets all input tensors of runner.

Prototype

List[xir.Tensor] get_input_tensors()

Parameters

None

Returns

All input tensors. A vector of raw pointer to the input tensor.

get_output_tensors

Gets all output tensors of runner.

Prototype

List[xir.Tensor] get_output_tensors()

Parameters

None

Returns

All output tensors. A vector of raw pointer to the output tensor.

get_inputs

Gets all input TensorBuffers of RunnerExt.

Prototype

List[vart.TensorBuffer] get_inputs()

Parameters

None.

Returns

All input TensorBuffers. A vector of raw pointer to the input TensorBuffer.

get_outputs

Gets all output TensorBuffers of RunnerExt.

Prototype

List[vart.TensorBuffer] get_outputs()

Parameters

None.

Returns

All output TensorBuffers. A vector of raw pointer to the output TensorBuffer.

create_graph_runner

Factory function to create an instance of runner by graph.

Prototype

vart.RunnerExt create_graph_runner(xir.Graph graph)

Parameters

The following table lists the create_graph_runner function arguments.

Table 7. create_graph_runner Arguments
Type Name Description
xir.Graph graph XIR Graph

Returns

An instance of runner.