IRuntimeConfig

tensorrt_rtx.ExecutionContextAllocationStrategy

Different memory allocation behaviors for IExecutionContext.

Members:

STATIC : Default static allocation with the maximum size across all profiles.

ON_PROFILE_CHANGE : Reallocate for a profile when it’s selected.

USER_MANAGED : The user supplies custom allocation to the execution context.

class tensorrt_rtx.IRuntimeConfig

A runtime configuration for an ICudaEngine .

__init__(*args, **kwargs)

create_runtime_cache(self: tensorrt_rtx.tensorrt_rtx.IRuntimeConfig) → nvinfer1::IRuntimeCache

Create a runtime cache.

Returns:

The runtime cache.

get_execution_context_allocation_strategy(self: tensorrt_rtx.tensorrt_rtx.IRuntimeConfig) → tensorrt_rtx.tensorrt_rtx.ExecutionContextAllocationStrategy

Get the execution context allocation strategy.

Returns:

The execution context allocation strategy.

get_runtime_cache(self: tensorrt_rtx.tensorrt_rtx.IRuntimeConfig) → nvinfer1::IRuntimeCache

Get the runtime cache.

Returns:

The runtime cache.

set_execution_context_allocation_strategy(self: tensorrt_rtx.tensorrt_rtx.IRuntimeConfig, strategy: tensorrt_rtx.tensorrt_rtx.ExecutionContextAllocationStrategy = <ExecutionContextAllocationStrategy.STATIC: 0>) → None

Set the execution context allocation strategy.

Parameters:

strategy – The execution context allocation strategy.

set_runtime_cache(self: tensorrt_rtx.tensorrt_rtx.IRuntimeConfig, cache: tensorrt_rtx.IRuntimeCache = None) → bool

Set the runtime cache.

Parameters:

cache – The runtime cache.