Specialized Configs for Triton Backends
The Python API provides specialized configuration classes that help provide only available options for the given type of model.
model_navigator.api.triton.BaseSpecializedModelConfig
dataclass
Bases: ABC
Common fields for specialized model configs.
Read more in Triton Inference server documentation
Parameters:
-
max_batch_size
(
int
, default:4
) –The maximal batch size that would be handled by model.
-
batching
(
bool
, default:True
) –Flag to enable/disable batching for model.
-
batcher
(
Union[DynamicBatcher, SequenceBatcher]
, default:dataclasses.field(default_factory=DynamicBatcher)
) –Configuration of Dynamic Batching for the model.
-
instance_groups
(
List[InstanceGroup]
, default:dataclasses.field(default_factory=lambda : [])
) –Instance groups configuration for multiple instances of the model
-
parameters
(
Dict[str, str]
, default:dataclasses.field(default_factory=lambda : {})
) –Custom parameters for model or backend
-
response_cache
(
bool
, default:False
) –Flag to enable/disable response cache for the model
-
warmup
(
Dict[str, ModelWarmup]
, default:dataclasses.field(default_factory=lambda : {})
) –Warmup configuration for model
model_navigator.api.triton.ONNXModelConfig
dataclass
Bases: BaseSpecializedModelConfig
Specialized model config for ONNX backend supported model.
Parameters:
-
platform
(
Optional[Platform]
, default:None
) –Override backend parameter with platform. Possible options: Platform.ONNXRuntimeONNX
-
optimization
(
Optional[ONNXOptimization]
, default:None
) –Possible optimization for ONNX models
model_navigator.api.triton.ONNXOptimization
dataclass
ONNX possible optimizations.
Parameters:
-
accelerator
(
Union[OpenVINOAccelerator, TensorRTAccelerator]
) –Execution accelerator for model
model_navigator.api.triton.PythonModelConfig
dataclass
Bases: BaseSpecializedModelConfig
Specialized model config for Python backend supported model.
Parameters:
-
inputs
(
Sequence[InputTensorSpec]
, default:dataclasses.field(default_factory=lambda : [])
) –Required definition of model inputs
-
outputs
(
Sequence[OutputTensorSpec]
, default:dataclasses.field(default_factory=lambda : [])
) –Required definition of model outputs
model_navigator.api.triton.PyTorchModelConfig
dataclass
Bases: BaseSpecializedModelConfig
Specialized model config for PyTorch backend supported model.
Parameters:
-
platform
(
Optional[Platform]
, default:None
) –Override backend parameter with platform. Possible options: Platform.PyTorchLibtorch
-
inputs
(
Sequence[InputTensorSpec]
, default:dataclasses.field(default_factory=lambda : [])
) –Required definition of model inputs
-
outputs
(
Sequence[OutputTensorSpec]
, default:dataclasses.field(default_factory=lambda : [])
) –Required definition of model outputs
model_navigator.api.triton.TensorFlowModelConfig
dataclass
Bases: BaseSpecializedModelConfig
Specialized model config for TensorFlow backend supported model.
Parameters:
-
platform
(
Optional[Platform]
, default:None
) –Override backend parameter with platform. Possible options: Platform.TensorFlowSavedModel, Platform.TensorFlowGraphDef
-
optimization
(
Optional[TensorFlowOptimization]
, default:None
) –Possible optimization for TensorFlow models
model_navigator.api.triton.TensorFlowOptimization
dataclass
TensorFlow possible optimizations.
Parameters:
-
accelerator
(
Union[AutoMixedPrecisionAccelerator, GPUIOAccelerator, TensorRTAccelerator]
) –Execution accelerator for model
model_navigator.api.triton.TensorRTModelConfig
dataclass
Bases: BaseSpecializedModelConfig
Specialized model config for TensorRT platform supported model.
Parameters:
-
platform
(
Optional[Platform]
, default:None
) –Override backend parameter with platform. Possible options: Platform.TensorRTPlan
-
optimization
(
Optional[TensorRTOptimization]
, default:None
) –Possible optimization for TensorRT models
model_navigator.api.triton.TensorRTOptimization
dataclass
TensorRT possible optimizations.
Parameters:
-
cuda_graphs
(
bool
, default:False
) –Use CUDA graphs API to capture model operations and execute them more efficiently.
-
gather_kernel_buffer_threshold
(
Optional[int]
, default:None
) –The backend may use a gather kernel to gather input data if the device has direct access to the source buffer and the destination buffer.
-
eager_batching
(
bool
, default:False
) –Start preparing the next batch before the model instance is ready for the next inference.