""" @generated by mypy-protobuf. Do not edit manually! isort:skip_file """ import builtins import collections.abc import sys import typing import google.protobuf.descriptor import google.protobuf.internal.containers import google.protobuf.internal.enum_type_wrapper import google.protobuf.message import tensorflow.core.framework.attr_value_pb2 import tensorflow.core.protobuf.verifier_config_pb2 if sys.version_info >= (3, 10): import typing as typing_extensions else: import typing_extensions DESCRIPTOR: google.protobuf.descriptor.FileDescriptor @typing_extensions.final class AutoParallelOptions(google.protobuf.message.Message): DESCRIPTOR: google.protobuf.descriptor.Descriptor ENABLE_FIELD_NUMBER: builtins.int NUM_REPLICAS_FIELD_NUMBER: builtins.int enable: builtins.bool num_replicas: builtins.int def __init__( self, *, enable: builtins.bool | None = ..., num_replicas: builtins.int | None = ..., ) -> None: ... def ClearField(self, field_name: typing_extensions.Literal["enable", b"enable", "num_replicas", b"num_replicas"]) -> None: ... global___AutoParallelOptions = AutoParallelOptions @typing_extensions.final class ScopedAllocatorOptions(google.protobuf.message.Message): DESCRIPTOR: google.protobuf.descriptor.Descriptor ENABLE_OP_FIELD_NUMBER: builtins.int @property def enable_op(self) -> google.protobuf.internal.containers.RepeatedScalarFieldContainer[builtins.str]: """If present, only perform optimization for these ops.""" def __init__( self, *, enable_op: collections.abc.Iterable[builtins.str] | None = ..., ) -> None: ... def ClearField(self, field_name: typing_extensions.Literal["enable_op", b"enable_op"]) -> None: ... global___ScopedAllocatorOptions = ScopedAllocatorOptions @typing_extensions.final class RewriterConfig(google.protobuf.message.Message): """Graph rewriting is experimental and subject to change, not covered by any API stability guarantees. """ DESCRIPTOR: google.protobuf.descriptor.Descriptor class _Toggle: ValueType = typing.NewType("ValueType", builtins.int) V: typing_extensions.TypeAlias = ValueType class _ToggleEnumTypeWrapper(google.protobuf.internal.enum_type_wrapper._EnumTypeWrapper[RewriterConfig._Toggle.ValueType], builtins.type): DESCRIPTOR: google.protobuf.descriptor.EnumDescriptor DEFAULT: RewriterConfig._Toggle.ValueType # 0 ON: RewriterConfig._Toggle.ValueType # 1 OFF: RewriterConfig._Toggle.ValueType # 2 AGGRESSIVE: RewriterConfig._Toggle.ValueType # 3 """Enable some aggressive optimizations that use assumptions that TF graphs may break. For example, assume the shape of a placeholder matches its actual feed. """ EXPERIMENTAL_MLIR: RewriterConfig._Toggle.ValueType # 4 """Run MLIR pass if there's one implemented in TFG, do nothing otherwise. I.e., if there's no corresponding TFG pass, it's an OFF. This is supposed to be mapped with `ON` and there's no `AGGRESSIVE` in MLIR pass now. """ EXPERIMENTAL_BOTH: RewriterConfig._Toggle.ValueType # 5 """Run both MLIR and Grappler passes consecutively and MLIR pass will come first. """ class Toggle(_Toggle, metaclass=_ToggleEnumTypeWrapper): """Configuration options for the meta-optimizer. Unless otherwise noted, these configuration options do not apply to explicitly triggered optimization passes in the optimizers field. """ DEFAULT: RewriterConfig.Toggle.ValueType # 0 ON: RewriterConfig.Toggle.ValueType # 1 OFF: RewriterConfig.Toggle.ValueType # 2 AGGRESSIVE: RewriterConfig.Toggle.ValueType # 3 """Enable some aggressive optimizations that use assumptions that TF graphs may break. For example, assume the shape of a placeholder matches its actual feed. """ EXPERIMENTAL_MLIR: RewriterConfig.Toggle.ValueType # 4 """Run MLIR pass if there's one implemented in TFG, do nothing otherwise. I.e., if there's no corresponding TFG pass, it's an OFF. This is supposed to be mapped with `ON` and there's no `AGGRESSIVE` in MLIR pass now. """ EXPERIMENTAL_BOTH: RewriterConfig.Toggle.ValueType # 5 """Run both MLIR and Grappler passes consecutively and MLIR pass will come first. """ class _CpuLayout: ValueType = typing.NewType("ValueType", builtins.int) V: typing_extensions.TypeAlias = ValueType class _CpuLayoutEnumTypeWrapper(google.protobuf.internal.enum_type_wrapper._EnumTypeWrapper[RewriterConfig._CpuLayout.ValueType], builtins.type): DESCRIPTOR: google.protobuf.descriptor.EnumDescriptor NO_CONVERSION_ON_CPU: RewriterConfig._CpuLayout.ValueType # 0 NCHW_TO_NHWC: RewriterConfig._CpuLayout.ValueType # 1 NHWC_TO_NCHW: RewriterConfig._CpuLayout.ValueType # 2 class CpuLayout(_CpuLayout, metaclass=_CpuLayoutEnumTypeWrapper): """Enum for layout conversion between NCHW and NHWC on CPU. Default is OFF.""" NO_CONVERSION_ON_CPU: RewriterConfig.CpuLayout.ValueType # 0 NCHW_TO_NHWC: RewriterConfig.CpuLayout.ValueType # 1 NHWC_TO_NCHW: RewriterConfig.CpuLayout.ValueType # 2 class _NumIterationsType: ValueType = typing.NewType("ValueType", builtins.int) V: typing_extensions.TypeAlias = ValueType class _NumIterationsTypeEnumTypeWrapper(google.protobuf.internal.enum_type_wrapper._EnumTypeWrapper[RewriterConfig._NumIterationsType.ValueType], builtins.type): DESCRIPTOR: google.protobuf.descriptor.EnumDescriptor DEFAULT_NUM_ITERS: RewriterConfig._NumIterationsType.ValueType # 0 ONE: RewriterConfig._NumIterationsType.ValueType # 1 TWO: RewriterConfig._NumIterationsType.ValueType # 2 class NumIterationsType(_NumIterationsType, metaclass=_NumIterationsTypeEnumTypeWrapper): """Enum controlling the number of times to run optimizers. The default is to run them twice. """ DEFAULT_NUM_ITERS: RewriterConfig.NumIterationsType.ValueType # 0 ONE: RewriterConfig.NumIterationsType.ValueType # 1 TWO: RewriterConfig.NumIterationsType.ValueType # 2 class _MemOptType: ValueType = typing.NewType("ValueType", builtins.int) V: typing_extensions.TypeAlias = ValueType class _MemOptTypeEnumTypeWrapper(google.protobuf.internal.enum_type_wrapper._EnumTypeWrapper[RewriterConfig._MemOptType.ValueType], builtins.type): DESCRIPTOR: google.protobuf.descriptor.EnumDescriptor DEFAULT_MEM_OPT: RewriterConfig._MemOptType.ValueType # 0 """The default setting (SCHEDULING and SWAPPING HEURISTICS only)""" NO_MEM_OPT: RewriterConfig._MemOptType.ValueType # 1 """Disabled in the meta-optimizer.""" MANUAL: RewriterConfig._MemOptType.ValueType # 2 """Driven by manual op-level annotations.""" SWAPPING_HEURISTICS: RewriterConfig._MemOptType.ValueType # 4 """Driven by heuristics. The behavior of these heuristics is subject to change. Currently includes an experimental recomputation and swapping heuristics. Manual annotations are respected, but additional nodes are selected automatically. Swapping heuristic will move a tensor from the GPU to the CPU and move it back when needed to reduce peak memory usage. """ RECOMPUTATION_HEURISTICS: RewriterConfig._MemOptType.ValueType # 5 """Recomputation heuristics will recompute ops (such as Relu activation) during backprop instead of storing them, reducing peak memory usage. """ SCHEDULING_HEURISTICS: RewriterConfig._MemOptType.ValueType # 6 """Scheduling will split big ops such as AddN and try to enforce a schedule of the new computations that decreases peak memory usage. """ HEURISTICS: RewriterConfig._MemOptType.ValueType # 3 """Use any combination of swapping and recomputation heuristics.""" class MemOptType(_MemOptType, metaclass=_MemOptTypeEnumTypeWrapper): ... DEFAULT_MEM_OPT: RewriterConfig.MemOptType.ValueType # 0 """The default setting (SCHEDULING and SWAPPING HEURISTICS only)""" NO_MEM_OPT: RewriterConfig.MemOptType.ValueType # 1 """Disabled in the meta-optimizer.""" MANUAL: RewriterConfig.MemOptType.ValueType # 2 """Driven by manual op-level annotations.""" SWAPPING_HEURISTICS: RewriterConfig.MemOptType.ValueType # 4 """Driven by heuristics. The behavior of these heuristics is subject to change. Currently includes an experimental recomputation and swapping heuristics. Manual annotations are respected, but additional nodes are selected automatically. Swapping heuristic will move a tensor from the GPU to the CPU and move it back when needed to reduce peak memory usage. """ RECOMPUTATION_HEURISTICS: RewriterConfig.MemOptType.ValueType # 5 """Recomputation heuristics will recompute ops (such as Relu activation) during backprop instead of storing them, reducing peak memory usage. """ SCHEDULING_HEURISTICS: RewriterConfig.MemOptType.ValueType # 6 """Scheduling will split big ops such as AddN and try to enforce a schedule of the new computations that decreases peak memory usage. """ HEURISTICS: RewriterConfig.MemOptType.ValueType # 3 """Use any combination of swapping and recomputation heuristics.""" @typing_extensions.final class CustomGraphOptimizer(google.protobuf.message.Message): """Message to describe custom graph optimizer and its parameters""" DESCRIPTOR: google.protobuf.descriptor.Descriptor @typing_extensions.final class ParameterMapEntry(google.protobuf.message.Message): DESCRIPTOR: google.protobuf.descriptor.Descriptor KEY_FIELD_NUMBER: builtins.int VALUE_FIELD_NUMBER: builtins.int key: builtins.str @property def value(self) -> tensorflow.core.framework.attr_value_pb2.AttrValue: ... def __init__( self, *, key: builtins.str | None = ..., value: tensorflow.core.framework.attr_value_pb2.AttrValue | None = ..., ) -> None: ... def HasField(self, field_name: typing_extensions.Literal["value", b"value"]) -> builtins.bool: ... def ClearField(self, field_name: typing_extensions.Literal["key", b"key", "value", b"value"]) -> None: ... NAME_FIELD_NUMBER: builtins.int PARAMETER_MAP_FIELD_NUMBER: builtins.int name: builtins.str @property def parameter_map(self) -> google.protobuf.internal.containers.MessageMap[builtins.str, tensorflow.core.framework.attr_value_pb2.AttrValue]: ... def __init__( self, *, name: builtins.str | None = ..., parameter_map: collections.abc.Mapping[builtins.str, tensorflow.core.framework.attr_value_pb2.AttrValue] | None = ..., ) -> None: ... def ClearField(self, field_name: typing_extensions.Literal["name", b"name", "parameter_map", b"parameter_map"]) -> None: ... CPU_LAYOUT_CONVERSION_FIELD_NUMBER: builtins.int LAYOUT_OPTIMIZER_FIELD_NUMBER: builtins.int CONSTANT_FOLDING_FIELD_NUMBER: builtins.int SHAPE_OPTIMIZATION_FIELD_NUMBER: builtins.int REMAPPING_FIELD_NUMBER: builtins.int COMMON_SUBGRAPH_ELIMINATION_FIELD_NUMBER: builtins.int ARITHMETIC_OPTIMIZATION_FIELD_NUMBER: builtins.int DEPENDENCY_OPTIMIZATION_FIELD_NUMBER: builtins.int LOOP_OPTIMIZATION_FIELD_NUMBER: builtins.int FUNCTION_OPTIMIZATION_FIELD_NUMBER: builtins.int DEBUG_STRIPPER_FIELD_NUMBER: builtins.int DISABLE_MODEL_PRUNING_FIELD_NUMBER: builtins.int SCOPED_ALLOCATOR_OPTIMIZATION_FIELD_NUMBER: builtins.int PIN_TO_HOST_OPTIMIZATION_FIELD_NUMBER: builtins.int IMPLEMENTATION_SELECTOR_FIELD_NUMBER: builtins.int AUTO_MIXED_PRECISION_FIELD_NUMBER: builtins.int AUTO_MIXED_PRECISION_MKL_FIELD_NUMBER: builtins.int AUTO_MIXED_PRECISION_ONEDNN_BFLOAT16_FIELD_NUMBER: builtins.int AUTO_MIXED_PRECISION_CPU_FIELD_NUMBER: builtins.int DISABLE_META_OPTIMIZER_FIELD_NUMBER: builtins.int USE_PLUGIN_OPTIMIZERS_FIELD_NUMBER: builtins.int EXPERIMENTAL_CONDITIONAL_CODE_MOTION_FIELD_NUMBER: builtins.int META_OPTIMIZER_ITERATIONS_FIELD_NUMBER: builtins.int MIN_GRAPH_NODES_FIELD_NUMBER: builtins.int EXPERIMENTAL_DISABLE_COMPRESSED_TENSOR_OPTIMIZATION_FIELD_NUMBER: builtins.int EXPERIMENTAL_DISABLE_FOLDING_QUANTIZATION_EMULATION_FIELD_NUMBER: builtins.int MEMORY_OPTIMIZATION_FIELD_NUMBER: builtins.int MEMORY_OPTIMIZER_TARGET_NODE_NAME_SCOPE_FIELD_NUMBER: builtins.int META_OPTIMIZER_TIMEOUT_MS_FIELD_NUMBER: builtins.int AUTO_PARALLEL_FIELD_NUMBER: builtins.int FAIL_ON_OPTIMIZER_ERRORS_FIELD_NUMBER: builtins.int SCOPED_ALLOCATOR_OPTS_FIELD_NUMBER: builtins.int OPTIMIZERS_FIELD_NUMBER: builtins.int CUSTOM_OPTIMIZERS_FIELD_NUMBER: builtins.int INTER_OPTIMIZER_VERIFIER_CONFIG_FIELD_NUMBER: builtins.int POST_OPTIMIZATION_VERIFIER_CONFIG_FIELD_NUMBER: builtins.int cpu_layout_conversion: global___RewriterConfig.CpuLayout.ValueType """CPU Conversion settings between NHCW and NCHW.""" layout_optimizer: global___RewriterConfig.Toggle.ValueType """Optimize tensor layouts (default is ON) e.g. This will try to use NCHW layout on GPU which is faster. """ constant_folding: global___RewriterConfig.Toggle.ValueType """Fold constants (default is ON) Statically infer the value of tensors when possible, and materialize the result using constants. """ shape_optimization: global___RewriterConfig.Toggle.ValueType """Shape optimizations (default is ON) Simplify computations made on shapes. """ remapping: global___RewriterConfig.Toggle.ValueType """Remapping (default is ON) Remap subgraphs onto more efficient implementations. """ common_subgraph_elimination: global___RewriterConfig.Toggle.ValueType """Common subgraph elimination (default is ON) e.g. Simplify arithmetic ops; merge ops with same value (like constants). """ arithmetic_optimization: global___RewriterConfig.Toggle.ValueType """Arithmetic optimizations (default is ON) e.g. Simplify arithmetic ops; merge ops with same value (like constants). """ dependency_optimization: global___RewriterConfig.Toggle.ValueType """Control dependency optimizations (default is ON). Remove redundant control dependencies, which may enable other optimization. """ loop_optimization: global___RewriterConfig.Toggle.ValueType """Loop optimizations (default is ON).""" function_optimization: global___RewriterConfig.Toggle.ValueType """Function optimizations (default is ON).""" debug_stripper: global___RewriterConfig.Toggle.ValueType """Strips debug-related nodes from the graph (off by default).""" disable_model_pruning: builtins.bool """If true, don't remove unnecessary ops from the graph""" scoped_allocator_optimization: global___RewriterConfig.Toggle.ValueType """Try to allocate some independent Op outputs contiguously in order to merge or eliminate downstream Ops (off by default). """ pin_to_host_optimization: global___RewriterConfig.Toggle.ValueType """Force small ops onto the CPU (default is OFF).""" implementation_selector: global___RewriterConfig.Toggle.ValueType """Enable the swap of kernel implementations based on the device placement (default is ON). """ auto_mixed_precision: global___RewriterConfig.Toggle.ValueType """Optimize data types for CUDA (default is OFF). This will try to use float16 on GPU which is faster. Note that this can change the numerical stability of the graph and may require the use of loss scaling to maintain model convergence. """ auto_mixed_precision_mkl: global___RewriterConfig.Toggle.ValueType """Optimize data types for oneDNN (default is OFF). This will try to use bfloat16 on CPUs, which is faster. Note that this can change the numerical stability of the graph. Note: this is deprecated. It is replaced by auto_mixed_precision_onednn_bfloat16 """ auto_mixed_precision_onednn_bfloat16: global___RewriterConfig.Toggle.ValueType """Optimize data types for oneDNN (default is OFF). This will try to use bfloat16 on CPUs, which is faster. Note that this can change the numerical stability of the graph. Note: this is equivalent to the deprecated option auto_mixed_precision_mkl """ auto_mixed_precision_cpu: global___RewriterConfig.Toggle.ValueType """Emulate a model using data type float16 on CPU (default is OFF). This will try to emulate the float16 inputs and outputs of an operator on CPU to have better correlation with float16 on GPU; however the computation in the operator is based on float32. Note that this can change the numerical stability of the graph. """ disable_meta_optimizer: builtins.bool """Disable the entire meta optimizer (off by default).""" use_plugin_optimizers: global___RewriterConfig.Toggle.ValueType """Optimizers registered by plugin (default is ON)""" experimental_conditional_code_motion: global___RewriterConfig.Toggle.ValueType """Conditional code motion (default is ON).""" meta_optimizer_iterations: global___RewriterConfig.NumIterationsType.ValueType """Controls how many times we run the optimizers in meta optimizer (default is once). """ min_graph_nodes: builtins.int """The minimum number of nodes in a graph to optimizer. For smaller graphs, optimization is skipped. 0 means the system picks an appropriate number. < 0 means do not skip optimization. """ experimental_disable_compressed_tensor_optimization: builtins.bool """Disable optimizations that assume compressed tensors. Note that this flag is experimental and may be removed in the future. """ experimental_disable_folding_quantization_emulation: builtins.bool """Disable folding quantization emulation ops such as FakeQuantWithMinMax* and QuantizeAndDequantize*. Some compilers (e.g. the TF-to-tflite converter) have to extract quantization configs (e.g. min/max range, number of bits, and per-channel) from the quantization emulation ops. Note that this flag is experimental and may be removed in the future. See b/174138564 for more details. """ memory_optimization: global___RewriterConfig.MemOptType.ValueType """Configures memory optimization passes through the meta-optimizer. Has no effect on manually requested memory optimization passes in the optimizers field. """ memory_optimizer_target_node_name_scope: builtins.str """A node name scope for node names which are valid outputs of recomputations. Inputs to nodes that match this scope may be recomputed (subject either to manual annotation of those input nodes or to manual annotation and heuristics depending on memory_optimization), but the nodes themselves will not be recomputed. This matches any sub-scopes as well, meaning the scope can appear not just as a top-level scope. For example, if the value is "gradients/", the default, it will match node name "gradients/foo", "foo/gradients/bar", but not "foo_gradients/" """ meta_optimizer_timeout_ms: builtins.int """Maximum number of milliseconds to spend optimizing a single graph before timing out. If less than or equal to 0 (default value) the optimizer will never time out. """ @property def auto_parallel(self) -> global___AutoParallelOptions: """Configures AutoParallel optimization passes either through the meta-optimizer or when manually specified through the optimizers field. """ fail_on_optimizer_errors: builtins.bool """If true, any optimization pass failing will cause the MetaOptimizer to stop with an error. By default - or when set to false, failing passes are skipped silently. """ @property def scoped_allocator_opts(self) -> global___ScopedAllocatorOptions: ... @property def optimizers(self) -> google.protobuf.internal.containers.RepeatedScalarFieldContainer[builtins.str]: """If non-empty, will use this as an alternative way to specify a list of optimizations to turn on and the order of the optimizations (replacing the meta-optimizer). Of the RewriterConfig options, only the AutoParallel configuration options (the auto_parallel field) apply to manually requested optimization passes ("autoparallel"). Memory optimization passes ("memory") invoked here are not configurable (in contrast to memory optimization passes through the meta-optimizer) and act only on manual op annotations. Custom optimizers (see custom_optimizers) that are not part of this schedule will be run after - in the order that they were specified. """ @property def custom_optimizers(self) -> google.protobuf.internal.containers.RepeatedCompositeFieldContainer[global___RewriterConfig.CustomGraphOptimizer]: """list of CustomGraphOptimizers to apply.""" @property def inter_optimizer_verifier_config(self) -> tensorflow.core.protobuf.verifier_config_pb2.VerifierConfig: """VerifierConfig specifying the verifiers to be run after every optimizer.""" @property def post_optimization_verifier_config(self) -> tensorflow.core.protobuf.verifier_config_pb2.VerifierConfig: """VerifierConfig specifying the verifiers to be run at the end, after all optimizers have run. """ def __init__( self, *, cpu_layout_conversion: global___RewriterConfig.CpuLayout.ValueType | None = ..., layout_optimizer: global___RewriterConfig.Toggle.ValueType | None = ..., constant_folding: global___RewriterConfig.Toggle.ValueType | None = ..., shape_optimization: global___RewriterConfig.Toggle.ValueType | None = ..., remapping: global___RewriterConfig.Toggle.ValueType | None = ..., common_subgraph_elimination: global___RewriterConfig.Toggle.ValueType | None = ..., arithmetic_optimization: global___RewriterConfig.Toggle.ValueType | None = ..., dependency_optimization: global___RewriterConfig.Toggle.ValueType | None = ..., loop_optimization: global___RewriterConfig.Toggle.ValueType | None = ..., function_optimization: global___RewriterConfig.Toggle.ValueType | None = ..., debug_stripper: global___RewriterConfig.Toggle.ValueType | None = ..., disable_model_pruning: builtins.bool | None = ..., scoped_allocator_optimization: global___RewriterConfig.Toggle.ValueType | None = ..., pin_to_host_optimization: global___RewriterConfig.Toggle.ValueType | None = ..., implementation_selector: global___RewriterConfig.Toggle.ValueType | None = ..., auto_mixed_precision: global___RewriterConfig.Toggle.ValueType | None = ..., auto_mixed_precision_mkl: global___RewriterConfig.Toggle.ValueType | None = ..., auto_mixed_precision_onednn_bfloat16: global___RewriterConfig.Toggle.ValueType | None = ..., auto_mixed_precision_cpu: global___RewriterConfig.Toggle.ValueType | None = ..., disable_meta_optimizer: builtins.bool | None = ..., use_plugin_optimizers: global___RewriterConfig.Toggle.ValueType | None = ..., experimental_conditional_code_motion: global___RewriterConfig.Toggle.ValueType | None = ..., meta_optimizer_iterations: global___RewriterConfig.NumIterationsType.ValueType | None = ..., min_graph_nodes: builtins.int | None = ..., experimental_disable_compressed_tensor_optimization: builtins.bool | None = ..., experimental_disable_folding_quantization_emulation: builtins.bool | None = ..., memory_optimization: global___RewriterConfig.MemOptType.ValueType | None = ..., memory_optimizer_target_node_name_scope: builtins.str | None = ..., meta_optimizer_timeout_ms: builtins.int | None = ..., auto_parallel: global___AutoParallelOptions | None = ..., fail_on_optimizer_errors: builtins.bool | None = ..., scoped_allocator_opts: global___ScopedAllocatorOptions | None = ..., optimizers: collections.abc.Iterable[builtins.str] | None = ..., custom_optimizers: collections.abc.Iterable[global___RewriterConfig.CustomGraphOptimizer] | None = ..., inter_optimizer_verifier_config: tensorflow.core.protobuf.verifier_config_pb2.VerifierConfig | None = ..., post_optimization_verifier_config: tensorflow.core.protobuf.verifier_config_pb2.VerifierConfig | None = ..., ) -> None: ... def HasField(self, field_name: typing_extensions.Literal["auto_parallel", b"auto_parallel", "inter_optimizer_verifier_config", b"inter_optimizer_verifier_config", "post_optimization_verifier_config", b"post_optimization_verifier_config", "scoped_allocator_opts", b"scoped_allocator_opts"]) -> builtins.bool: ... def ClearField(self, field_name: typing_extensions.Literal["arithmetic_optimization", b"arithmetic_optimization", "auto_mixed_precision", b"auto_mixed_precision", "auto_mixed_precision_cpu", b"auto_mixed_precision_cpu", "auto_mixed_precision_mkl", b"auto_mixed_precision_mkl", "auto_mixed_precision_onednn_bfloat16", b"auto_mixed_precision_onednn_bfloat16", "auto_parallel", b"auto_parallel", "common_subgraph_elimination", b"common_subgraph_elimination", "constant_folding", b"constant_folding", "cpu_layout_conversion", b"cpu_layout_conversion", "custom_optimizers", b"custom_optimizers", "debug_stripper", b"debug_stripper", "dependency_optimization", b"dependency_optimization", "disable_meta_optimizer", b"disable_meta_optimizer", "disable_model_pruning", b"disable_model_pruning", "experimental_conditional_code_motion", b"experimental_conditional_code_motion", "experimental_disable_compressed_tensor_optimization", b"experimental_disable_compressed_tensor_optimization", "experimental_disable_folding_quantization_emulation", b"experimental_disable_folding_quantization_emulation", "fail_on_optimizer_errors", b"fail_on_optimizer_errors", "function_optimization", b"function_optimization", "implementation_selector", b"implementation_selector", "inter_optimizer_verifier_config", b"inter_optimizer_verifier_config", "layout_optimizer", b"layout_optimizer", "loop_optimization", b"loop_optimization", "memory_optimization", b"memory_optimization", "memory_optimizer_target_node_name_scope", b"memory_optimizer_target_node_name_scope", "meta_optimizer_iterations", b"meta_optimizer_iterations", "meta_optimizer_timeout_ms", b"meta_optimizer_timeout_ms", "min_graph_nodes", b"min_graph_nodes", "optimizers", b"optimizers", "pin_to_host_optimization", b"pin_to_host_optimization", "post_optimization_verifier_config", b"post_optimization_verifier_config", "remapping", b"remapping", "scoped_allocator_optimization", b"scoped_allocator_optimization", "scoped_allocator_opts", b"scoped_allocator_opts", "shape_optimization", b"shape_optimization", "use_plugin_optimizers", b"use_plugin_optimizers"]) -> None: ... global___RewriterConfig = RewriterConfig