Module executorch_sys::torch::executor

Modules

• util

Structs

• ArrayRef
  Represents a constant reference to an array (0 or more elements consecutively in memory), i.e. a start pointer and a length. It allows various APIs to take consecutive elements easily and conveniently.
• BoxedEvalueList
• DataLoader
  Loads from a data source.
• EValue
• EventTracer
  EventTracer is a class that users can inherit and implement to log/serialize/stream etc. the profiling and debugging events that are generated at runtime for a model. An example of this is the ETDump implementation in the SDK codebase that serializes these events to a flatbuffer.
• FreeableBuffer
  A read-only buffer than can be freed.
• HierarchicalAllocator
  A group of buffers that can be used to represent a device’s memory hierarchy.
• MemoryAllocator
  A class that does simple allocation based on a size and returns the pointer to the memory address. It bookmarks a buffer with certain size. The allocation is simply checking space and growing the cur_ pointer with each allocation request.
• MemoryManager
  A container class for allocators used during Method load and execution.
• Method
  An executable method of an executorch program. Maps to a python method like forward() on the original nn.Module.
• MethodMeta
  Describes a a method in an ExecuTorch program.
• Module
  A facade class for loading programs and executing methods within them.
• Program
  A deserialized ExecuTorch program binary.
• Result
  Result type wrapping either a value of type T or an error.
• Span
  Represent a reference to an array (0 or more elements consecutively in memory), i.e. a start pointer and a length. It allows various APIs to take consecutive elements easily and conveniently.
• Tensor
  A minimal Tensor type whose API is a source compatible subset of at::Tensor.
• TensorImpl
  Manages the storage behind an ETensor (torch::executor::Tensor).
• TensorInfo
  Metadata about a specific tensor of an ExecuTorch Program.
• optional
  Leaner optional class, subset of c10, std, and boost optional APIs.
• optional_trivial_init_t

Enums

• Error
  ExecuTorch Error type.
• Module_MlockConfig
  Enum to define memory locking behavior.
• Program_HeaderStatus
  Describes the presence of an ExecuTorch program header.
• Program_Verification
  Types of validation that the Program can do before parsing the data.
• ScalarType
  Data types (dtypes) that can be used as element types in ETensors.
• Tag
• TensorShapeDynamism
  Rank can never change in executorch. But shape sometimes can. This enum is used to help provide better safety in kernels on what tensors are resizable

Constants

• DelegateDebugIdType_kInt
  Indicates a delegate event logged using an integer delegate debug identifier.
• DelegateDebugIdType_kNone
  Default value, indicates that it’s not a delegate event.
• DelegateDebugIdType_kStr
  Indicates a delegate event logged using a string delegate debug identifier i.e. the delegate debug id is a pointer to a string table managed by the class implementing EventTracer functionality.
• HierarchicalAllocator_kSpanArraySize
• LoggedEValueType_kIntermediateOutput
  Intermediate output from an operator.
• LoggedEValueType_kProgramOutput
  Output at the program level. This is essentially the output of the model.
• MemoryAllocator_kDefaultAlignment
  Default alignment of memory returned by this class. Ensures that pointer fields of structs will be aligned. Larger types like long double may not be, however, depending on the toolchain and architecture.

Functions

• MethodMeta_input_tag^⚠
  Get the tag of the specified input.
• MethodMeta_input_tensor_meta^⚠
  Get metadata about the specified input.
• MethodMeta_memory_planned_buffer_size^⚠
  Get the size in bytes of the specified memory-planned buffer.
• MethodMeta_name^⚠
  Get the name of this method.
• MethodMeta_num_inputs^⚠
  Get the number of inputs to this method.
• MethodMeta_num_memory_planned_buffers^⚠
  Get the number of memory-planned buffers this method requires.
• MethodMeta_num_outputs^⚠
  Get the number of outputs to this method.
• MethodMeta_output_tag^⚠
  Get the tag of the specified output.
• MethodMeta_output_tensor_meta^⚠
  Get metadata about the specified output.
• Method_Method_destructor^⚠
• Method_execute^⚠
  Execute the method.
• Method_experimental_reset_execution^⚠
  Resets execution state to the start of the Method. For use with the experimental_step() API.
• Method_experimental_step^⚠
  Advances/executes a single instruction in the method.
• Method_get_event_tracer^⚠
• Method_get_input^⚠
  DEPRECATED: Use MethodMeta instead to access metadata, and set_input to update Method inputs.
• Method_get_inputs^⚠
  Copies the method’s inputs into the provided array.
• Method_get_output^⚠
  Retrieves the output at the specified index.
• Method_get_outputs^⚠
  Copies the method’s outputs into the provided array.
• Method_inputs_size^⚠
  Returns the number of inputs the Method expects.
• Method_method_meta^⚠
  Returns the MethodMeta that corresponds to the calling Method.
• Method_mutable_input^⚠
  DEPRECATED: Use MethodMeta instead to access metadata, and set_input to update Method inputs.
• Method_mutable_output^⚠
  DEPRECATED: Use MethodMeta instead to access metadata, and get_output to retrieve Method outputs.
• Method_outputs_size^⚠
  Returns the number of outputs the Method returns.
• Method_set_input^⚠
  Sets the internal input value to be equivalent to the to the provided value.
• Method_set_inputs^⚠
  Sets the values of all method inputs.
• Method_set_output_data_ptr^⚠
  Sets the data buffer of the specified method output to the provided value.
• Module_Module1^⚠
  Constructs an instance with the provided data loader and memory allocator.
• Module_is_loaded^⚠
  Checks if the program is loaded.
• Module_load^⚠
  Loads the program using the specified data loader and memory allocator.
• Module_method_names^⚠
  Get a list of method names available in the loaded program. Loads the program and method if needed.
• Program_check_header^⚠
  Looks for an ExecuTorch program header in the provided data.
• Program_get_constant_buffer_data^⚠
  Get the constant buffer inside Program with index buffer_idx. @param[in] buffer_idx the index of the buffer in the constant_buffer. @param[in] nbytes the number of bytes to read from the buffer. @return The buffer with corresponding index.
• Program_get_method_name^⚠
  Returns the name of the method at particular index.
• Program_get_non_const_buffer_size^⚠
  DEPRECATED: Use MethodMeta instead.
• Program_get_output_flattening_encoding^⚠
  DEPRECATED: Get the pytree encoding string for the output. Deprecated as this functionality will eventually move out of the core program into a higher level structure, but that does not exist at this time. @param[in] method_name The name of the method to get the encoding for.
• Program_load^⚠
  Loads a Program from the provided loader. The Program will hold a pointer to the loader, which must outlive the returned Program instance.
• Program_load_method^⚠
  Loads the named method and prepares it for execution.
• Program_method_meta^⚠
  Gathers metadata for the named method.
• Program_num_methods^⚠
  Returns the number of methods in the program.
• Program_num_non_const_buffers^⚠
  DEPRECATED: Use MethodMeta instead.
• TensorImpl_TensorImpl^⚠
  @param type The type of the data (int, float, bool). @param dim Number of dimensions, and the length of the sizes array. @param sizes Sizes of the tensor at each dimension. Must contain dim entries. @param data Pointer to the data, whose size is determined by type, dim, and sizes. The tensor will not own this memory. @param dim_order Order in which dimensions are laid out in memory. @param strides Strides of the tensor at each dimension. Must contain dim entries. @param dynamism The mutability of the shape of the tensor.
• TensorImpl_element_size^⚠
  Returns the size in bytes of one element of the tensor.
• TensorImpl_nbytes^⚠
  Returns the size of the tensor in bytes.
• TensorInfo_dim_order^⚠
  Returns the dim order of the tensor.
• TensorInfo_nbytes^⚠
  Returns the size of the tensor in bytes.
• TensorInfo_scalar_type^⚠
  Returns the scalar type of the input/output.
• TensorInfo_sizes^⚠
  Returns the sizes of the tensor.

Type Aliases

• AllocatorID
  Represents an allocator id returned by track_allocator.
• ArrayRef_const_iterator
• ArrayRef_iterator
• ArrayRef_size_type
• ArrayRef_value_type
• ChainID
  Represents the chain id that will be passed in by the user during event logging.
• DebugHandle
  Represents the debug handle that is generally associated with each op executed in the runtime.
• DelegateDebugIdType
  Different types of delegate debug identifiers that are supported currently.
• FreeableBuffer_FreeFn
• LoggedEValueType
  Indicates the type of the EValue that was logged. These values could be serialized and should not be changed.
• Result_value_type
  value_type member for generic programming.
• Span_iterator
• Span_size_type
• TensorImpl_DimOrderType
  The type used for elements of dim_order().
• TensorImpl_SizesType
  The type used for elements of sizes().
• TensorImpl_StridesType
  The type used for elements of strides().
• Tensor_DimOrderType
  The type used for elements of dim_order().
• Tensor_SizesType
  The type used for elements of sizes().
• Tensor_StridesType
  The type used for elements of strides().
• error_code_t
• optional_value_type
  The type wrapped by the optional class.

Unions

• EValue_Payload
• EValue_Payload_TriviallyCopyablePayload
• Result__bindgen_ty_1
• optional_storage_t
  A wrapper type that lets us avoid constructing a T when there is no value. If there is a value present, the optional class must destroy it.