diff options
Diffstat (limited to 'model-integration/src/main/protobuf')
-rw-r--r-- | model-integration/src/main/protobuf/onnx.proto | 464 |
1 files changed, 464 insertions, 0 deletions
diff --git a/model-integration/src/main/protobuf/onnx.proto b/model-integration/src/main/protobuf/onnx.proto new file mode 100644 index 00000000000..dc6542867e0 --- /dev/null +++ b/model-integration/src/main/protobuf/onnx.proto @@ -0,0 +1,464 @@ +// +// WARNING: This file is automatically generated! Please edit onnx.in.proto. +// + + +// Copyright (c) Facebook Inc. and Microsoft Corporation. +// Licensed under the MIT license. + +syntax = "proto2"; + +package onnx; + +// Overview +// +// ONNX is an open specification that is comprised of the following components: +// +// 1) A definition of an extensible computation graph model. +// 2) Definitions of standard data types. +// 3) Definitions of built-in operators. +// +// This document describes the syntax of models and their computation graphs, +// as well as the standard data types. Together, they are referred to as the ONNX +// Intermediate Representation, or 'IR' for short. +// +// The normative semantic specification of the ONNX IR is found in docs/IR.md. +// Definitions of the built-in neural network operators may be found in docs/Operators.md. + +// Notes +// +// Release +// +// We are still in the very early stage of defining ONNX. The current +// version of ONNX is a starting point. While we are actively working +// towards a complete spec, we would like to get the community involved +// by sharing our working version of ONNX. +// +// Protobuf compatibility +// +// To simplify framework compatibility, ONNX is defined using the subset of protobuf +// that is compatible with both protobuf v2 and v3. This means that we do not use any +// protobuf features that are only available in one of the two versions. +// +// Here are the most notable contortions we have to carry out to work around +// these limitations: +// +// - No 'map' (added protobuf 3.0). We instead represent mappings as lists +// of key-value pairs, where order does not matter and duplicates +// are not allowed. + + +// Versioning +// +// ONNX versioning is specified in docs/IR.md and elaborated on in docs/Versioning.md +// +// To be compatible with both proto2 and proto3, we will use a version number +// that is not defined by the default value but an explicit enum number. +enum Version { + // proto3 requires the first enum value to be zero. + // We add this just to appease the compiler. + _START_VERSION = 0; + // The version field is always serialized and we will use it to store the + // version that the graph is generated from. This helps us set up version + // control. We should use version as + // xx(major) - xx(minor) - xxxx(bugfix) + // and we are starting with 0x00000001 (0.0.1), which was the + // version we published on Oct 10, 2017. + IR_VERSION_2017_10_10 = 0x00000001; + + // IR_VERSION 0.0.2 published on Oct 30, 2017 + // - Added type discriminator to AttributeProto to support proto3 users + IR_VERSION_2017_10_30 = 0x00000002; + + // IR VERSION 0.0.3 published on Nov 3, 2017 + // - For operator versioning: + // - Added new message OperatorSetIdProto + // - Added opset_import in ModelProto + // - For vendor extensions, added domain in NodeProto + IR_VERSION = 0x00000003; +} + +// Attributes +// +// A named attribute containing either singular float, integer, string, graph, +// and tensor values, or repeated float, integer, string, graph, and tensor values. +// An AttributeProto MUST contain the name field, and *only one* of the +// following content fields, effectively enforcing a C/C++ union equivalent. +message AttributeProto { + + // Note: this enum is structurally identical to the OpSchema::AttrType + // enum defined in schema.h. If you rev one, you likely need to rev the other. + enum AttributeType { + UNDEFINED = 0; + FLOAT = 1; + INT = 2; + STRING = 3; + TENSOR = 4; + GRAPH = 5; + + FLOATS = 6; + INTS = 7; + STRINGS = 8; + TENSORS = 9; + GRAPHS = 10; + } + + // The name field MUST be present for this version of the IR. + optional string name = 1; // namespace Attribute + + // if ref_attr_name is not empty, ref_attr_name is the attribute name in parent function. + // In this case, this AttributeProto does not contain data, and it's a reference of attribute + // in parent scope. + // NOTE: This should ONLY be used in function (sub-graph). It's invalid to be used in main graph. + optional string ref_attr_name = 21; + + // A human-readable documentation for this attribute. Markdown is allowed. + optional string doc_string = 13; + + // The type field MUST be present for this version of the IR. + // For 0.0.1 versions of the IR, this field was not defined, and + // implementations needed to use has_field hueristics to determine + // which value field was in use. For IR_VERSION 0.0.2 or later, this + // field MUST be set and match the f|i|s|t|... field in use. This + // change was made to accomodate proto3 implementations. + optional AttributeType type = 20; // discriminator that indicates which field below is in use + + // Exactly ONE of the following fields must be present for this version of the IR + optional float f = 2; // float + optional int64 i = 3; // int + optional bytes s = 4; // UTF-8 string + optional TensorProto t = 5; // tensor value + optional GraphProto g = 6; // graph + // Do not use field below, it's deprecated. + // optional ValueProto v = 12; // value - subsumes everything but graph + + repeated float floats = 7; // list of floats + repeated int64 ints = 8; // list of ints + repeated bytes strings = 9; // list of UTF-8 strings + repeated TensorProto tensors = 10; // list of tensors + repeated GraphProto graphs = 11; // list of graph +} + +// Defines information on value, including the name, the type, and +// the shape of the value. +message ValueInfoProto { + // This field MUST be present in this version of the IR. + optional string name = 1; // namespace Value + // This field MUST be present in this version of the IR. + optional TypeProto type = 2; + // A human-readable documentation for this value. Markdown is allowed. + optional string doc_string = 3; +} + +// Nodes +// +// Computation graphs are made up of a DAG of nodes, which represent what is +// commonly called a "layer" or "pipeline stage" in machine learning frameworks. +// +// For example, it can be a node of type "Conv" that takes in an image, a filter +// tensor and a bias tensor, and produces the convolved output. +message NodeProto { + repeated string input = 1; // namespace Value + repeated string output = 2; // namespace Value + + // An optional identifier for this node in a graph. + // This field MAY be absent in ths version of the IR. + optional string name = 3; // namespace Node + + // The symbolic identifier of the Operator to execute. + optional string op_type = 4; // namespace Operator + // The domain of the OperatorSet that specifies the operator named by op_type. + optional string domain = 7; // namespace Domain + + // Additional named attributes. + repeated AttributeProto attribute = 5; + + // A human-readable documentation for this node. Markdown is allowed. + optional string doc_string = 6; +} + +// Models +// +// ModelProto is a top-level file/container format for bundling a ML model and +// associating its computation graph with metadata. +// +// The semantics of the model are described by the associated GraphProto. +message ModelProto { + // The version of the IR this model targets. See Version enum above. + // This field MUST be present. + optional int64 ir_version = 1; + + // The OperatorSets this model relies on. + // All ModelProtos MUST have at least one entry that + // specifies which version of the ONNX OperatorSet is + // being imported. + // + // All nodes in the ModelProto's graph will bind against the operator + // with the same-domain/same-op_type operator with the HIGHEST version + // in the referenced operator sets. + repeated OperatorSetIdProto opset_import = 8; + + // The name of the framework or tool used to generate this model. + // This field SHOULD be present to indicate which implementation/tool/framework + // emitted the model. + optional string producer_name = 2; + + // The version of the framework or tool used to generate this model. + // This field SHOULD be present to indicate which implementation/tool/framework + // emitted the model. + optional string producer_version = 3; + + // Domain name of the model. + // We use reverse domain names as name space indicators. For example: + // `com.facebook.fair` or `com.microsoft.cognitiveservices` + // + // Together with `model_version` and GraphProto.name, this forms the unique identity of + // the graph. + optional string domain = 4; + + // The version of the graph encoded. See Version enum below. + optional int64 model_version = 5; + + // A human-readable documentation for this model. Markdown is allowed. + optional string doc_string = 6; + + // The parameterized graph that is evaluated to execute the model. + optional GraphProto graph = 7; + + // Named metadata values; keys should be distinct. + repeated StringStringEntryProto metadata_props = 14; +}; + +// StringStringEntryProto follows the pattern for cross-proto-version maps. +// See https://developers.google.com/protocol-buffers/docs/proto3#maps +message StringStringEntryProto { + optional string key = 1; + optional string value= 2; +}; + +// Graphs +// +// A graph defines the computational logic of a model and is comprised of a parameterized +// list of nodes that form a directed acyclic graph based on their inputs and outputs. +// This is the equivalent of the "network" or "graph" in many deep learning +// frameworks. +message GraphProto { + // The nodes in the graph, sorted topologically. + repeated NodeProto node = 1; + + // The name of the graph. + optional string name = 2; // namespace Graph + + // A list of named tensor values, used to specify constant inputs of the graph. + // Each TensorProto entry must have a distinct name (within the list) that + // also appears in the input list. + repeated TensorProto initializer = 5; + + // A human-readable documentation for this graph. Markdown is allowed. + optional string doc_string = 10; + + // The inputs and outputs of the graph. + repeated ValueInfoProto input = 11; + repeated ValueInfoProto output = 12; + + // Information for the values in the graph. The ValueInfoProto.name's + // must be distinct. It is optional for a value to appear in value_info list. + repeated ValueInfoProto value_info = 13; + + // DO NOT USE the following fields, they were deprecated from earlier versions. + // repeated string input = 3; + // repeated string output = 4; + // optional int64 ir_version = 6; + // optional int64 producer_version = 7; + // optional string producer_tag = 8; + // optional string domain = 9; +} + +// Tensors +// +// A serialized tensor value. +message TensorProto { + enum DataType { + UNDEFINED = 0; + // Basic types. + FLOAT = 1; // float + UINT8 = 2; // uint8_t + INT8 = 3; // int8_t + UINT16 = 4; // uint16_t + INT16 = 5; // int16_t + INT32 = 6; // int32_t + INT64 = 7; // int64_t + STRING = 8; // string + BOOL = 9; // bool + + // Advanced types + FLOAT16 = 10; + DOUBLE = 11; + UINT32 = 12; + UINT64 = 13; + COMPLEX64 = 14; // complex with float32 real and imaginary components + COMPLEX128 = 15; // complex with float64 real and imaginary components + // Future extensions go here. + } + + // The shape of the tensor. + repeated int64 dims = 1; + + // The data type of the tensor. + optional DataType data_type = 2; + + // For very large tensors, we may want to store them in chunks, in which + // case the following fields will specify the segment that is stored in + // the current TensorProto. + message Segment { + optional int64 begin = 1; + optional int64 end = 2; + } + optional Segment segment = 3; + + // Tensor content must be organized in row-major order. + // + // Depending on the data_type field, exactly one of the fields below with + // name ending in _data is used to store the elements of the tensor. + + // For float and complex64 values + // Complex64 tensors are encoded as a single array of floats, + // with the real components appearing in odd numbered positions, + // and the corresponding imaginary component apparing in the + // subsequent even numbered position. (e.g., [1.0 + 2.0i, 3.0 + 4.0i] + // is encoded as [1.0, 2.0 ,3.0 ,4.0] + // When this field is present, the data_type field MUST be FLOAT or COMPLEX64. + repeated float float_data = 4 [packed = true]; + + // For int32, uint8, int8, uint16, int16, bool, and float16 values + // float16 values must be bit-wise converted to an uint16_t prior + // to writing to the buffer. + // When this field is present, the data_type field MUST be + // INT32, INT16, INT8, UINT16, INT8, BOOL, or FLOAT32 + repeated int32 int32_data = 5 [packed = true]; + + // For strings. + // Each element of string_data is a UTF-8 encoded Unicode + // string. No trailing null, no leading BOM. The protobuf "string" + // scalar type is not used to match ML community conventions. + // When this field is present, the data_type field MUST be STRING + repeated bytes string_data = 6; + + // For int64. + // When this field is present, the data_type field MUST be INT64 + repeated int64 int64_data = 7 [packed = true]; + + // Optionally, a name for the tensor. + optional string name = 8; // namespace Value + + // A human-readable documentation for this tensor. Markdown is allowed. + optional string doc_string = 12; + + // Serializations can either use one of the fields above, or use this + // raw bytes field. The only exception is the string case, where one is + // required to store the content in the repeated bytes string_data field. + // + // When this raw_data field is used to store tensor value, elements MUST + // be stored in as fixed-width, little-endian order. + // Floating-point data types MUST be stored in IEEE 754 format. + // Complex64 elements must be written as two consecutive FLOAT values, real component first. + // Complex128 elements must be written as two consecutive DOUBLE values, real component first. + // Boolean type MUST be written one byte per tensor element (00000001 for true, 00000000 for false). + // + // Note: the advantage of specific field rather than the raw_data field is + // that in some cases (e.g. int data), protobuf does a better packing via + // variable length storage, and may lead to smaller binary footprint. + // When this field is present, the data_type field MUST NOT be STRING or UNDEFINED + optional bytes raw_data = 9; + + // For double + // Complex64 tensors are encoded as a single array of doubles, + // with the real components appearing in odd numbered positions, + // and the corresponding imaginary component apparing in the + // subsequent even numbered position. (e.g., [1.0 + 2.0i, 3.0 + 4.0i] + // is encoded as [1.0, 2.0 ,3.0 ,4.0] + // When this field is present, the data_type field MUST be DOUBLE or COMPLEX128 + repeated double double_data = 10 [packed = true]; + + // For uint64 and uint32 values + // When this field is present, the data_type field MUST be + // UINT32 or UINT64 + repeated uint64 uint64_data = 11 [packed = true]; +} + +// Defines a tensor shape. A dimension can be either an integer value +// or a symbolic variable. A symbolic variable represents an unknown +// dimension. +message TensorShapeProto { + message Dimension { + oneof value { + int64 dim_value = 1; + string dim_param = 2; // namespace Shape + }; + // Standard denotation can optionally be used to denote tensor + // dimensions with standard semantic descriptions to ensure + // that operations are applied to the correct axis of a tensor. + optional string denotation = 3; + }; + repeated Dimension dim = 1; +} + +// A set of pre-defined constants to be used as values for +// the standard denotation field in TensorShapeProto.Dimension +// for semantic description of the tensor dimension. +message DenotationConstProto { + // Describe a batch number dimension. + optional string DATA_BATCH = 1 [default = "DATA_BATCH"]; + // Describe a channel dimension. + optional string DATA_CHANNEL = 2 [default = "DATA_CHANNEL"]; + // Describe a time dimension. + optional string DATA_TIME = 3 [default = "DATA_TIME"]; + // Describe a feature dimension. This is typically a feature + // dimension in RNN and/or spatial dimension in CNN. + optional string DATA_FEATURE = 4 [default = "DATA_FEATURE"]; + // Describe a filter in-channel dimension. This is the dimension + // that is identical (in size) to the channel dimension of the input + // image feature maps. + optional string FILTER_IN_CHANNEL = 5 [default = "FILTER_IN_CHANNEL"]; + // Describe a filter out channel dimension. This is the dimension + // that is identical (int size) to the channel dimension of the output + // image feature maps. + optional string FILTER_OUT_CHANNEL = 6 [default = "FILTER_OUT_CHANNEL"]; + // Describe a filter spatial dimension. + optional string FILTER_SPATIAL = 7 [default = "FILTER_SPATIAL"]; +} + +// Types +// +// The standard ONNX data types. +message TypeProto { + + message Tensor { + // This field MUST NOT have the value of UNDEFINED + // This field MUST be present for this version of the IR. + optional TensorProto.DataType elem_type = 1; + optional TensorShapeProto shape = 2; + } + + + oneof value { + // The type of a tensor. + Tensor tensor_type = 1; + + } +} + +// Operator Sets +// +// OperatorSets are uniquely identified by a (domain, opset_version) pair. +message OperatorSetIdProto { + // The domain of the operator set being identified. + // The empty string ("") or absence of this field implies the operator + // set that is defined as part of the ONNX specification. + // This field MUST be present in this version of the IR when referring to any other operator set. + optional string domain = 1; + + // The version of the operator set being identified. + // This field MUST be present in this version of the IR. + optional int64 version = 2; +}
\ No newline at end of file |