package onnx

Get desktop application:
View/edit binary Protocol Buffers messages

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.

Used in: NodeProto

• string name = 1

  The name field MUST be present for this version of the IR.

  namespace Attribute

• string ref_attr_name = 21

  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.

• string doc_string = 13

  A human-readable documentation for this attribute. Markdown is allowed.

• AttributeProto.AttributeType type = 20

  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 heuristics 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 accommodate proto3 implementations.

  discriminator that indicates which field below is in use

• float f = 2

  Exactly ONE of the following fields must be present for this version of the IR

  float

• int64 i = 3

  int

• bytes s = 4

  UTF-8 string

• optional TensorProto t = 5

  tensor value

• optional GraphProto g = 6

  graph

• optional SparseTensorProto sparse_tensor = 22

  sparse tensor value

• optional TypeProto tp = 14

  Do not use field below, it's deprecated. optional ValueProto v = 12; // value - subsumes everything but graph

  type proto

• 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

• repeated SparseTensorProto sparse_tensors = 23

  list of sparse tensors

• repeated TypeProto type_protos = 15

  list of type protos

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.

Used in: AttributeProto

• UNDEFINED = 0

• FLOAT = 1

• INT = 2

• STRING = 3

• TENSOR = 4

• GRAPH = 5

• SPARSE_TENSOR = 11

• TYPE_PROTO = 13

• FLOATS = 6

• INTS = 7

• STRINGS = 8

• TENSORS = 9

• GRAPHS = 10

• SPARSE_TENSORS = 12

• TYPE_PROTOS = 14

Used in: ModelProto, OperatorSetProto

• string name = 1

  The name of the function, similar usage of op_type in OperatorProto. Combined with FunctionProto.domain, this forms the unique identity of the FunctionProto.

• repeated string input = 4

  The inputs and outputs of the function.

• repeated string output = 5

• repeated string attribute = 6

  The attributes of the function.

• repeated NodeProto node = 7

  The nodes in the function.

• string doc_string = 8

  A human-readable documentation for this function. Markdown is allowed.

• repeated OperatorSetIdProto opset_import = 9

• string domain = 10

  The domain which this function belongs to. Combined with FunctionProto.name, this forms the unique identity of the FunctionProto.

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.

Used in: AttributeProto, ModelProto, TrainingInfoProto

• repeated NodeProto node = 1

  The nodes in the graph, sorted topologically.

• string name = 2

  The name of the graph.

  namespace Graph

• repeated TensorProto initializer = 5

  A list of named tensor values, used to specify constant inputs of the graph. Each initializer (both TensorProto as well SparseTensorProto) MUST have a name. The name MUST be unique across both initializer and sparse_initializer, but the name MAY also appear in the input list.

• repeated SparseTensorProto sparse_initializer = 15

  Initializers (see above) stored in sparse format.

• string doc_string = 10

  A human-readable documentation for this graph. Markdown is allowed.

• repeated ValueInfoProto input = 11

  The inputs and outputs of the graph.

• repeated ValueInfoProto output = 12

• repeated ValueInfoProto value_info = 13

  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 TensorAnnotation quantization_annotation = 14

  This field carries information to indicate the mapping among a tensor and its quantization parameter tensors. For example: For tensor 'a', it may have {'SCALE_TENSOR', 'a_scale'} and {'ZERO_POINT_TENSOR', 'a_zero_point'} annotated, which means, tensor 'a_scale' and tensor 'a_zero_point' are scale and zero point of tensor 'a' in the model.

Maps Specifies an associative table, defined by keys and values. MapProto is formed with a repeated field of keys (of type INT8, INT16, INT32, INT64, UINT8, UINT16, UINT32, UINT64, or STRING) and values (of type TENSOR, SPARSE_TENSOR, SEQUENCE, or MAP). Key types and value types have to remain the same throughout the instantiation of the MapProto.

Used in: OptionalProto, SequenceProto

• string name = 1

• int32 key_type = 2

  The data type of the key. This field MUST have a valid TensorProto.DataType value of INT8, INT16, INT32, INT64, UINT8, UINT16, UINT32, UINT64, or STRING

• repeated int64 keys = 3

  Every element of keys has to be one of the following data types INT8, INT16, INT32, INT64, UINT8, UINT16, UINT32, UINT64, or STRING. The integer cases are represented by the repeated int64 field keys below.

• repeated bytes string_keys = 4

  If keys are strings, they are represented by the repeated bytes field string_keys below.

• optional SequenceProto values = 5

  MapProto values are represented in a SequenceProto of the same length as the repeated keys field and have to be one of the following data types TENSOR, SPARSE_TENSOR, MAP, SEQUENCE.

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's.

• int64 ir_version = 1

  The version of the IR this model targets. See Version enum above. This field MUST be present.

• repeated OperatorSetIdProto opset_import = 8

  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.

• string producer_name = 2

  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.

• string producer_version = 3

  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.

• string domain = 4

  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.

• int64 model_version = 5

  The version of the graph encoded. See Version enum below.

• string doc_string = 6

  A human-readable documentation for this model. Markdown is allowed.

• optional GraphProto graph = 7

  The parameterized graph that is evaluated to execute the model.

• repeated StringStringEntryProto metadata_props = 14

  Named metadata values; keys should be distinct.

• repeated TrainingInfoProto training_info = 20

  Training-specific information. Sequentially executing all stored `TrainingInfoProto.algorithm`s and assigning their outputs following the corresponding `TrainingInfoProto.update_binding`s is one training iteration. Similarly, to initialize the model (as if training hasn't happened), the user should sequentially execute all stored `TrainingInfoProto.initialization`s and assigns their outputs using `TrainingInfoProto.initialization_binding`s. If this field is empty, the training behavior of the model is undefined.

• repeated FunctionProto functions = 25

  A list of function protos local to the model. Name of the function "FunctionProto.name" should be unique within the domain "FunctionProto.domain". In case of any conflicts the behavior (whether the model local functions are given higher priority, or standard opserator sets are given higher priotity or this is treated as error) is defined by the runtimes. The operator sets imported by FunctionProto should be compatible with the ones imported by ModelProto and other model local FunctionProtos. Example, if same operator set say 'A' is imported by a FunctionProto and ModelProto or by 2 FunctionProtos then versions for the operator set may be different but, the operator schema returned for op_type, domain, version combination for both the versions should be same for every node in the function body. One FunctionProto can reference other FunctionProto in the model, however, recursive reference is not allowed.

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.

Used in: FunctionProto, GraphProto

• repeated string input = 1

  namespace Value

• repeated string output = 2

  namespace Value

• string name = 3

  An optional identifier for this node in a graph. This field MAY be absent in ths version of the IR.

  namespace Node

• string op_type = 4

  The symbolic identifier of the Operator to execute.

  namespace Operator

• string domain = 7

  The domain of the OperatorSet that specifies the operator named by op_type.

  namespace Domain

• repeated AttributeProto attribute = 5

  Additional named attributes.

• string doc_string = 6

  A human-readable documentation for this node. Markdown is allowed.

An OperatorProto represents the immutable specification of the signature and semantics of an operator. Operators are declared as part of an OperatorSet, which also defines the domain name for the set. Operators are uniquely identified by a three part identifier (domain, op_type, since_version) where *domain* is the domain of an operator set that contains this operator specification. *op_type* is the name of the operator as referenced by a NodeProto.op_type *since_version* is the version of the operator set that this operator was initially declared in.

Used in: OperatorSetProto

• string op_type = 1

  The name of the operator within a domain. This field MUST be present in this version of the IR.

• int64 since_version = 2

  The version of the operator set that first introduced this operator. This value MUST be the same value as the opset_version of the operator set that first published this operator. Subsequent versions of the operator set MUST NOT alter the signature or semantics of the operator once published as STABLE. This field MUST be present in this version of the IR.

• OperatorStatus status = 3

  This field indicates whether the syntax, semantics, or presence of this operator is in an experimental or stable stage. Once an operator is published as STABLE, it's syntax and semantics MUST NOT change in subsequent versions of the operator set. When an operator is published as EXPERIMENTAL, the syntax and semantics of the operator MAY change across operator set versions. Operators "become" stable by deprecating the experimental version and introducing a new stable operator with the same op_type.

• string doc_string = 10

  A human-readable documentation for this operator. Markdown is allowed.

Operator Sets OperatorSets are uniquely identified by a (domain, opset_version) pair.

Used in: FunctionProto, ModelProto

• string domain = 1

  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.

• int64 version = 2

  The version of the operator set being identified. This field MUST be present in this version of the IR.

An OperatorSetProto represents an immutable set of immutable operator specifications. The domain of the set (OperatorSetProto.domain) is a reverse-DNS name that disambiguates operator sets defined by independent entities. The version of the set (opset_version) is a monotonically increasing integer that indicates changes to the membership of the operator set. Operator sets are uniquely identified by a two part identifier (domain, opset_version) Like ModelProto, OperatorSetProto is intended as a top-level file/wire format, and thus has the standard format headers in addition to the operator set information.

• string magic = 1

  All OperatorSetProtos start with a distingushed byte sequence to disambiguate protobuf files containing OperatorSets from other content. This field MUST be "ONNXOPSET" This field MUST be present in this version of the IR

• int64 ir_version = 2

  All OperatorSetProtos indicate the version of the IR syntax and semantics they adhere to. It is always IR_VERSION. This field MUST be present in this version of the IR

• string ir_version_prerelease = 3

  The prerelease component of the SemVer of the IR. This field MAY be absent in this version of the IR

• string ir_build_metadata = 7

  The build metadata component of the SemVer of the IR. This field MAY be absent in this version of the IR

• string domain = 4

  Domain name of the operator set, in reverse DNS form (e.g., com.acme.dnnops).

• int64 opset_version = 5

  The version of the set of operators. This is a simple int value that is monotonically increasing as new versions of the operator set are published. All operators in this set MUST have since_version <= opset_version.

• string doc_string = 6

  A human-readable documentation for this set of operators. Markdown is allowed.

• repeated OperatorProto operator = 8

  The operators specified by this operator set. The (name, version) MUST be unique across all OperatorProtos in operator

• repeated FunctionProto functions = 9

  The functions specified by this operator set. The (name, version) MUST be unique across all OperatorProtos/FunctionProtos in operator/functions

Operator/function status.

Used in: OperatorProto

• EXPERIMENTAL = 0

• STABLE = 1

Optional

Used in: SequenceProto

• string name = 1

• int32 elem_type = 2

  The data type of the element, identifies if the OptionalProto value is Tensor, Sparse Tensor, Sequence, Map, or Optional. The type of the optional value MUST match the elem_type specified. This field MUST have a valid OptionalProto.DataType value.

• optional TensorProto tensor_value = 3

  For TensorProto value. When this field is present, the elem_type field MUST be TENSOR.

• optional SparseTensorProto sparse_tensor_value = 4

  For SparseTensorProto value. When this field is present, the elem_type field MUST be SPARSE_TENSOR.

• optional SequenceProto sequence_value = 5

  For SequenceProto value. When this field is present, the elem_type field MUST be SEQUENCE.

• optional MapProto map_value = 6

  For MapProto value. When this field is present, the elem_type field MUST be MAP.

• optional OptionalProto optional_value = 7

  For OptionalProto value, allowing optional to be of itself (completeness) When this field is present, the elem_type field MUST be OPTIONAL.

• UNDEFINED = 0

• TENSOR = 1

• SPARSE_TENSOR = 2

• SEQUENCE = 3

• MAP = 4

• OPTIONAL = 5

Sequences Defines a dense, ordered, collection of elements that are of homogeneous types. Sequences can be made out of tensors, maps, or sequences. If a sequence is made out of tensors, the tensors must have the same element type (i.e. int32). In some cases, the tensors in a sequence can have different shapes. Whether the tensors can have different shapes or not depends on the type/shape associated with the corresponding "ValueInfo". For example, "Sequence<Tensor<float, [M,N]>" means that all tensors have same shape. However, "Sequence<Tensor<float, [omitted,omitted]>" means they can have different shapes (all of rank 2), where "omitted" means the corresponding dimension has no symbolic/constant value. Finally, "Sequence<Tensor<float, omitted>>" means that the different tensors can have different ranks, when the "shape" itself is omitted from the tensor-type. For a more complete description, refer to https://github.com/onnx/onnx/blob/master/docs/IR.md#static-tensor-shapes.

Used in: MapProto, OptionalProto

• string name = 1

• int32 elem_type = 2

  The data type of the element. This field MUST have a valid SequenceProto.DataType value

• repeated TensorProto tensor_values = 3

  For TensorProto values. When this field is present, the elem_type field MUST be TENSOR.

• repeated SparseTensorProto sparse_tensor_values = 4

  For SparseTensorProto values. When this field is present, the elem_type field MUST be SPARSE_TENSOR.

• repeated SequenceProto sequence_values = 5

  For SequenceProto values, allowing sequences to be of themselves. When this field is present, the elem_type field MUST be SEQUENCE.

• repeated MapProto map_values = 6

  For MapProto values. When this field is present, the elem_type field MUST be MAP.

• repeated OptionalProto optional_values = 7

  For OptionalProto values. When this field is present, the elem_type field MUST be Optional.

• UNDEFINED = 0

• TENSOR = 1

• SPARSE_TENSOR = 2

• SEQUENCE = 3

• MAP = 4

• OPTIONAL = 5

A serialized sparse-tensor value

Used in: AttributeProto, GraphProto, OptionalProto, SequenceProto

• optional TensorProto values = 1

  The sequence of non-default values are encoded as a tensor of shape [NNZ]. The default-value is zero for numeric tensors, and empty-string for string tensors. values must have a non-empty name present which serves as a name for SparseTensorProto when used in sparse_initializer list.

• optional TensorProto indices = 2

  The indices of the non-default values, which may be stored in one of two formats. (a) Indices can be a tensor of shape [NNZ, rank] with the [i,j]-th value corresponding to the j-th index of the i-th value (in the values tensor). (b) Indices can be a tensor of shape [NNZ], in which case the i-th value must be the linearized-index of the i-th value (in the values tensor). The linearized-index can be converted into an index tuple (k_1,...,k_rank) using the shape provided below. The indices must appear in ascending order without duplication. In the first format, the ordering is lexicographic-ordering: e.g., index-value [1,4] must appear before [2,1]

• repeated int64 dims = 3

  The shape of the underlying dense-tensor: [dim_1, dim_2, ... dim_rank]

StringStringEntryProto follows the pattern for cross-proto-version maps. See https://developers.google.com/protocol-buffers/docs/proto3#maps

Used in: ModelProto, TensorAnnotation, TensorProto, TrainingInfoProto

• string key = 1

• string value = 2

Used in: GraphProto

• string tensor_name = 1

• repeated StringStringEntryProto quant_parameter_tensor_names = 2

  <key, value> pairs to annotate tensor specified by <tensor_name> above. The keys used in the mapping below must be pre-defined in ONNX spec. For example, for 8-bit linear quantization case, 'SCALE_TENSOR', 'ZERO_POINT_TENSOR' will be pre-defined as quantization parameter keys.

Tensors A serialized tensor value.

Used in: AttributeProto, GraphProto, OptionalProto, SequenceProto, SparseTensorProto

• repeated int64 dims = 1

  The shape of the tensor.

• int32 data_type = 2

  The data type of the tensor. This field MUST have a valid TensorProto.DataType value

• optional TensorProto.Segment segment = 3

• repeated float float_data = 4

  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 appearing 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 int32 int32_data = 5

  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, UINT8, BOOL, or FLOAT16

• repeated bytes string_data = 6

  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 int64 int64_data = 7

  For int64. When this field is present, the data_type field MUST be INT64

• string name = 8

  Optionally, a name for the tensor.

  namespace Value

• string doc_string = 12

  A human-readable documentation for this tensor. Markdown is allowed.

• bytes raw_data = 9

  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

• repeated StringStringEntryProto external_data = 13

  Data can be stored inside the protobuf file using type-specific fields or raw_data. Alternatively, raw bytes data can be stored in an external file, using the external_data field. external_data stores key-value pairs describing data location. Recognized keys are: - "location" (required) - POSIX filesystem path relative to the directory where the ONNX protobuf model was stored - "offset" (optional) - position of byte at which stored data begins. Integer stored as string. Offset values SHOULD be multiples 4096 (page size) to enable mmap support. - "length" (optional) - number of bytes containing data. Integer stored as string. - "checksum" (optional) - SHA1 digest of file specified in under 'location' key.

• TensorProto.DataLocation data_location = 14

  If value not set, data is stored in raw_data (if set) otherwise in type-specified field.

• repeated double double_data = 10

  For double Complex128 tensors are encoded as a single array of doubles, with the real components appearing in odd numbered positions, and the corresponding imaginary component appearing 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 uint64 uint64_data = 11

  For uint64 and uint32 values When this field is present, the data_type field MUST be UINT32 or UINT64

Location of the data for this tensor. MUST be one of: - DEFAULT - data stored inside the protobuf message. Data is stored in raw_data (if set) otherwise in type-specified field. - EXTERNAL - data stored in an external location as described by external_data field.

Used in: TensorProto

• DEFAULT = 0

• EXTERNAL = 1

• UNDEFINED = 0

• FLOAT = 1

  Basic types.

  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

• FLOAT16 = 10

  IEEE754 half-precision floating-point format (16 bits wide). This format has 1 sign bit, 5 exponent bits, and 10 mantissa bits.

• DOUBLE = 11

• UINT32 = 12

• UINT64 = 13

• COMPLEX64 = 14

  complex with float32 real and imaginary components

• COMPLEX128 = 15

  complex with float64 real and imaginary components

• BFLOAT16 = 16

  Non-IEEE floating-point format based on IEEE754 single-precision floating-point number truncated to 16 bits. This format has 1 sign bit, 8 exponent bits, and 7 mantissa bits.

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.

Used in: TensorProto

• int64 begin = 1

• int64 end = 2

Defines a tensor shape. A dimension can be either an integer value or a symbolic variable. A symbolic variable represents an unknown dimension.

Used in: TypeProto.SparseTensor, TypeProto.Tensor

• repeated TensorShapeProto.Dimension dim = 1

Used in: TensorShapeProto

• oneof value

  • int64 dim_value = 1

  • string dim_param = 2

    namespace Shape

• string denotation = 3

  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. Refer to https://github.com/onnx/onnx/blob/master/docs/DimensionDenotation.md#denotation-definition for pre-defined dimension denotations.

Training information TrainingInfoProto stores information for training a model. In particular, this defines two functionalities: an initialization-step and a training-algorithm-step. Initialization resets the model back to its original state as if no training has been performed. Training algorithm improves the model based on input data. The semantics of the initialization-step is that the initializers in ModelProto.graph and in TrainingInfoProto.algorithm are first initialized as specified by the initializers in the graph, and then updated by the "initialization_binding" in every instance in ModelProto.training_info. The field "algorithm" defines a computation graph which represents a training algorithm's step. After the execution of a TrainingInfoProto.algorithm, the initializers specified by "update_binding" may be immediately updated. If the targeted training algorithm contains consecutive update steps (such as block coordinate descent methods), the user needs to create a TrainingInfoProto for each step.

Used in: ModelProto

• optional GraphProto initialization = 1

  This field describes a graph to compute the initial tensors upon starting the training process. Initialization graph has no input and can have multiple outputs. Usually, trainable tensors in neural networks are randomly initialized. To achieve that, for each tensor, the user can put a random number operator such as RandomNormal or RandomUniform in TrainingInfoProto.initialization.node and assign its random output to the specific tensor using "initialization_binding". This graph can also set the initializers in "algorithm" in the same TrainingInfoProto; a use case is resetting the number of training iteration to zero. By default, this field is an empty graph and its evaluation does not produce any output. Thus, no initializer would be changed by default.

• optional GraphProto algorithm = 2

  This field represents a training algorithm step. Given required inputs, it computes outputs to update initializers in its own or inference graph's initializer lists. In general, this field contains loss node, gradient node, optimizer node, increment of iteration count. An execution of the training algorithm step is performed by executing the graph obtained by combining the inference graph (namely "ModelProto.graph") and the "algorithm" graph. That is, the actual the actual input/initializer/output/node/value_info/sparse_initializer list of the training graph is the concatenation of "ModelProto.graph.input/initializer/output/node/value_info/sparse_initializer" and "algorithm.input/initializer/output/node/value_info/sparse_initializer" in that order. This combined graph must satisfy the normal ONNX conditions. Now, let's provide a visualization of graph combination for clarity. Let the inference graph (i.e., "ModelProto.graph") be tensor_a, tensor_b -> MatMul -> tensor_c -> Sigmoid -> tensor_d and the "algorithm" graph be tensor_d -> Add -> tensor_e The combination process results tensor_a, tensor_b -> MatMul -> tensor_c -> Sigmoid -> tensor_d -> Add -> tensor_e Notice that an input of a node in the "algorithm" graph may reference the output of a node in the inference graph (but not the other way round). Also, inference node cannot reference inputs of "algorithm". With these restrictions, inference graph can always be run independently without training information. By default, this field is an empty graph and its evaluation does not produce any output. Evaluating the default training step never update any initializers.

• repeated StringStringEntryProto initialization_binding = 3

  This field specifies the bindings from the outputs of "initialization" to some initializers in "ModelProto.graph.initializer" and the "algorithm.initializer" in the same TrainingInfoProto. See "update_binding" below for details. By default, this field is empty and no initializer would be changed by the execution of "initialization".

• repeated StringStringEntryProto update_binding = 4

  Gradient-based training is usually an iterative procedure. In one gradient descent iteration, we apply x = x - r * g where "x" is the optimized tensor, "r" stands for learning rate, and "g" is gradient of "x" with respect to a chosen loss. To avoid adding assignments into the training graph, we split the update equation into y = x - r * g x = y The user needs to save "y = x - r * g" into TrainingInfoProto.algorithm. To tell that "y" should be assigned to "x", the field "update_binding" may contain a key-value pair of strings, "x" (key of StringStringEntryProto) and "y" (value of StringStringEntryProto). For a neural network with multiple trainable (mutable) tensors, there can be multiple key-value pairs in "update_binding". The initializers appears as keys in "update_binding" are considered mutable variables. This implies some behaviors as described below. 1. We have only unique keys in all "update_binding"s so that two variables may not have the same name. This ensures that one variable is assigned up to once. 2. The keys must appear in names of "ModelProto.graph.initializer" or "TrainingInfoProto.algorithm.initializer". 3. The values must be output names of "algorithm" or "ModelProto.graph.output". 4. Mutable variables are initialized to the value specified by the corresponding initializer, and then potentially updated by "initializer_binding"s and "update_binding"s in "TrainingInfoProto"s. This field usually contains names of trainable tensors (in ModelProto.graph), optimizer states such as momentums in advanced stochastic gradient methods (in TrainingInfoProto.graph), and number of training iterations (in TrainingInfoProto.graph). By default, this field is empty and no initializer would be changed by the execution of "algorithm".

Types The standard ONNX data types.

Used in: AttributeProto, TypeProto.Map, TypeProto.Optional, TypeProto.Sequence, ValueInfoProto

• oneof value

  • TypeProto.Tensor tensor_type = 1

    The type of a tensor.

  • TypeProto.Sequence sequence_type = 4

    The type of a sequence.

  • TypeProto.Map map_type = 5

    The type of a map.

  • TypeProto.Optional optional_type = 9

    The type of an optional.

  • TypeProto.SparseTensor sparse_tensor_type = 8

    Type of the sparse tensor

• string denotation = 6

  An optional denotation can be used to denote the whole type with a standard semantic description as to what is stored inside. Refer to https://github.com/onnx/onnx/blob/master/docs/TypeDenotation.md#type-denotation-definition for pre-defined type denotations.

map<K,V>

Used in: TypeProto

• int32 key_type = 1

  This field MUST have a valid TensorProto.DataType value This field MUST be present for this version of the IR. This field MUST refer to an integral type ([U]INT{8|16|32|64}) or STRING

• optional TypeProto value_type = 2

  This field MUST be present for this version of the IR.

wrapper for Tensor, Sequence, or Map

Used in: TypeProto

• optional TypeProto elem_type = 1

  The type and optional shape of the element wrapped. This field MUST be present for this version of the IR. Possible values correspond to OptionalProto.DataType enum

repeated T

Used in: TypeProto

• optional TypeProto elem_type = 1

  The type and optional shape of each element of the sequence. This field MUST be present for this version of the IR.

Used in: TypeProto

• int32 elem_type = 1

  This field MUST NOT have the value of UNDEFINED This field MUST have a valid TensorProto.DataType value This field MUST be present for this version of the IR.

• optional TensorShapeProto shape = 2

Used in: TypeProto

• int32 elem_type = 1

  This field MUST NOT have the value of UNDEFINED This field MUST have a valid TensorProto.DataType value This field MUST be present for this version of the IR.

• optional TensorShapeProto shape = 2

Defines information on value, including the name, the type, and the shape of the value.

Used in: GraphProto

• string name = 1

  This field MUST be present in this version of the IR.

  namespace Value

• optional TypeProto type = 2

  This field MUST be present in this version of the IR for inputs and outputs of the top-level graph.

• string doc_string = 3

  A human-readable documentation for this value. Markdown is 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.

• _START_VERSION = 0

  proto3 requires the first enum value to be zero. We add this just to appease the compiler.

• IR_VERSION_2017_10_10 = 1

  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. For the IR, we are using simple numbers starting with 0x00000001, which was the version we published on Oct 10, 2017.

• IR_VERSION_2017_10_30 = 2

  IR_VERSION 2 published on Oct 30, 2017 - Added type discriminator to AttributeProto to support proto3 users

• IR_VERSION_2017_11_3 = 3

  IR VERSION 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_2019_1_22 = 4

  IR VERSION 4 published on Jan 22, 2019 - Relax constraint that initializers should be a subset of graph inputs - Add type BFLOAT16

• IR_VERSION_2019_3_18 = 5

  IR VERSION 5 published on March 18, 2019 - Add message TensorAnnotation. - Add quantization annotation in GraphProto to map tensor with its scale and zero point quantization parameters.

• IR_VERSION_2019_9_19 = 6

  IR VERSION 6 published on Sep 19, 2019 - Add support for sparse tensor constants stored in model. - Add message SparseTensorProto - Add sparse initializers

• IR_VERSION_2020_5_8 = 7

  IR VERSION 7 published on May 8, 2020 - Add support to allow function body graph to rely on multiple external opreator sets. - Add a list to promote inference graph's initializers to global and mutable variables. Global variables are visible in all graphs of the stored models. - Add message TrainingInfoProto to store initialization method and training algorithm. The execution of TrainingInfoProto can modify the values of mutable variables. - Implicitly add inference graph into each TrainingInfoProto's algorithm.

• IR_VERSION = 8

  IR VERSION 8 published on <TBD> Introduce TypeProto.SparseTensor Introduce TypeProto.Optional Added a list of FunctionProtos local to the model Deprecated since_version and operator status from FunctionProto