daceml.onnx¶

has_onnx_node(name)[source]¶

Check if an ONNX operator is supported.

Parameters: name (str) – the operator name.
Return type: bool

get_onnx_node(name)[source]¶

Get the ONNX Operator node for an operator by name.

Parameters: name (str) – the operator name
Return type: ONNXOp

class ONNXModel(name, model, infer_shapes=True, cuda=False, apply_strict=False, auto_optimize=True, fold_constants=True, parent_pytorch_module=None, storage=None, save_transients=None, auto_merge=False)[source]¶

Loads an ONNX model into an SDFG.

Example

First download an ONNX model, such as efficientnet.

import onnx
import os
import numpy as np
from daceml.onnx import ONNXModel

model_path = os.path.join("..", "tests", "onnx_files", "efficientnet.onnx")
model = onnx.load(model_path)
dace_model = ONNXModel("efficientnet", model)

test_input = np.random.rand(1, 3, 224, 224).astype(np.float32)
dace_model(test_input)

Parameters

name (str) – the name for the SDFG.
model (ModelProto) – the model to import.
infer_shapes (bool) – whether to infer shapes for the model. If this is False, the model must have value infos (with shapes) for all arrays, including intermediate values.
cuda (bool) – if True, the model will be executed on the GPU.
apply_strict (bool) – if True, apply strict transformations after all nodes have been expanded calling (warning: this can be very slow!)
auto_optimize (bool) – if True, apply automatic optimizations before calling.
parent_pytorch_module (Optional[Module]) – when not None, the weight tensors are loaded from the parameters of this model rather than the ONNX graph.
storage (Optional[StorageType]) – the storage type of the parameters, inputs and outputs. If None, will be set according to cuda.
save_transients (Optional[Dict[str, Tensor]]) – if not None, save transients to this dict (for debugging).
auto_merge (bool) – whether to automatically merge symbolic shapes in symbolic shape inference.
fold_constants (bool) –

__call__(*args, **kwargs)[source]¶

Execute the model.

Parameters

args – positional arguments to the model. The i-th argument will be passed as the i-th input of the model.
kwargs – named arguments to the model. The passed names should match the names in the ONNX model.

Return type

Union[Tensor, ndarray, Tuple[Union[Tensor, ndarray]]]

Returns

the output of the model (or a tuple of outputs if there are multiple).

compile_and_init()[source]¶

Compile the SDFG and load parameters into GPU memory.

Return type: CompiledSDFG

inputs: List[str]¶: the inputs to the model

outputs: List[str]¶: the outputs of the model

sdfg: SDFG¶: the generated SDFG.

state: SDFGState¶: the state containing the model computation.

weights: Dict[str, torch.Tensor]¶: mapping from weight name to array

class ONNXOp(*args, **kwargs)[source]¶

Bases: dace.sdfg.nodes.LibraryNode

Abstract superclass for all ONNX ops. Do not use this class, use the concrete subclasses (e.g. ONNXConv) instead.

backward_implementation¶: Which implementation this library node will expand into in the backward pass.

iter_edges(state)[source]¶

Returns an iterator over tuples of an edge and a boolean that indicates whether that edge is an input, ordered by the order required by the schema. This method assumes that this node has been validated.

Parameters: state (SDFGState) – the state containing this node.
Return type: Iterator[Tuple[MultiConnectorEdge, bool]]

iter_inputs_in_onnx_order(state)[source]¶

Iterate through the output edges in the same order as they would appear in an ONNX node proto. This assumes that the node has been validated!

Parameters: state (SDFGState) – the state containing this node.
Return type: List[MultiConnectorEdge]
Returns: the in edges in the order as they would appear in the node proto.

iter_outputs_in_onnx_order(state)[source]¶

Iterate through the input edges in the same order as they would appear in an ONNX node proto. This assumes that the node has been validated!

Parameters: state (SDFGState) – the state containing this node.
Return type: List[MultiConnectorEdge]
Returns: the out edges in the order as they would appear in the node proto.

schema¶: The operator’s ONNX OpSchema

validate(sdfg, state)[source]¶

Validate this node.

Parameters

sdfg (SDFG) – the parent sdfg.
state (SDFGState) – the parent state.

Schema Representation¶

onnx_representation(represents, **mapping)[source]¶

Decorator for python representations of ONNX protobufs.

The decorator will monkey patch in the following methods:

__init__ (a constructor based on the class properties)
construct_from_onnx_proto
construct_from_json

Parameters

represents – The onnx protobuf type that the decorated class represents
mapping –
a mapping from class property names to either:

a string s
In this case, convert_onnx_attribute will be applied on the protobuf attribute with the name s to get the property value.

a function f
In this case, f will be called with the protobuf, and the property value will be set to the return value of that call.

If a property name is not present in mapping, the property name itself will be used to access the protobuf attribute.

class ONNXParameterType(value=<object object>, names=None, module=None, type=None, start=1, boundary=None)[source]¶

An enumeration.

Optional = 2¶: optional parameters

Single = 1¶: single/required parameters

Variadic = 3¶: variadic parameters

class ONNXAttributeType(value=<object object>, names=None, module=None, type=None, start=1, boundary=None)[source]¶

An enumeration.

Float = 2¶: Float (python representation is float)

Floats = 5¶: Floats (python representation is List [float])

Int = 1¶: Integer (python representation is int)

Ints = 4¶: Ints (python representation is List [int])

String = 3¶: String (python representation is str)

Strings = 6¶: Strings (python representation is List [str])

Tensor = 7¶: Tensor (python representation is numpy.ndarray)

Unsupported = 8¶: Any unsupported attribute type

class ONNXAttribute(name='', description='', required=False, attribute_type=<ONNXAttributeType.Int: 1>, default_value=None)[source]¶

Python representation of an ONNX attribute.

Parameters

name (str, default '') – The attribute name
description (str, default '') – A description this attribute
required (bool, default False) – Whether this attribute is required
attribute_type (ONNXAttributeType, default <ONNXAttributeType.Int: 1>) – The type of this attribute
default_value (NoneType, default None) – The default value of this attribute

attribute_type¶: The type of this attribute

default_value¶: The default value of this attribute

description¶: A description this attribute

classmethod from_json(json, context=None)[source]¶: Construct an object json

classmethod from_onnx_proto(onnx_proto)[source]¶: Construct an object from an ONNX proto of type <class 'onnx.onnx_cpp2py_export.defs.OpSchema.Attribute'>.

name¶: The attribute name

required¶: Whether this attribute is required

to_json()[source]¶: Serialize to json

class ONNXTypeConstraint(type_str='', types=[])[source]¶

Python representation of an ONNX type constraint.

Parameters

type_str (str, default '') – The type parameter string
types (list, default []) – The possible types. Note that only tensor types are currently supported.

classmethod from_json(json, context=None)[source]¶: Construct an object json

classmethod from_onnx_proto(onnx_proto)[source]¶: Construct an object from an ONNX proto of type <class 'onnx.onnx_cpp2py_export.defs.OpSchema.TypeConstraintParam'>.

to_json()[source]¶: Serialize to json

type_str¶: The type parameter string

types¶: The possible types. Note that only tensor types are currently supported.

class ONNXParameter(name='', description='', type_str='', param_type=<ONNXParameterType.Single: 1>, homogeneous=False)[source]¶

Python representation of an ONNX parameter.

Parameters

name (str, default '') – The parameter name
description (str, default '') – A description of the parameter
type_str (str, default '') – The type string of this parameter
param_type (ONNXParameterType, default <ONNXParameterType.Single: 1>) – The type of the this parameter
homogeneous (bool, default False) – Whether this parameter is homogeneous

description¶: A description of the parameter

classmethod from_json(json, context=None)[source]¶: Construct an object json

classmethod from_onnx_proto(onnx_proto)[source]¶: Construct an object from an ONNX proto of type <class 'onnx.onnx_cpp2py_export.defs.OpSchema.FormalParameter'>.

homogeneous¶: Whether this parameter is homogeneous

name¶: The parameter name

param_type¶: The type of the this parameter

to_json()[source]¶: Serialize to json

type_str¶: The type string of this parameter

class ONNXSchema(name='', domain='', doc='', since_version=0, attributes={}, type_constraints={}, inputs=[], outputs=[])[source]¶

Python representation of an ONNX schema

Parameters

name (str, default '') – The operator name
domain (str, default '') – The operator domain
doc (str, default '') – The operator’s docstring
since_version (int, default 0) – The version of the operator
attributes (dict, default {}) – The operator attributes. Keys should contain the name of the attribute, and values should have type ONNXAttribute.
type_constraints (dict, default {}) – The type constraints for inputs and outputs. Keys should contain the type string of the constraint, values should have type ONNXTypeConstraint.
inputs (list, default []) – The operator input parameter descriptors. Entries should have type ONNXParameter.
outputs (list, default []) – The operator output parameter descriptors. Entries should have type ONNXParameter.

attributes¶: The operator attributes. Keys should contain the name of the attribute, and values should have type ONNXAttribute.

doc¶: The operator’s docstring

domain¶: The operator domain

classmethod from_json(json, context=None)[source]¶: Construct an object json

classmethod from_onnx_proto(onnx_proto)[source]¶: Construct an object from an ONNX proto of type <class 'onnx.onnx_cpp2py_export.defs.OpSchema'>.

inputs¶: The operator input parameter descriptors. Entries should have type ONNXParameter.

name¶: The operator name

outputs¶: The operator output parameter descriptors. Entries should have type ONNXParameter.

since_version¶: The version of the operator

to_json()[source]¶: Serialize to json

type_constraints¶: The type constraints for inputs and outputs. Keys should contain the type string of the constraint, values should have type ONNXTypeConstraint.

Op Implementation Registration¶

class ONNXForward[source]¶

ABC for ONNX op forward implementations.

The register function expects an argument op containing the ONNX op name (string).

abstract static forward(node, state, sdfg)[source]¶

Expand node and return its expansion.

Parameters

node (ONNXOp) – the candidate node.
state (SDFGState) – the candidate state.
sdfg (SDFG) – the candidate sdfg.

Return type

Union[Node, SDFG]

Returns

the expanded node.

static forward_can_be_applied(node, state, sdfg)[source]¶

Return whether this expansion can be applied.

Parameters

node (ONNXOp) – the candidate node.
state (SDFGState) – the candidate state.
sdfg (SDFG) – the candidate sdfg.

Return type

bool

Returns

whether this expansion can be applied.

op_implementation(op, name)[source]¶

A decorator that registers an op implementation. It should be used on classes that extend ONNXForward.

Parameters

op – the ONNX name of the op to register for.
name – the name of the implementation.

SDFG-based ONNX Implementations¶

class PureCast[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Implementation name: "pure"

class PureEinsum[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Implementation name: "pure"

class PureErf[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Implementation name: "pure"

class PureFlatten[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Flatten Expansion, reuses Reshape implementation

Implementation name

"pure"

class PureGemm[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Implementation name: "pure"

class PureLog[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Implementation name: "pure"

class PureLogSoftmax[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Implementation name: "pure"

class PureMatMul[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Implementation name: "pure"

class PurePow[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Implementation name: "pure"

class PureReciprocal[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Implementation name: "pure"

class PureReduceMax[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Implementation name: "pure"

class PureReduceMean[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Implementation name: "pure"

class PureReduceMin[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Implementation name: "pure"

class PureReduceSum[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Implementation name: "pure"

class PureRelu[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Implementation name: "pure"

class PureReshape[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Reshape expansion: this relies on views

Implementation name

"pure"

class PureSigmoid[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Implementation name: "pure"

class PureSoftmax[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Implementation name: "pure"

class PureSqrt[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Implementation name: "pure"

class PureSum[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Implementation name: "pure"

class PureTranspose[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Implementation name: "pure"

Image-related Implementations¶

class PureBatchNormalization[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Implementation name: "pure"

class PureConv2D[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

The “trivial” convolution implementation, i.e. two nested maps.

Implementation name: "pure"

class PureMaxPool2D[source]¶

Bases: daceml.onnx.forward_implementation_abc.ONNXForward

Implementation name: "pure"

Dace CMake Environments¶

class ONNXRuntime(*args, **kwargs)[source]¶: Environment used to run ONNX operator nodes using ONNX Runtime. See Installing ONNXRuntime for installation instructions.

class ONNXRuntimeCUDA(*args, **kwargs)[source]¶: Environment used to run ONNX operator nodes using ONNX Runtime, with the CUDA execution provider. See Installing ONNXRuntime for installation instructions.

Supported ONNX Operators¶

The following documentation is mostly automatically generated from the ONNX documentation, except for the removal of unsupported attributes and nodes.

class ONNXAbs(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Absolute takes one input data (Tensor<T>) and produces one output data (Tensor<T>) where the absolute is, y = abs(x), is applied to the tensor elementwise.

Node Inputs

X (T, single) – Input tensor

Node Outputs

Y (T, single) – Output tensor

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64

Parameters

name – the name of the node.

class ONNXAcos(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Calculates the arccosine (inverse of cosine) of the given input tensor, element-wise.

Node Inputs

input (T, single) – Input tensor

Node Outputs

output (T, single) – The arccosine of the input tensor computed element-wise

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXAcosh(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Calculates the hyperbolic arccosine of the given input tensor element-wise.

Node Inputs

input (T, single) – Input tensor

Node Outputs

output (T, single) – The hyperbolic arccosine values of the input tensor computed element-wise

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXAdagrad(name, *, decay_factor=0.0, epsilon=9.999999974752427e-07, norm_coefficient=0.0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Compute one iteration of ADAGRAD, a stochastic gradient based optimization algorithm. This operator can conduct the optimization of multiple tensor variables.

Let’s define the behavior of this operator. As you can imagine, ADAGRAD requires some parameters:

The initial learning-rate “R”.

The update count “T”. That is, the number of training iterations conducted.

A L2-norm regularization coefficient “norm_coefficient”.

A learning-rate decay factor “decay_factor”.

A small constant “epsilon” to avoid dividing-by-zero.

At each ADAGRAD iteration, the optimized tensors are moved along a direction computed based on their estimated gradient and accumulated squared gradient. Assume that only a single tensor “X” is updated by this operator. We need the value of “X”, its gradient “G”, and its accumulated squared gradient “H”. Therefore, variables in this operator’s input list are sequentially “R”, “T”, “X”, “G”, and “H”. Other parameters are given as attributes because they are usually constants. Also, the corresponding output tensors are the new value of “X” (called “X_new”), and then the new accumulated squared gradient (called “H_new”). Those outputs are computed from the given inputs following the pseudo code below.

Let “+”, “-“, “*”, and “/” are all element-wise arithmetic operations with numpy-style broadcasting support. The pseudo code to compute those outputs is:

// Compute a scalar learning-rate factor. At the first update of X, T is generally // 0 (0-based update index) or 1 (1-based update index). r = R / (1 + T * decay_factor);

// Add gradient of 0.5 * norm_coefficient * ||X||_2^2, where ||X||_2 is the 2-norm. G_regularized = norm_coefficient * X + G;

// Compute new accumulated squared gradient. H_new = H + G_regularized * G_regularized;

// Compute the adaptive part of per-coordinate learning rate. Note that Sqrt(…) // computes element-wise square-root. H_adaptive = Sqrt(H_new) + epsilon

// Compute the new value of “X”. X_new = X - r * G_regularized / H_adaptive;

If one assign this operators to optimize multiple inputs, for example, “X_1” and “X_2”, the same pseudo code may be extended to handle all tensors jointly. More specifically, we can view “X” as a concatenation of “X_1” and “X_2” (of course, their gradient and accumulate gradient should be concatenated too) and then just reuse the entire pseudo code.

Note that ADAGRAD was first proposed in http://jmlr.org/papers/volume12/duchi11a/duchi11a.pdf. In that reference paper, this operator is a special case of the Figure 1’s composite mirror descent update.

Node Inputs

R (T1, single) – The initial learning rate.
T (T2, single) – The update count of “X”. It should be a scalar.
inputs (T3, variadic) – The current values of optimized tensors, followed by their respective gradients, followed by their respective accumulated squared gradients.For example, if two tensor “X_1” and “X_2” are optimized, The input list would be [“X_1”, “X_2”, gradient of “X_1”, gradient of “X_2”, accumulated squared gradient of “X_1”, accumulated squared gradient of “X_2”].

Node Outputs

outputs (T3, variadic) – Updated values of optimized tensors, followed by their updated values of accumulated squared gradients. For example, if two tensor “X_1” and “X_2” are optimized, the output list would be [new value of “X_1,” new value of “X_2” new accumulated squared gradient of “X_1”, new accumulated squared gradient of “X_2”].

Type Constraints

T1 – float32, float64
T2 – int64
T3 – float32, float64

Parameters

name – the name of the node.
decay_factor (Optional [float], default=0.0) – The decay factor of learning rate after one update.The effective learning rate is computed by r = R / (1 + T * decay_factor). Default to 0 so that increasing update counts doesn’t reduce the learning rate.
epsilon (Optional [float], default=9.999999974752427e-07) – Small scalar to avoid dividing by zero.
norm_coefficient (Optional [float], default=0.0) – Regularization coefficient in 0.5 * norm_coefficient * ||X||_2^2. Default to 0, which means no regularization.

decay_factor¶: The decay factor of learning rate after one update.The effective learning rate is computed by r = R / (1 + T * decay_factor). Default to 0 so that increasing update counts doesn’t reduce the learning rate.

epsilon¶: Small scalar to avoid dividing by zero.

norm_coefficient¶: Regularization coefficient in 0.5 * norm_coefficient * ||X||_2^2. Default to 0, which means no regularization.

class ONNXAdam(name, *, alpha=0.8999999761581421, beta=0.9990000128746033, epsilon=9.999999974752427e-07, norm_coefficient=0.0, norm_coefficient_post=0.0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Compute one iteration of Adam, a stochastic gradient based optimization algorithm. This operator can conduct the optimization of multiple tensor variables.

Let’s define the behavior of this operator. First of all, Adam requires some parameters:

The learning-rate “R”.

The update count “T”. That is, the number of training iterations conducted.

A L2-norm regularization coefficient “norm_coefficient”.

A small constant “epsilon” to avoid dividing-by-zero.

Two coefficients, “alpha” and “beta”.

At each Adam iteration, the optimized tensors are moved along a direction computed based on their exponentially-averaged historical gradient and exponentially-averaged historical squared gradient. Assume that only a tensor “X” is being optimized. The rest of required information is

the value of “X”,

“X“‘s gradient (denoted by “G”),

“X“‘s exponentially-averaged historical gradient (denoted by “V”), and

“X“‘s exponentially-averaged historical squared gradient (denoted by “H”).

Some of those parameters are passed into this operator as input tensors and others are stored as this operator’s attributes. Specifically, this operator’s input tensor list is [“R”, “T”, “X”, “G”, “V”, “H”]. That is, “R” is the first input, “T” is the second input, and so on. Other parameters are given as attributes because they are constants. Moreover, the corresponding output tensors are

the new value of “X” (called “X_new”),

the new exponentially-averaged historical gradient (denoted by “V_new”), and

the new exponentially-averaged historical squared gradient (denoted by “H_new”).

Those outputs are computed following the pseudo code below.

Let “+”, “-“, “*”, and “/” are all element-wise arithmetic operations with numpy-style broadcasting support. The pseudo code to compute those outputs is:

// Add gradient of 0.5 * norm_coefficient * ||X||_2^2, where ||X||_2 is the 2-norm. G_regularized = norm_coefficient * X + G

// Update exponentially-averaged historical gradient. V_new = alpha * V + (1 - alpha) * G_regularized

// Update exponentially-averaged historical squared gradient. H_new = beta * H + (1 - beta) * G_regularized * G_regularized

// Compute the element-wise square-root of H_new. V_new will be element-wisely // divided by H_sqrt for a better update direction. H_sqrt = Sqrt(H_new) + epsilon

// Compute learning-rate. Note that “alpha**T”/”beta**T” is alpha’s/beta’s T-th power. R_adjusted = T > 0 ? R * Sqrt(1 - beta**T) / (1 - alpha**T) : R

// Compute new value of “X”. X_new = X - R_adjusted * V_new / H_sqrt

// Post-update regularization. X_final = (1 - norm_coefficient_post) * X_new

If there are multiple inputs to be optimized, the pseudo code will be applied independently to each of them.

Node Inputs

R (T1, single) – The initial learning rate.
T (T2, single) – The update count of “X”. It should be a scalar.
inputs (T3, variadic) – The tensors to be optimized, followed by their respective gradients, followed by their respective accumulated gradients (aka momentum), followed by their respective accumulated squared gradients. For example, to optimize tensors “X_1” and “X_2,”, the input list would be [“X_1”, “X_2”, gradient of “X_1”, gradient of “X_2”, accumulated gradient of “X_1”, accumulated gradient of “X_2”, accumulated squared gradient of “X_1”, accumulated squared gradient of “X_2”].

Node Outputs

outputs (T3, variadic) – New values of optimized tensors, followed by their respective new accumulated gradients, followed by their respective new accumulated squared gradients. For example, if two tensors “X_1” and “X_2” are optimized, the outputs list would be [new value of “X_1”, new value of “X_2”, new accumulated gradient of “X_1”, new accumulated gradient of “X_2”, new accumulated squared gradient of “X_1”, new accumulated squared gradient of “X_2”].

Type Constraints

T1 – float32, float64
T2 – int64
T3 – float32, float64

Parameters

name – the name of the node.
alpha (Optional [float], default=0.8999999761581421) – Coefficient of previously accumulated gradient in running average. Default to 0.9.
beta (Optional [float], default=0.9990000128746033) – Coefficient of previously accumulated squared-gradient in running average. Default to 0.999.
epsilon (Optional [float], default=9.999999974752427e-07) – Small scalar to avoid dividing by zero.
norm_coefficient (Optional [float], default=0.0) – Regularization coefficient of 0.5 * norm_coefficient * ||X||_2^2. Default to 0, which means no regularization.
norm_coefficient_post (Optional [float], default=0.0) – Regularization coefficient of 0.5 * norm_coefficient * ||X||_2^2. Default to 0, which means no regularization.

alpha¶: Coefficient of previously accumulated gradient in running average. Default to 0.9.

beta¶: Coefficient of previously accumulated squared-gradient in running average. Default to 0.999.

epsilon¶: Small scalar to avoid dividing by zero.

norm_coefficient¶: Regularization coefficient of 0.5 * norm_coefficient * ||X||_2^2. Default to 0, which means no regularization.

norm_coefficient_post¶: Regularization coefficient of 0.5 * norm_coefficient * ||X||_2^2. Default to 0, which means no regularization.

class ONNXAdd(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Performs element-wise binary addition (with Numpy-style broadcasting support).

This operator supports multidirectional (i.e., Numpy-style) broadcasting; for more details please check [the doc](Broadcasting.md).

Node Inputs

A (T, single) – First operand.
B (T, single) – Second operand.

Node Outputs

C (T, single) – Result, has same element type as two inputs

Type Constraints

T – uint32, uint64, int32, int64, float16, float32, float64

Parameters

name – the name of the node.

class ONNXAnd(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Returns the tensor resulted from performing the and logical operation elementwise on the input tensors A and B (with Numpy-style broadcasting support).

This operator supports multidirectional (i.e., Numpy-style) broadcasting; for more details please check [the doc](Broadcasting.md).

Node Inputs

A (T, single) – First input operand for the logical operator.
B (T, single) – Second input operand for the logical operator.

Node Outputs

C (T1, single) – Result tensor.

Type Constraints

T – bool_
T1 – bool_

Parameters

name – the name of the node.

class ONNXArgMax(name, *, axis=0, keepdims=1, select_last_index=0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Computes the indices of the max elements of the input tensor’s element along the provided axis. The resulting tensor has the same rank as the input if keepdims equal 1. If keepdims equal 0, then the resulting tensor have the reduced dimension pruned. If select_last_index is True (default False), the index of the last occurrence of the max is selected if the max appears more than once in the input. Otherwise the index of the first occurrence is selected. The type of the output tensor is integer.

Node Inputs

data (T, single) – An input tensor.

Node Outputs

reduced (reduced_constraint, single) – Reduced output tensor with integer data type.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64
reduced_constraint – int64

Parameters

name – the name of the node.
axis (Optional [int], default=0) – The axis in which to compute the arg indices. Accepted range is [-r, r-1] where r = rank(data).
keepdims (Optional [int], default=1) – Keep the reduced dimension or not, default 1 mean keep reduced dimension.
select_last_index (Optional [int], default=0) – Whether to select the last index or the first index if the {name} appears in multiple indices, default is False (first index).

axis¶: The axis in which to compute the arg indices. Accepted range is [-r, r-1] where r = rank(data).

keepdims¶: Keep the reduced dimension or not, default 1 mean keep reduced dimension.

select_last_index¶: Whether to select the last index or the first index if the {name} appears in multiple indices, default is False (first index).

class ONNXArgMin(name, *, axis=0, keepdims=1, select_last_index=0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Computes the indices of the min elements of the input tensor’s element along the provided axis. The resulting tensor has the same rank as the input if keepdims equal 1. If keepdims equal 0, then the resulting tensor have the reduced dimension pruned. If select_last_index is True (default False), the index of the last occurrence of the min is selected if the min appears more than once in the input. Otherwise the index of the first occurrence is selected. The type of the output tensor is integer.

Node Inputs

data (T, single) – An input tensor.

Node Outputs

reduced (reduced_constraint, single) – Reduced output tensor with integer data type.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64
reduced_constraint – int64

Parameters

name – the name of the node.
axis (Optional [int], default=0) – The axis in which to compute the arg indices. Accepted range is [-r, r-1] where r = rank(data).
keepdims (Optional [int], default=1) – Keep the reduced dimension or not, default 1 mean keep reduced dimension.
select_last_index (Optional [int], default=0) – Whether to select the last index or the first index if the {name} appears in multiple indices, default is False (first index).

axis¶: The axis in which to compute the arg indices. Accepted range is [-r, r-1] where r = rank(data).

keepdims¶: Keep the reduced dimension or not, default 1 mean keep reduced dimension.

select_last_index¶: Whether to select the last index or the first index if the {name} appears in multiple indices, default is False (first index).

class ONNXArrayFeatureExtractor(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Select elements of the input tensor based on the indices passed. The indices are applied to the last axes of the tensor.

Node Inputs

X (T, single) – Data to be selected
Y (Y_constraint, single) – The indices, based on 0 as the first index of any dimension.

Node Outputs

Z (T, single) – Selected output data as an array

Type Constraints

T – float32, float64, int64, int32
Y_constraint – int64

Parameters

name – the name of the node.

class ONNXAsin(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Calculates the arcsine (inverse of sine) of the given input tensor, element-wise.

Node Inputs

input (T, single) – Input tensor

Node Outputs

output (T, single) – The arcsine of the input tensor computed element-wise

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXAsinh(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Calculates the hyperbolic arcsine of the given input tensor element-wise.

Node Inputs

input (T, single) – Input tensor

Node Outputs

output (T, single) – The hyperbolic arcsine values of the input tensor computed element-wise

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXAtan(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Calculates the arctangent (inverse of tangent) of the given input tensor, element-wise.

Node Inputs

input (T, single) – Input tensor

Node Outputs

output (T, single) – The arctangent of the input tensor computed element-wise

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXAtanh(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Calculates the hyperbolic arctangent of the given input tensor element-wise.

Node Inputs

input (T, single) – Input tensor

Node Outputs

output (T, single) – The hyperbolic arctangent values of the input tensor computed element-wise

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXAveragePool(name, *, auto_pad='NOTSET', ceil_mode=0, count_include_pad=0, kernel_shape, pads=None, strides=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

AveragePool consumes an input tensor X and applies average pooling across the tensor according to kernel sizes, stride sizes, and pad lengths. average pooling consisting of computing the average on all values of a subset of the input tensor according to the kernel size and downsampling the data into the output tensor Y for further processing. The output spatial shape will be following: ` output_spatial_shape[i] = floor((input_spatial_shape[i] + pad_shape[i] - kernel_spatial_shape[i]) / strides_spatial_shape[i] + 1) ` or ` output_spatial_shape[i] = ceil((input_spatial_shape[i] + pad_shape[i] - kernel_spatial_shape[i]) / strides_spatial_shape[i] + 1) ` if ceil_mode is enabled

` * pad_shape[i] is sum of pads along axis i `

auto_pad is a DEPRECATED attribute. If you are using them currently, the output spatial shape will be following: ` VALID: output_spatial_shape[i] = ceil((input_spatial_shape[i] - kernel_spatial_shape[i] + 1) / strides_spatial_shape[i]) SAME_UPPER or SAME_LOWER: output_spatial_shape[i] = ceil(input_spatial_shape[i] / strides_spatial_shape[i]) ` And pad shape will be following if SAME_UPPER or SAME_LOWER: ` pad_shape[i] = (output_spatial_shape[i] - 1) * strides_spatial_shape[i] + kernel_spatial_shape[i] - input_spatial_shape[i] ` The output of each pooling window is divided by the number of elements (exclude pad when attribute count_include_pad is zero).

Node Inputs

X (T, single) – Input data tensor from the previous operator; dimensions for image case are (N x C x H x W), where N is the batch size, C is the number of channels, and H and W are the height and the width of the data. For non image case, the dimensions are in the form of (N x C x D1 x D2 … Dn), where N is the batch size. Optionally, if dimension denotation is in effect, the operation expects the input data tensor to arrive with the dimension denotation of [DATA_BATCH, DATA_CHANNEL, DATA_FEATURE, DATA_FEATURE …].

Node Outputs

Y (T, single) – Output data tensor from average or max pooling across the input tensor. Dimensions will vary based on various kernel, stride, and pad sizes. Floor value of the dimension is used

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.
auto_pad (Optional [str], default=’NOTSET’) – auto_pad must be either NOTSET, SAME_UPPER, SAME_LOWER or VALID. Where default value is NOTSET, which means explicit padding is used. SAME_UPPER or SAME_LOWER mean pad the input so that the output spatial size match the input.In case of odd number add the extra padding at the end for SAME_UPPER and at the beginning for SAME_LOWER. VALID mean no padding.
ceil_mode (Optional [int], default=0) – Whether to use ceil or floor (default) to compute the output shape.
count_include_pad (Optional [int], default=0) – Whether include pad pixels when calculating values for the edges. Default is 0, doesn’t count include pad.
kernel_shape (List [int]) – The size of the kernel along each axis.
pads (Optional [List [int]], default=None) – Padding for the beginning and ending along each spatial axis, it can take any value greater than or equal to 0. The value represent the number of pixels added to the beginning and end part of the corresponding axis. pads format should be as follow [x1_begin, x2_begin…x1_end, x2_end,…], where xi_begin the number of pixels added at the beginning of axis i and xi_end, the number of pixels added at the end of axis i. This attribute cannot be used simultaneously with auto_pad attribute. If not present, the padding defaults to 0 along start and end of each spatial axis.
strides (Optional [List [int]], default=None) – Stride along each spatial axis. If not present, the stride defaults to 1 along each spatial axis.

auto_pad¶: auto_pad must be either NOTSET, SAME_UPPER, SAME_LOWER or VALID. Where default value is NOTSET, which means explicit padding is used. SAME_UPPER or SAME_LOWER mean pad the input so that the output spatial size match the input.In case of odd number add the extra padding at the end for SAME_UPPER and at the beginning for SAME_LOWER. VALID mean no padding.

ceil_mode¶: Whether to use ceil or floor (default) to compute the output shape.

count_include_pad¶: Whether include pad pixels when calculating values for the edges. Default is 0, doesn’t count include pad.

kernel_shape¶: The size of the kernel along each axis.

pads¶: Padding for the beginning and ending along each spatial axis, it can take any value greater than or equal to 0. The value represent the number of pixels added to the beginning and end part of the corresponding axis. pads format should be as follow [x1_begin, x2_begin…x1_end, x2_end,…], where xi_begin the number of pixels added at the beginning of axis i and xi_end, the number of pixels added at the end of axis i. This attribute cannot be used simultaneously with auto_pad attribute. If not present, the padding defaults to 0 along start and end of each spatial axis.

strides¶: Stride along each spatial axis. If not present, the stride defaults to 1 along each spatial axis.

class ONNXBatchNormalization(name, *, epsilon=9.999999747378752e-06, momentum=0.8999999761581421)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Carries out batch normalization as described in the paper https://arxiv.org/abs/1502.03167. Depending on the mode it is being run, there are multiple cases for the number of outputs, which we list below:

Output case #1: Y, mean, var, saved_mean, saved_var (training mode) Output case #2: Y (test mode)

For previous (depreciated) non-spatial cases, implementors are suggested to flatten the input shape to (N x C*D1*D2 ..*Dn) before a BatchNormalization Op. This operator has optional inputs/outputs. See [the doc](IR.md) for more details about the representation of optional arguments. An empty string may be used in the place of an actual argument’s name to indicate a missing argument. Trailing optional arguments (those not followed by an argument that is present) may also be simply omitted.

Node Inputs

X (T, single) – Input data tensor from the previous operator; dimensions are in the form of (N x C x D1 x D2 … Dn), where N is the batch size, C is the number of channels. Statistics are computed for every channel of C over N and D1 to Dn dimensions. For image data, input dimensions become (N x C x H x W). The op also accepts single dimension input of size N in which case C is assumed to be 1
scale (T, single) – Scale tensor of shape (C).
B (T, single) – Bias tensor of shape (C).
in_mean (T, single) – running (training) or estimated (testing) mean tensor of shape (C).
in_var (T, single) – running (training) or estimated (testing) variance tensor of shape (C).

Node Outputs

Y (T, single) – The output tensor of the same shape as X
out_mean (T, optional) – The running mean after the BatchNormalization operator.
out_var (T, optional) – The running variance after the BatchNormalization operator.
saved_mean (T, optional) – Saved mean used during training to speed up gradient computation.
saved_var (T, optional) – Saved variance used during training to speed up gradient computation.

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.
epsilon (Optional [float], default=9.999999747378752e-06) – The epsilon value to use to avoid division by zero.
momentum (Optional [float], default=0.8999999761581421) – Factor used in computing the running mean and variance.e.g., running_mean = running_mean * momentum + mean * (1 - momentum).

epsilon¶: The epsilon value to use to avoid division by zero.

momentum¶: Factor used in computing the running mean and variance.e.g., running_mean = running_mean * momentum + mean * (1 - momentum).

class ONNXBinarizer(name, *, threshold=0.0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Maps the values of the input tensor to either 0 or 1, element-wise, based on the outcome of a comparison against a threshold value.

Node Inputs

X (T, single) – Data to be binarized

Node Outputs

Y (T, single) – Binarized output data

Type Constraints

T – float32, float64, int64, int32

Parameters

name – the name of the node.
threshold (Optional [float], default=0.0) – Values greater than this are mapped to 1, others to 0.

threshold¶: Values greater than this are mapped to 1, others to 0.

class ONNXBitShift(name, *, direction)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Bitwise shift operator performs element-wise operation. For each input element, if the

attribute “direction” is “RIGHT”, this operator moves its binary representation toward the right side so that the input value is effectively decreased. If the attribute “direction” is “LEFT”, bits of binary representation moves toward the left side, which results the increase of its actual value. The input X is the tensor to be shifted and another input Y specifies the amounts of shifting. For example, if “direction” is “Right”, X is [1, 4], and S is [1, 1], the corresponding output Z would be [0, 2]. If “direction” is “LEFT” with X=[1, 2] and S=[1, 2], the corresponding output Y would be [2, 8].

Because this operator supports Numpy-style broadcasting, X’s and Y’s shapes are not necessarily identical.

This operator supports multidirectional (i.e., Numpy-style) broadcasting; for more details please check [the doc](Broadcasting.md).

Node Inputs

X (T, single) – First operand, input to be shifted.
Y (T, single) – Second operand, amounts of shift.

Node Outputs

Z (T, single) – Output tensor

Type Constraints

T – uint8, uint16, uint32, uint64

Parameters

name – the name of the node.
direction (str) – Direction of moving bits. It can be either “RIGHT” (for right shift) or “LEFT” (for left shift).

direction¶: Direction of moving bits. It can be either “RIGHT” (for right shift) or “LEFT” (for left shift).

class ONNXCast(name, *, to)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

The operator casts the elements of a given input tensor to a data type specified by the ‘to’ argument and returns an output tensor of the same size in the converted type. The ‘to’ argument must be one of the data types specified in the ‘DataType’ enum field in the TensorProto message.

Casting from string tensor in plain (e.g., “3.14” and “1000”) and scientific numeric representations (e.g., “1e-5” and “1E8”) to float types is supported. For example, converting string “100.5” to an integer may result 100. There are some string literals reserved for special floating-point values; “+INF” (and “INF”), “-INF”, and “NaN” are positive infinity, negative infinity, and not-a-number, respectively. Any string which can exactly match “+INF” in a case-insensitive way would be mapped to positive infinite. Similarly, this case-insensitive rule is applied to “INF” and “NaN”. When casting from numeric tensors to string tensors, plain floating-point representation (such as “314.15926”) would be used. Converting non-numerical-literal string such as “Hello World!” is an undefined behavior. Cases of converting string representing floating-point arithmetic value, such as “2.718”, to INT is an undefined behavior.

Conversion from a numerical type to any numerical type is always allowed. User must be aware of precision loss and value change caused by range difference between two types. For example, a 64-bit float 3.1415926459 may be round to a 32-bit float 3.141592. Similarly, converting an integer 36 to Boolean may produce 1 because we truncate bits which can’t be stored in the targeted type.

Node Inputs

input (T1, single) – Input tensor to be cast.

Node Outputs

output (T2, single) – Output tensor with the same shape as input with type specified by the ‘to’ argument

Type Constraints

T1 – float16, float32, float64, int8, int16, int32, int64, uint8, uint16, uint32, uint64, bool_
T2 – float16, float32, float64, int8, int16, int32, int64, uint8, uint16, uint32, uint64, bool_

Parameters

name – the name of the node.
to (int) – The data type to which the elements of the input tensor are cast. Strictly must be one of the types from DataType enum in TensorProto

to¶: The data type to which the elements of the input tensor are cast. Strictly must be one of the types from DataType enum in TensorProto

class ONNXCategoryMapper(name, *, cats_int64s=None, cats_strings=None, default_int64=- 1, default_string='_Unused')¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Converts strings to integers and vice versa. Two sequences of equal length are used to map between integers and strings, with strings and integers at the same index detailing the mapping. Each operator converts either integers to strings or strings to integers, depending on which default value attribute is provided. Only one default value attribute should be defined. If the string default value is set, it will convert integers to strings. If the int default value is set, it will convert strings to integers.

Node Inputs

X (T1, single) – Input data

Node Outputs

Y (T2, single) – Output data. If strings are input, the output values are integers, and vice versa.

Type Constraints

T1 – int64
T2 – int64

Parameters

name – the name of the node.
cats_int64s (Optional [List [int]], default=None) – The integers of the map. This sequence must be the same length as the ‘cats_strings’ sequence.
cats_strings (Optional [List [str]], default=None) – The strings of the map. This sequence must be the same length as the ‘cats_int64s’ sequence
default_int64 (Optional [int], default=-1) – An integer to use when an input string value is not found in the map. One and only one of the ‘default_*’ attributes must be defined.
default_string (Optional [str], default=’_Unused’) – A string to use when an input integer value is not found in the map. One and only one of the ‘default_*’ attributes must be defined.

cats_int64s¶: The integers of the map. This sequence must be the same length as the ‘cats_strings’ sequence.

cats_strings¶: The strings of the map. This sequence must be the same length as the ‘cats_int64s’ sequence

default_int64¶: An integer to use when an input string value is not found in the map. One and only one of the ‘default_*’ attributes must be defined.

default_string¶: A string to use when an input integer value is not found in the map. One and only one of the ‘default_*’ attributes must be defined.

class ONNXCeil(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Ceil takes one input data (Tensor<T>) and produces one output data (Tensor<T>) where the ceil is, y = ceil(x), is applied to the tensor elementwise.

Node Inputs

X (T, single) – Input tensor

Node Outputs

Y (T, single) – Output tensor

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXCelu(name, *, alpha=1.0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Continuously Differentiable Exponential Linear Units: Perform the linear unit element-wise on the input tensor X using formula:

` max(0,x) + min(0,alpha*(exp(x/alpha)-1)) `

Node Inputs

X (T, single) – Input tensor

Node Outputs

Y (T, single) – Output tensor

Type Constraints

T – float32

Parameters

name – the name of the node.
alpha (Optional [float], default=1.0) – The Alpha value in Celu formula which control the shape of the unit. The default value is 1.0.

alpha¶: The Alpha value in Celu formula which control the shape of the unit. The default value is 1.0.

class ONNXClip(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Clip operator limits the given input within an interval. The interval is specified by the inputs ‘min’ and ‘max’. They default to numeric_limits::lowest() and numeric_limits::max(), respectively.

Node Inputs

input (T, single) – Input tensor whose elements to be clipped
min (T, optional) – Minimum value, under which element is replaced by min. It must be a scalar(tensor of empty shape).
max (T, optional) – Maximum value, above which element is replaced by max. It must be a scalar(tensor of empty shape).

Node Outputs

output (T, single) – Output tensor with clipped input elements

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64

Parameters

name – the name of the node.

class ONNXCompress(name, *, axis=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Selects slices from an input tensor along a given axis where condition evaluates to True for each axis index. In case axis is not provided, input is flattened before elements are selected. Compress behaves like numpy.compress: https://docs.scipy.org/doc/numpy/reference/generated/numpy.compress.html

Node Inputs

input (T, single) – Tensor of rank r >= 1.
condition (T1, single) – Rank 1 tensor of booleans to indicate which slices or data elements to be selected. Its length can be less than the input length along the axis or the flattened input size if axis is not specified. In such cases data slices or elements exceeding the condition length are discarded.

Node Outputs

output (T, single) – Tensor of rank r if axis is specified. Otherwise output is a Tensor of rank 1.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128
T1 – bool_

Parameters

name – the name of the node.
axis (Optional [int], default=None) – (Optional) Axis along which to take slices. If not specified, input is flattened before elements being selected. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(input).

axis¶: (Optional) Axis along which to take slices. If not specified, input is flattened before elements being selected. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(input).

class ONNXConcat(name, *, axis)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Concatenate a list of tensors into a single tensor. All input tensors must have the same shape, except for the dimension size of the axis to concatenate on.

Node Inputs

inputs (T, variadic) – List of tensors for concatenation

Node Outputs

concat_result (T, single) – Concatenated tensor

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128

Parameters

name – the name of the node.
axis (int) – Which axis to concat on. A negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(inputs)..

axis¶: Which axis to concat on. A negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(inputs)..

class ONNXConstant(name, *, sparse_value=None, value=None, value_float=None, value_floats=None, value_int=None, value_ints=None, value_string=None, value_strings=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

This operator produces a constant tensor. Exactly one of the provided attributes, either value, sparse_value, or value_* must be specified.

Node Inputs

Node Outputs

output (T, single) – Output tensor containing the same value of the provided tensor.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128

Parameters

name – the name of the node.
value (Optional [numpy.ndarray], default=None) – The value for the elements of the output tensor.
value_float (Optional [float], default=None) – The value for the sole element for the scalar, float32, output tensor.
value_floats (Optional [List [float]], default=None) – The values for the elements for the 1D, float32, output tensor.
value_int (Optional [int], default=None) – The value for the sole element for the scalar, int64, output tensor.
value_ints (Optional [List [int]], default=None) – The values for the elements for the 1D, int64, output tensor.
value_string (Optional [str], default=None) – The value for the sole element for the scalar, UTF-8 string, output tensor.
value_strings (Optional [List [str]], default=None) – The values for the elements for the 1D, UTF-8 string, output tensor.

value¶: The value for the elements of the output tensor.

value_float¶: The value for the sole element for the scalar, float32, output tensor.

value_floats¶: The values for the elements for the 1D, float32, output tensor.

value_int¶: The value for the sole element for the scalar, int64, output tensor.

value_ints¶: The values for the elements for the 1D, int64, output tensor.

value_string¶: The value for the sole element for the scalar, UTF-8 string, output tensor.

value_strings¶: The values for the elements for the 1D, UTF-8 string, output tensor.

class ONNXConstantOfShape(name, *, value=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Generate a tensor with given value and shape.

Node Inputs

input (T1, single) – 1D tensor. The shape of the expected output tensor. If empty tensor is given, the output would be a scalar. All values must be >= 0.

Node Outputs

output (T2, single) – Output tensor of shape specified by ‘input’.If attribute ‘value’ is specified, the value and datatype of the output tensor is taken from ‘value’.If attribute ‘value’ is not specified, the value in the output defaults to 0, and the datatype defaults to float32.

Type Constraints

T1 – int64
T2 – float16, float32, float64, int8, int16, int32, int64, uint8, uint16, uint32, uint64, bool_

Parameters

name – the name of the node.
value (Optional [numpy.ndarray], default=None) – (Optional) The value of the output elements.Should be a one-element tensor. If not specified, it defaults to a tensor of value 0 and datatype float32

value¶: (Optional) The value of the output elements.Should be a one-element tensor. If not specified, it defaults to a tensor of value 0 and datatype float32

class ONNXConv(name, *, auto_pad='NOTSET', dilations=None, group=1, kernel_shape=None, pads=None, strides=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

The convolution operator consumes an input tensor and a filter, and computes the output.

Node Inputs

X (T, single) – Input data tensor from previous layer; has size (N x C x H x W), where N is the batch size, C is the number of channels, and H and W are the height and width. Note that this is for the 2D image. Otherwise the size is (N x C x D1 x D2 … x Dn). Optionally, if dimension denotation is in effect, the operation expects input data tensor to arrive with the dimension denotation of [DATA_BATCH, DATA_CHANNEL, DATA_FEATURE, DATA_FEATURE …].
W (T, single) – The weight tensor that will be used in the convolutions; has size (M x C/group x kH x kW), where C is the number of channels, and kH and kW are the height and width of the kernel, and M is the number of feature maps. For more than 2 dimensions, the kernel shape will be (M x C/group x k1 x k2 x … x kn), where (k1 x k2 x … kn) is the dimension of the kernel. Optionally, if dimension denotation is in effect, the operation expects the weight tensor to arrive with the dimension denotation of [FILTER_OUT_CHANNEL, FILTER_IN_CHANNEL, FILTER_SPATIAL, FILTER_SPATIAL …]. X.shape[1] == (W.shape[1] * group) == C (assuming zero based indices for the shape array). Or in other words FILTER_IN_CHANNEL should be equal to DATA_CHANNEL.
B (T, optional) – Optional 1D bias to be added to the convolution, has size of M.

Node Outputs

Y (T, single) – Output data tensor that contains the result of the convolution. The output dimensions are functions of the kernel size, stride size, and pad lengths.

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.
auto_pad (Optional [str], default=’NOTSET’) – auto_pad must be either NOTSET, SAME_UPPER, SAME_LOWER or VALID. Where default value is NOTSET, which means explicit padding is used. SAME_UPPER or SAME_LOWER mean pad the input so that the output spatial size match the input.In case of odd number add the extra padding at the end for SAME_UPPER and at the beginning for SAME_LOWER. VALID mean no padding.
dilations (Optional [List [int]], default=None) – dilation value along each spatial axis of the filter. If not present, the dilation defaults is 1 along each spatial axis.
group (Optional [int], default=1) – number of groups input channels and output channels are divided into.
kernel_shape (Optional [List [int]], default=None) – The shape of the convolution kernel. If not present, should be inferred from input W.
pads (Optional [List [int]], default=None) – Padding for the beginning and ending along each spatial axis, it can take any value greater than or equal to 0. The value represent the number of pixels added to the beginning and end part of the corresponding axis. pads format should be as follow [x1_begin, x2_begin…x1_end, x2_end,…], where xi_begin the number of pixels added at the beginning of axis i and xi_end, the number of pixels added at the end of axis i. This attribute cannot be used simultaneously with auto_pad attribute. If not present, the padding defaults to 0 along start and end of each spatial axis.
strides (Optional [List [int]], default=None) – Stride along each spatial axis. If not present, the stride defaults is 1 along each spatial axis.

auto_pad¶: auto_pad must be either NOTSET, SAME_UPPER, SAME_LOWER or VALID. Where default value is NOTSET, which means explicit padding is used. SAME_UPPER or SAME_LOWER mean pad the input so that the output spatial size match the input.In case of odd number add the extra padding at the end for SAME_UPPER and at the beginning for SAME_LOWER. VALID mean no padding.

dilations¶: dilation value along each spatial axis of the filter. If not present, the dilation defaults is 1 along each spatial axis.

group¶: number of groups input channels and output channels are divided into.

kernel_shape¶: The shape of the convolution kernel. If not present, should be inferred from input W.

pads¶: Padding for the beginning and ending along each spatial axis, it can take any value greater than or equal to 0. The value represent the number of pixels added to the beginning and end part of the corresponding axis. pads format should be as follow [x1_begin, x2_begin…x1_end, x2_end,…], where xi_begin the number of pixels added at the beginning of axis i and xi_end, the number of pixels added at the end of axis i. This attribute cannot be used simultaneously with auto_pad attribute. If not present, the padding defaults to 0 along start and end of each spatial axis.

strides¶: Stride along each spatial axis. If not present, the stride defaults is 1 along each spatial axis.

class ONNXConvInteger(name, *, auto_pad='NOTSET', dilations=None, group=1, kernel_shape=None, pads=None, strides=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

The integer convolution operator consumes an input tensor, its zero-point, a filter, and its zero-point, and computes the output. The production MUST never overflow. The accumulation may overflow if and only if in 32 bits.

Node Inputs

x (T1, single) – Input data tensor from previous layer; has size (N x C x H x W), where N is the batch size, C is the number of channels, and H and W are the height and width. Note that this is for the 2D image. Otherwise the size is (N x C x D1 x D2 … x Dn). Optionally, if dimension denotation is in effect, the operation expects input data tensor to arrive with the dimension denotation of [DATA_BATCH, DATA_CHANNEL, DATA_FEATURE, DATA_FEATURE …].
w (T2, single) – The weight tensor that will be used in the convolutions; has size (M x C/group x kH x kW), where C is the number of channels, and kH and kW are the height and width of the kernel, and M is the number of feature maps. For more than 2 dimensions, the kernel shape will be (M x C/group x k1 x k2 x … x kn), where (k1 x k2 x … kn) is the dimension of the kernel. Optionally, if dimension denotation is in effect, the operation expects the weight tensor to arrive with the dimension denotation of [FILTER_OUT_CHANNEL, FILTER_IN_CHANNEL, FILTER_SPATIAL, FILTER_SPATIAL …]. X.shape[1] == (W.shape[1] * group) == C (assuming zero based indices for the shape array). Or in other words FILTER_IN_CHANNEL should be equal to DATA_CHANNEL.
x_zero_point (T1, optional) – Zero point tensor for input ‘x’. It’s optional and default value is 0. It’s a scalar, which means a per-tensor/layer quantization.
w_zero_point (T2, optional) – Zero point tensor for input ‘w’. It’s optional and default value is 0. It could be a scalar or a 1-D tensor, which means a per-tensor/layer or per output channel quantization. If it’s a 1-D tensor, its number of elements should be equal to the number of output channels (M)

Node Outputs

y (T3, single) – Output data tensor that contains the result of the convolution. The output dimensions are functions of the kernel size, stride size, and pad lengths.

Type Constraints

T1 – int8, uint8
T2 – int8, uint8
T3 – int32

Parameters

name – the name of the node.
auto_pad (Optional [str], default=’NOTSET’) – auto_pad must be either NOTSET, SAME_UPPER, SAME_LOWER or VALID. Where default value is NOTSET, which means explicit padding is used. SAME_UPPER or SAME_LOWER mean pad the input so that the output spatial size match the input.In case of odd number add the extra padding at the end for SAME_UPPER and at the beginning for SAME_LOWER. VALID mean no padding.
dilations (Optional [List [int]], default=None) – dilation value along each spatial axis of the filter. If not present, the dilation defaults to 1 along each axis.
group (Optional [int], default=1) – number of groups input channels and output channels are divided into. default is 1.
kernel_shape (Optional [List [int]], default=None) – The shape of the convolution kernel. If not present, should be inferred from input ‘w’.
pads (Optional [List [int]], default=None) – Padding for the beginning and ending along each spatial axis, it can take any value greater than or equal to 0.The value represent the number of pixels added to the beginning and end part of the corresponding axis.`pads` format should be as follow [x1_begin, x2_begin…x1_end, x2_end,…], where xi_begin the number ofpixels added at the beginning of axis i and xi_end, the number of pixels added at the end of axis i.This attribute cannot be used simultaneously with auto_pad attribute. If not present, the padding defaultsto 0 along start and end of each spatial axis.
strides (Optional [List [int]], default=None) – Stride along each spatial axis. If not present, the stride defaults to 1 along each axis.

auto_pad¶: auto_pad must be either NOTSET, SAME_UPPER, SAME_LOWER or VALID. Where default value is NOTSET, which means explicit padding is used. SAME_UPPER or SAME_LOWER mean pad the input so that the output spatial size match the input.In case of odd number add the extra padding at the end for SAME_UPPER and at the beginning for SAME_LOWER. VALID mean no padding.

dilations¶: dilation value along each spatial axis of the filter. If not present, the dilation defaults to 1 along each axis.

group¶: number of groups input channels and output channels are divided into. default is 1.

kernel_shape¶: The shape of the convolution kernel. If not present, should be inferred from input ‘w’.

pads¶: Padding for the beginning and ending along each spatial axis, it can take any value greater than or equal to 0.The value represent the number of pixels added to the beginning and end part of the corresponding axis.`pads` format should be as follow [x1_begin, x2_begin…x1_end, x2_end,…], where xi_begin the number ofpixels added at the beginning of axis i and xi_end, the number of pixels added at the end of axis i.This attribute cannot be used simultaneously with auto_pad attribute. If not present, the padding defaultsto 0 along start and end of each spatial axis.

strides¶: Stride along each spatial axis. If not present, the stride defaults to 1 along each axis.

class ONNXConvTranspose(name, *, auto_pad='NOTSET', dilations=None, group=1, kernel_shape=None, output_padding=None, output_shape=None, pads=None, strides=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

The convolution transpose operator consumes an input tensor and a filter, and computes the output.

If the pads parameter is provided the shape of the output is calculated via the following equation:

output_shape[i] = stride[i] * (input_size[i] - 1) + output_padding[i] + ((kernel_shape[i] - 1) * dilations[i] + 1) - pads[start_i] - pads[end_i]

output_shape can also be explicitly specified in which case pads values are auto generated using these equations:

total_padding[i] = stride[i] * (input_size[i] - 1) + output_padding[i] + ((kernel_shape[i] - 1) * dilations[i] + 1) - output_shape[i] If (auto_pads != SAME_UPPER): pads[start_i] = total_padding[i]/2; pads[end_i] = total_padding[i] - (total_padding[i]/2) Else: pads[start_i] = total_padding[i] - (total_padding[i]/2); pads[end_i] = (total_padding[i]/2).

Node Inputs

X (T, single) – Input data tensor from previous layer; has size (N x C x H x W), where N is the batch size, C is the number of channels, and H and W are the height and width. Note that this is for the 2D image. Otherwise the size is (N x C x D1 x D2 … x Dn)
W (T, single) – The weight tensor that will be used in the convolutions; has size (C x M/group x kH x kW), where C is the number of channels, and kH and kW are the height and width of the kernel, and M is the number of feature maps. For more than 2 dimensions, the weight shape will be (C x M/group x k1 x k2 x … x kn), where (k1 x k2 x … x kn) is the dimension of the kernel. The number of channels in the output should be equal to W.shape[1] * group (assuming zero based indices of the shape array)
B (T, optional) – Optional 1D bias to be added to the convolution, has size of M.

Node Outputs

Y (T, single) – Output data tensor that contains the result of the convolution. The output dimensions are functions of the kernel size, stride size, pad lengths and group count. The number of channels in the output should be equal to W.shape[1] * group (assuming zero based indices of the shape array)

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.
auto_pad (Optional [str], default=’NOTSET’) – auto_pad must be either NOTSET, SAME_UPPER, SAME_LOWER or VALID. Where default value is NOTSET, which means explicit padding is used. SAME_UPPER or SAME_LOWER mean pad the input so that the output spatial size match the input.In case of odd number add the extra padding at the end for SAME_UPPER and at the beginning for SAME_LOWER. VALID mean no padding.
dilations (Optional [List [int]], default=None) – dilation value along each spatial axis of the filter. If not present, the dilation defaults to 1 along each spatial axis.
group (Optional [int], default=1) – number of groups input channels and output channels are divided into.
kernel_shape (Optional [List [int]], default=None) – The shape of the convolution kernel. If not present, should be inferred from input W.
output_padding (Optional [List [int]], default=None) – Additional elements added to the side with higher coordinate indices in the output. Each padding value in “output_padding” must be less than the corresponding stride/dilation dimension. By default, this attribute is a zero vector. Note that this attribute doesn’t directly affect the computed output values. It only controls the selection of the computed values, so changing this attribute only adds or removes output elements. If “output_shape” is explicitly provided, “output_padding” does not contribute additional size to “output_shape” but participates in the computation of the needed padding amount. This is also called adjs or adjustment in some frameworks.
output_shape (Optional [List [int]], default=None) – The shape of the output can be explicitly set which will cause pads values to be auto generated. If output_shape is specified pads values are ignored. See doc for details for equations to generate pads
pads (Optional [List [int]], default=None) – Padding for the beginning and ending along each spatial axis, it can take any value greater than or equal to 0. The value represent the number of pixels added to the beginning and end part of the corresponding axis. pads format should be as follow [x1_begin, x2_begin…x1_end, x2_end,…], where xi_begin the number of pixels added at the beginning of axis i and xi_end, the number of pixels added at the end of axis i. This attribute cannot be used simultaneously with auto_pad attribute. If not present, the padding defaults to 0 along start and end of each spatial axis.
strides (Optional [List [int]], default=None) – Stride along each spatial axis. If not present, the stride defaults to 1 along each spatial axis.

auto_pad¶: auto_pad must be either NOTSET, SAME_UPPER, SAME_LOWER or VALID. Where default value is NOTSET, which means explicit padding is used. SAME_UPPER or SAME_LOWER mean pad the input so that the output spatial size match the input.In case of odd number add the extra padding at the end for SAME_UPPER and at the beginning for SAME_LOWER. VALID mean no padding.

dilations¶: dilation value along each spatial axis of the filter. If not present, the dilation defaults to 1 along each spatial axis.

group¶: number of groups input channels and output channels are divided into.

kernel_shape¶: The shape of the convolution kernel. If not present, should be inferred from input W.

output_padding¶: Additional elements added to the side with higher coordinate indices in the output. Each padding value in “output_padding” must be less than the corresponding stride/dilation dimension. By default, this attribute is a zero vector. Note that this attribute doesn’t directly affect the computed output values. It only controls the selection of the computed values, so changing this attribute only adds or removes output elements. If “output_shape” is explicitly provided, “output_padding” does not contribute additional size to “output_shape” but participates in the computation of the needed padding amount. This is also called adjs or adjustment in some frameworks.

output_shape¶: The shape of the output can be explicitly set which will cause pads values to be auto generated. If output_shape is specified pads values are ignored. See doc for details for equations to generate pads

pads¶: Padding for the beginning and ending along each spatial axis, it can take any value greater than or equal to 0. The value represent the number of pixels added to the beginning and end part of the corresponding axis. pads format should be as follow [x1_begin, x2_begin…x1_end, x2_end,…], where xi_begin the number of pixels added at the beginning of axis i and xi_end, the number of pixels added at the end of axis i. This attribute cannot be used simultaneously with auto_pad attribute. If not present, the padding defaults to 0 along start and end of each spatial axis.

strides¶: Stride along each spatial axis. If not present, the stride defaults to 1 along each spatial axis.

class ONNXCos(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Calculates the cosine of the given input tensor, element-wise.

Node Inputs

input (T, single) – Input tensor

Node Outputs

output (T, single) – The cosine of the input tensor computed element-wise

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXCosh(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Calculates the hyperbolic cosine of the given input tensor element-wise.

Node Inputs

input (T, single) – Input tensor

Node Outputs

output (T, single) – The hyperbolic cosine values of the input tensor computed element-wise

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXCumSum(name, *, exclusive=0, reverse=0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Performs cumulative sum of the input elements along the given axis. By default, it will do the sum inclusively meaning the first element is copied as is. Through an exclusive attribute, this behavior can change to exclude the first element. It can also perform summation in the opposite direction of the axis. For that, set reverse attribute to 1.

Example: ` input_x = [1, 2, 3] axis=0 output = [1, 3, 6] exclusive=1 output = [0, 1, 3] exclusive=0 reverse=1 output = [6, 5, 3] exclusive=1 reverse=1 output = [5, 3, 0] `

Node Inputs

x (T, single) – An input tensor that is to be processed.
axis (T2, single) – (Optional) A 0-D tensor. Must be in the range [-rank(x), rank(x)-1]. Negative value means counting dimensions from the back.

Node Outputs

y (T, single) – Output tensor of the same type as ‘x’ with cumulative sums of the x’s elements

Type Constraints

T – uint32, uint64, int32, int64, float32, float64
T2 – int32, int64

Parameters

name – the name of the node.
exclusive (Optional [int], default=0) – If set to 1 will return exclusive sum in which the top element is not included. In other terms, if set to 1, the j-th output element would be the sum of the first (j-1) elements. Otherwise, it would be the sum of the first j elements.
reverse (Optional [int], default=0) – If set to 1 will perform the sums in reverse direction.

exclusive¶: If set to 1 will return exclusive sum in which the top element is not included. In other terms, if set to 1, the j-th output element would be the sum of the first (j-1) elements. Otherwise, it would be the sum of the first j elements.

reverse¶: If set to 1 will perform the sums in reverse direction.

class ONNXDepthToSpace(name, *, blocksize, mode='DCR')¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

DepthToSpace rearranges (permutes) data from depth into blocks of spatial data. This is the reverse transformation of SpaceToDepth. More specifically, this op outputs a copy of the input tensor where values from the depth dimension are moved in spatial blocks to the height and width dimensions. By default, mode = DCR. In the DCR mode, elements along the depth dimension from the input tensor are rearranged in the following order: depth, column, and then row. The output y is computed from the input x as below:

b, c, h, w = x.shape

tmp = np.reshape(x, [b, blocksize, blocksize, c // (blocksize**2), h, w])

tmp = np.transpose(tmp, [0, 3, 4, 1, 5, 2])

y = np.reshape(tmp, [b, c // (blocksize**2), h * blocksize, w * blocksize])

In the CRD mode, elements along the depth dimension from the input tensor are rearranged in the following order: column, row, and the depth. The output y is computed from the input x as below:

b, c, h, w = x.shape

tmp = np.reshape(x, [b, c // (blocksize ** 2), blocksize, blocksize, h, w])

tmp = np.transpose(tmp, [0, 1, 4, 2, 5, 3])

y = np.reshape(tmp, [b, c // (blocksize ** 2), h * blocksize, w * blocksize])

Node Inputs

input (T, single) – Input tensor of [N,C,H,W], where N is the batch axis, C is the channel or depth, H is the height and W is the width.

Node Outputs

output (T, single) – Output tensor of [N, C/(blocksize * blocksize), H * blocksize, W * blocksize].

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128

Parameters

name – the name of the node.
blocksize (int) – Blocks of [blocksize, blocksize] are moved.
mode (Optional [str], default=’DCR’) – DCR (default) for depth-column-row order re-arrangement. Use CRD for column-row-depth order.

blocksize¶: Blocks of [blocksize, blocksize] are moved.

mode¶: DCR (default) for depth-column-row order re-arrangement. Use CRD for column-row-depth order.

class ONNXDequantizeLinear(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

The linear dequantization operator. It consumes a quantized tensor, a scale, a zero point to compute the full precision tensor. The dequantization formula is y = (x - x_zero_point) * x_scale. ‘x_scale’ and ‘x_zero_point’ must have same shape. ‘x_zero_point’ and ‘x’ must have same type. ‘x’ and ‘y’ must have same shape. In the case of dequantizing int32, there’s no zero point (zero point is supposed to be 0).

Node Inputs

x (T, single) – N-D quantized input tensor to be de-quantized.
x_scale (x_scale_constraint, single) – Scale for input ‘x’. It’s a scalar, which means a per-tensor/layer quantization.
x_zero_point (T, optional) – Zero point for input ‘x’. It’s a scalar, which means a per-tensor/layer quantization. It’s optional. 0 is the default value when it’s not specified.

Node Outputs

y (y_constraint, single) – N-D full precision output tensor. It has same shape as input ‘x’.

Type Constraints

T – int8, uint8, int32
x_scale_constraint – float32
y_constraint – float32

Parameters

name – the name of the node.

class ONNXDet(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Det calculates determinant of a square matrix or batches of square matrices. Det takes one input tensor of shape [*, M, M], where * is zero or more batch dimensions, and the inner-most 2 dimensions form square matrices. The output is a tensor of shape [*], containing the determinants of all input submatrices. e.g., When the input is 2-D, the output is a scalar(shape is empty: []).

Node Inputs

X (T, single) – Input tensor

Node Outputs

Y (T, single) – Output tensor

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXDiv(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Performs element-wise binary division (with Numpy-style broadcasting support).

This operator supports multidirectional (i.e., Numpy-style) broadcasting; for more details please check [the doc](Broadcasting.md).

Node Inputs

A (T, single) – First operand.
B (T, single) – Second operand.

Node Outputs

C (T, single) – Result, has same element type as two inputs

Type Constraints

T – uint32, uint64, int32, int64, float16, float32, float64

Parameters

name – the name of the node.

class ONNXDropout(name, *, seed=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Dropout takes an input floating-point tensor, an optional input ratio (floating-point scalar) and an optional input training_mode (boolean scalar). It produces two tensor outputs, output (floating-point tensor) and mask (optional Tensor<bool>). If training_mode is true then the output Y will be a random dropout; Note that this Dropout scales the masked input data by the following equation, so to convert the trained model into inference mode, the user can simply not pass training_mode input or set it to false. ` output = scale * data * mask, ` where ` scale = 1. / (1. - ratio). ` This operator has optional inputs/outputs. See [the doc](IR.md) for more details about the representation of optional arguments. An empty string may be used in the place of an actual argument’s name to indicate a missing argument. Trailing optional arguments (those not followed by an argument that is present) may also be simply omitted.

Node Inputs

data (T, single) – The input data as Tensor.
ratio (T1, optional) – The ratio of random dropout, with value in [0, 1). If this input was not set, or if it was set to 0, the output would be a simple copy of the input. If it’s non-zero, output will be a random dropout of the scaled input, which is typically the case during training. It is an optional value, if not specified it will default to 0.5.
training_mode (T2, optional) – If set to true then it indicates dropout is being used for training. It is an optional value hence unless specified explicitly, it is false. If it is false, ratio is ignored and the operation mimics inference mode where nothing will be dropped from the input data and if mask is requested as output it will contain all ones.

Node Outputs

output (T, single) – The output.
mask (T2, optional) – The output mask.

Type Constraints

T – float16, float32, float64
T1 – float16, float32, float64
T2 – bool_

Parameters

name – the name of the node.
seed (Optional [int], default=None) – (Optional) Seed to the random generator, if not specified we will auto generate one.

seed¶: (Optional) Seed to the random generator, if not specified we will auto generate one.

class ONNXDynamicQuantizeLinear(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

A Function to fuse calculation for Scale, Zero Point and FP32->8Bit convertion of FP32 Input data. Outputs Scale, ZeroPoint and Quantized Input for a given FP32 Input. Scale is calculated as: ```

y_scale = (max(x) - min(x))/(qmax - qmin) * where qmax and qmin are max and min values for quantization range .i.e [0, 255] in case of uint8 * data range is adjusted to include 0.

` Zero point is calculated as: ` intermediate_zero_point = qmin - min(x)/y_scale y_zero_point = cast(round(saturate(itermediate_zero_point))) * where qmax and qmin are max and min values for quantization range .i.e [0, 255] in case of uint8 * for saturation, it saturates to [0, 255] if it’s uint8, or [-127, 127] if it’s int8. Right now only uint8 is supported. * rounding to nearest ties to even. ` Data quantization formula is: ` y = saturate (round (x / y_scale) + y_zero_point) * for saturation, it saturates to [0, 255] if it’s uint8, or [-127, 127] if it’s int8. Right now only uint8 is supported. * rounding to nearest ties to even. ```

Node Inputs

x (T1, single) – Input tensor

Node Outputs

y (T2, single) – Quantized output tensor
y_scale (y_scale_constraint, single) – Output scale. It’s a scalar, which means a per-tensor/layer quantization.
y_zero_point (T2, single) – Output zero point. It’s a scalar, which means a per-tensor/layer quantization.

Type Constraints

T1 – float32
T2 – uint8
y_scale_constraint – float32

Parameters

name – the name of the node.

class ONNXEinsum(name, *, equation)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

An einsum of the form `term1, term2 -> output-term` produces an output tensor using the following equation

`output[output-term] = reduce-sum( input1[term1] * input2[term] )`

where the reduce-sum performs a summation over all the indices occurring in in the input terms (term1, term2) that do not occur in the output-term.

The Einsum operator evaluates algebraic tensor operations on a sequence of tensors, using the Einstein summation convention. The equation string contains a comma-separated sequence of lower case letters. Each term corresponds to an operand tensor, and the characters within the terms correspond to operands dimensions.

This sequence may be followed by “->” to separate the left and right hand side of the equation. If the equation contains “->” followed by the right-hand side, the explicit (not classical) form of the Einstein summation is performed, and the right-hand side indices indicate output tensor dimensions. In other cases, output indices are (implicitly) set to the alphabetically sorted sequence of indices appearing exactly once in the equation.

When a dimension character is repeated in the left-hand side, it represents summation along the dimension.

The equation may contain ellipsis (“…”) to enable broadcasting. Ellipsis must indicate a fixed number of dimensions. Specifically, every occurrence of ellipsis in the equation must represent the same number of dimensions. The right-hand side may contain exactly one ellipsis. In implicit mode, the ellipsis dimensions are set to the beginning of the output. The equation string may contain space (U+0020) character.

Node Inputs

Inputs (T, variadic) – Operands

Node Outputs

Output (T, single) – Output tensor

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64

Parameters

name – the name of the node.
equation (str) – Einsum expression string.

equation¶: Einsum expression string.

class ONNXElu(name, *, alpha=1.0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Elu takes one input data (Tensor<T>) and produces one output data (Tensor<T>) where the function f(x) = alpha * (exp(x) - 1.) for x < 0, f(x) = x for x >= 0., is applied to the tensor elementwise.

Node Inputs

X (T, single) – 1D input tensor

Node Outputs

Y (T, single) – 1D input tensor

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.
alpha (Optional [float], default=1.0) – Coefficient of ELU.

alpha¶: Coefficient of ELU.

class ONNXEqual(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Returns the tensor resulted from performing the equal logical operation elementwise on the input tensors A and B (with Numpy-style broadcasting support).

This operator supports multidirectional (i.e., Numpy-style) broadcasting; for more details please check [the doc](Broadcasting.md).

Node Inputs

A (T, single) – First input operand for the logical operator.
B (T, single) – Second input operand for the logical operator.

Node Outputs

C (T1, single) – Result tensor.

Type Constraints

T – bool_, uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64
T1 – bool_

Parameters

name – the name of the node.

class ONNXErf(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Computes the error function of the given input tensor element-wise.

Node Inputs

input (T, single) – Input tensor

Node Outputs

output (T, single) – The error function of the input tensor computed element-wise. It has the same shape and type of the input.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64

Parameters

name – the name of the node.

class ONNXExp(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Calculates the exponential of the given input tensor, element-wise.

Node Inputs

input (T, single) – Input tensor

Node Outputs

output (T, single) – The exponential of the input tensor computed element-wise

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXExpand(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Broadcast the input tensor following the given shape and the broadcast rule. The broadcast rule is similar to numpy.array(input) * numpy.ones(shape): Dimensions are right alignment; Two corresponding dimension must have the same value, or one of them is equal to 1. Also, this operator is similar to numpy.broadcast_to(input, shape), but the major difference is numpy.broadcast_to() does not allow shape to be smaller than input.size(). It is possible that the output.shape is not equal to shape, when some dimensions in shape is equal to 1, or the shape.ndim < input.shape.ndim.

Node Inputs

input (T, single) – Input tensor
shape (shape_constraint, single) – A 1-D tensor indicates the shape you want to expand to, following the broadcast rule

Node Outputs

output (T, single) – Output tensor

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128
shape_constraint – int64

Parameters

name – the name of the node.

class ONNXEyeLike(name, *, dtype=None, k=0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Generate a 2D tensor (matrix) with ones on the diagonal and zeros everywhere else. Only 2D tensors are supported, i.e. input T1 must be of rank 2. The shape of the output tensor is the same as the input tensor. The data type can be specified by the ‘dtype’ argument. If ‘dtype’ is not specified, then the type of input tensor is used. By default, the main diagonal is populated with ones, but attribute ‘k’ can be used to populate upper or lower diagonals. The ‘dtype’ argument must be one of the data types specified in the ‘DataType’ enum field in the TensorProto message and be valid as an output type.

Node Inputs

input (T1, single) – 2D input tensor to copy shape, and optionally, type information from.

Node Outputs

output (T2, single) – Output tensor, same shape as input tensor T1.

Type Constraints

T1 – float16, float32, float64, int8, int16, int32, int64, uint8, uint16, uint32, uint64, bool_
T2 – float16, float32, float64, int8, int16, int32, int64, uint8, uint16, uint32, uint64, bool_

Parameters

name – the name of the node.
dtype (Optional [int], default=None) – (Optional) The data type for the elements of the output tensor. If not specified,the data type of the input tensor T1 is used. If input tensor T1 is also notspecified, then type defaults to ‘float’.
k (Optional [int], default=0) – (Optional) Index of the diagonal to be populated with ones. Default is 0. If T2 is the output, this op sets T2[i, i+k] = 1. k = 0 populates the main diagonal, k > 0 populates an upper diagonal, and k < 0 populates a lower diagonal.

dtype¶: (Optional) The data type for the elements of the output tensor. If not specified,the data type of the input tensor T1 is used. If input tensor T1 is also notspecified, then type defaults to ‘float’.

k¶: (Optional) Index of the diagonal to be populated with ones. Default is 0. If T2 is the output, this op sets T2[i, i+k] = 1. k = 0 populates the main diagonal, k > 0 populates an upper diagonal, and k < 0 populates a lower diagonal.

class ONNXFeatureVectorizer(name, *, inputdimensions=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Concatenates input tensors into one continuous output. All input shapes are 2-D and are concatenated along the second dimention. 1-D tensors are treated as [1,C]. Inputs are copied to the output maintaining the order of the input arguments. All inputs must be integers or floats, while the output will be all floating point values.

Node Inputs

X (T1, variadic) – An ordered collection of tensors, all with the same element type.

Node Outputs

Y (Y_constraint, single) – The output array, elements ordered as the inputs.

Type Constraints

T1 – int32, int64, float32, float64
Y_constraint – float32

Parameters

name – the name of the node.
inputdimensions (Optional [List [int]], default=None) – The size of each input in the input list

inputdimensions¶: The size of each input in the input list

class ONNXFlatten(name, *, axis=1)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Flattens the input tensor into a 2D matrix. If input tensor has shape (d_0, d_1, … d_n) then the output will have shape (d_0 X d_1 … d_(axis-1), d_axis X d_(axis+1) … X dn).

Node Inputs

input (T, single) – A tensor of rank >= axis.

Node Outputs

output (T, single) – A 2D tensor with the contents of the input tensor, with input dimensions up to axis flattened to the outer dimension of the output and remaining input dimensions flattened into the inner dimension of the output.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128

Parameters

name – the name of the node.
axis (Optional [int], default=1) – Indicate up to which input dimensions (exclusive) should be flattened to the outer dimension of the output. The value for axis must be in the range [-r, r], where r is the rank of the input tensor. Negative value means counting dimensions from the back. When axis = 0, the shape of the output tensor is (1, (d_0 X d_1 … d_n), where the shape of the input tensor is (d_0, d_1, … d_n).

axis¶: Indicate up to which input dimensions (exclusive) should be flattened to the outer dimension of the output. The value for axis must be in the range [-r, r], where r is the rank of the input tensor. Negative value means counting dimensions from the back. When axis = 0, the shape of the output tensor is (1, (d_0 X d_1 … d_n), where the shape of the input tensor is (d_0, d_1, … d_n).

class ONNXFloor(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Floor takes one input data (Tensor<T>) and produces one output data (Tensor<T>) where the floor is, y = floor(x), is applied to the tensor elementwise.

Node Inputs

X (T, single) – Input tensor

Node Outputs

Y (T, single) – Output tensor

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXGRU(name, *, activation_alpha=None, activation_beta=None, activations=None, clip=None, direction='forward', hidden_size=None, linear_before_reset=0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Computes an one-layer GRU. This operator is usually supported via some custom implementation such as CuDNN.

Notations:

X - input tensor

z - update gate

r - reset gate

h - hidden gate

t - time step (t-1 means previous time step)

W[zrh] - W parameter weight matrix for update, reset, and hidden gates

R[zrh] - R recurrence weight matrix for update, reset, and hidden gates

Wb[zrh] - W bias vectors for update, reset, and hidden gates

Rb[zrh] - R bias vectors for update, reset, and hidden gates

WB[zrh] - W parameter weight matrix for backward update, reset, and hidden gates

RB[zrh] - R recurrence weight matrix for backward update, reset, and hidden gates

WBb[zrh] - W bias vectors for backward update, reset, and hidden gates

RBb[zrh] - R bias vectors for backward update, reset, and hidden gates

H - Hidden state

num_directions - 2 if direction == bidirectional else 1

Activation functions:

Relu(x) - max(0, x)

Tanh(x) - (1 - e^{-2x})/(1 + e^{-2x})

Sigmoid(x) - 1/(1 + e^{-x})

(NOTE: Below are optional)

Affine(x) - alpha*x + beta

LeakyRelu(x) - x if x >= 0 else alpha * x

ThresholdedRelu(x) - x if x >= alpha else 0

ScaledTanh(x) - alpha*Tanh(beta*x)

HardSigmoid(x) - min(max(alpha*x + beta, 0), 1)

Elu(x) - x if x >= 0 else alpha*(e^x - 1)

Softsign(x) - x/(1 + |x|)

Softplus(x) - log(1 + e^x)

Equations (Default: f=Sigmoid, g=Tanh):

zt = f(Xt*(Wz^T) + Ht-1*(Rz^T) + Wbz + Rbz)

rt = f(Xt*(Wr^T) + Ht-1*(Rr^T) + Wbr + Rbr)

ht = g(Xt*(Wh^T) + (rt (.) Ht-1)*(Rh^T) + Rbh + Wbh) # default, when linear_before_reset = 0

ht = g(Xt*(Wh^T) + (rt (.) (Ht-1*(Rh^T) + Rbh)) + Wbh) # when linear_before_reset != 0

Ht = (1 - zt) (.) ht + zt (.) Ht-1

This operator has optional inputs/outputs. See [the doc](IR.md) for more details about the representation of optional arguments. An empty string may be used in the place of an actual argument’s name to indicate a missing argument. Trailing optional arguments (those not followed by an argument that is present) may also be simply omitted.

Node Inputs

X (T, single) – The input sequences packed (and potentially padded) into one 3-D tensor with the shape of [seq_length, batch_size, input_size].
W (T, single) – The weight tensor for the gates. Concatenation of W[zrh] and WB[zrh] (if bidirectional) along dimension 0. This tensor has shape [num_directions, 3*hidden_size, input_size].
R (T, single) – The recurrence weight tensor. Concatenation of R[zrh] and RB[zrh] (if bidirectional) along dimension 0. This tensor has shape [num_directions, 3*hidden_size, hidden_size].
B (T, optional) – The bias tensor for the gates. Concatenation of [Wb[zrh], Rb[zrh]] and [WBb[zrh], RBb[zrh]] (if bidirectional) along dimension 0. This tensor has shape [num_directions, 6*hidden_size]. Optional: If not specified - assumed to be 0
sequence_lens (T1, optional) – Optional tensor specifying lengths of the sequences in a batch. If not specified - assumed all sequences in the batch to have length seq_length. It has shape [batch_size].
initial_h (T, optional) – Optional initial value of the hidden. If not specified - assumed to be 0. It has shape [num_directions, batch_size, hidden_size].

Node Outputs

Y (T, optional) – A tensor that concats all the intermediate output values of the hidden. It has shape [seq_length, num_directions, batch_size, hidden_size].
Y_h (T, optional) – The last output value of the hidden. It has shape [num_directions, batch_size, hidden_size].

Type Constraints

T – float16, float32, float64
T1 – int32

Parameters

name – the name of the node.
activation_alpha (Optional [List [float]], default=None) – Optional scaling values used by some activation functions. The values are consumed in the order of activation functions, for example (f, g, h) in LSTM. Default values are the same as of corresponding ONNX operators.For example with LeakyRelu, the default alpha is 0.01.
activation_beta (Optional [List [float]], default=None) – Optional scaling values used by some activation functions. The values are consumed in the order of activation functions, for example (f, g, h) in LSTM. Default values are the same as of corresponding ONNX operators.
activations (Optional [List [str]], default=None) – A list of 2 (or 4 if bidirectional) activation functions for update, reset, and hidden gates. The activation functions must be one of the activation functions specified above. Optional: See the equations for default if not specified.
clip (Optional [float], default=None) – Cell clip threshold. Clipping bounds the elements of a tensor in the range of [-threshold, +threshold] and is applied to the input of activations. No clip if not specified.
direction (Optional [str], default=’forward’) – Specify if the RNN is forward, reverse, or bidirectional. Must be one of forward (default), reverse, or bidirectional.
hidden_size (Optional [int], default=None) – Number of neurons in the hidden layer
linear_before_reset (Optional [int], default=0) – When computing the output of the hidden gate, apply the linear transformation before multiplying by the output of the reset gate.

activation_alpha¶: Optional scaling values used by some activation functions. The values are consumed in the order of activation functions, for example (f, g, h) in LSTM. Default values are the same as of corresponding ONNX operators.For example with LeakyRelu, the default alpha is 0.01.

activation_beta¶: Optional scaling values used by some activation functions. The values are consumed in the order of activation functions, for example (f, g, h) in LSTM. Default values are the same as of corresponding ONNX operators.

activations¶: A list of 2 (or 4 if bidirectional) activation functions for update, reset, and hidden gates. The activation functions must be one of the activation functions specified above. Optional: See the equations for default if not specified.

clip¶: Cell clip threshold. Clipping bounds the elements of a tensor in the range of [-threshold, +threshold] and is applied to the input of activations. No clip if not specified.

direction¶: Specify if the RNN is forward, reverse, or bidirectional. Must be one of forward (default), reverse, or bidirectional.

hidden_size¶: Number of neurons in the hidden layer

linear_before_reset¶: When computing the output of the hidden gate, apply the linear transformation before multiplying by the output of the reset gate.

class ONNXGather(name, *, axis=0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Given data tensor of rank r >= 1, and indices tensor of rank q, gather entries of the axis dimension of data (by default outer-most one as axis=0) indexed by indices, and concatenates them in an output tensor of rank q + (r - 1).

axis = 0 :

Let k = indices[i_{0}, …, i_{q-1}] Then output[i_{0}, …, i_{q-1}, j_{0}, …, j_{r-2}] = input[k , j_{0}, …, j_{r-2}]

```

data = [: [1.0, 1.2], [2.3, 3.4], [4.5, 5.7],

] indices = [

[0, 1], [1, 2],

] output = [

[
[1.0, 1.2], [2.3, 3.4],

], [

[2.3, 3.4], [4.5, 5.7],

],

]

``` axis = 1 :

Let k = indices[i_{0}, …, i_{q-1}] Then output[i_{0}, …, i_{q-1}, j_{0}, …, j_{r-2}] = input[j_{0}, k, j_{1}, …, j_{r-2}]

```

data = [: [1.0, 1.2, 1.9], [2.3, 3.4, 3.9], [4.5, 5.7, 5.9],

] indices = [

[0, 2],

] axis = 1, output = [

[
[1.0, 1.9], [2.3, 3.9], [4.5, 5.9],

],

]

```

Node Inputs

data (T, single) – Tensor of rank r >= 1.
indices (Tind, single) – Tensor of int32/int64 indices, of any rank q. All index values are expected to be within bounds [-s, s-1] along axis of size s. It is an error if any of the index values are out of bounds.

Node Outputs

output (T, single) – Tensor of rank q + (r - 1).

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128
Tind – int32, int64

Parameters

name – the name of the node.
axis (Optional [int], default=0) – Which axis to gather on. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(data).

axis¶: Which axis to gather on. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(data).

class ONNXGatherElements(name, *, axis=0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

GatherElements takes two inputs data and indices of the same rank r >= 1 and an optional attribute axis that identifies an axis of data (by default, the outer-most axis, that is axis 0). It is an indexing operation that produces its output by indexing into the input data tensor at index positions determined by elements of the indices tensor. Its output shape is the same as the shape of indices and consists of one value (gathered from the data) for each element in indices.

For instance, in the 3-D case (r = 3), the output produced is determined by the following equations: ```

out[i][j][k] = input[index[i][j][k]][j][k] if axis = 0, out[i][j][k] = input[i][index[i][j][k]][k] if axis = 1, out[i][j][k] = input[i][j][index[i][j][k]] if axis = 2,

```

This operator is also the inverse of ScatterElements. It is similar to Torch’s gather operation.

Example 1: ```

data = [
[1, 2], [3, 4],

] indices = [

[0, 0], [1, 0],

] axis = 1 output = [

[
[1, 1], [4, 3],

],

]

` Example 2: `

data = [
[1, 2, 3], [4, 5, 6], [7, 8, 9],

] indices = [

[1, 2, 0], [2, 0, 0],

] axis = 0 output = [

[
[4, 8, 3], [7, 2, 3],

],

]

```

Node Inputs

data (T, single) – Tensor of rank r >= 1.
indices (Tind, single) – Tensor of int32/int64 indices, with the same rank r as the input. All index values are expected to be within bounds [-s, s-1] along axis of size s. It is an error if any of the index values are out of bounds.

Node Outputs

output (T, single) – Tensor of the same shape as indices.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128
Tind – int32, int64

Parameters

name – the name of the node.
axis (Optional [int], default=0) – Which axis to gather on. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(data).

axis¶: Which axis to gather on. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(data).

class ONNXGatherND(name, *, batch_dims=0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Given data tensor of rank r >= 1, indices tensor of rank q >= 1, and batch_dims integer b, this operator gathers slices of data into an output tensor of rank q + r - indices_shape[-1] - 1 - b.

indices is an q-dimensional integer tensor, best thought of as a (q-1)-dimensional tensor of index-tuples into data, where each element defines a slice of data

batch_dims (denoted as b) is an integer indicating the number of batch dimensions, i.e the leading b number of dimensions of data tensor and indices are representing the batches, and the gather starts from the b+1 dimension.

Some salient points about the inputs’ rank and shape:

r >= 1 and q >= 1 are to be honored. There is no dependency condition to be met between ranks r and q
The first b dimensions of the shape of indices tensor and data tensor must be equal.
b < min(q, r) is to be honored.
The indices_shape[-1] should have a value between 1 (inclusive) and rank r-b (inclusive)
All values in indices are expected to be within bounds [-s, s-1] along axis of size s (i.e.) -data_shape[i] <= indices[…,i] <= data_shape[i] - 1. It is an error if any of the index values are out of bounds.

The output is computed as follows:

The output tensor is obtained by mapping each index-tuple in the indices tensor to the corresponding slice of the input data.

If indices_shape[-1] > r-b => error condition
If indices_shape[-1] == r-b, since the rank of indices is q, indices can be thought of as N (q-b-1)-dimensional tensors containing 1-D tensors of dimension r-b, where N is an integer equals to the product of 1 and all the elements in the batch dimensions of the indices_shape. Let us think of each such r-b ranked tensor as indices_slice. Each scalar value corresponding to data[0:b-1,indices_slice] is filled into the corresponding location of the (q-b-1)-dimensional tensor to form the output tensor (Example 1 below)
If indices_shape[-1] < r-b, since the rank of indices is q, indices can be thought of as N (q-b-1)-dimensional tensor containing 1-D tensors of dimension < r-b. Let us think of each such tensors as indices_slice. Each tensor slice corresponding to data[0:b-1, indices_slice , :] is filled into the corresponding location of the (q-b-1)-dimensional tensor to form the output tensor (Examples 2, 3, 4 and 5 below)

This operator is the inverse of ScatterND.

Example 1

batch_dims = 0

data = [[0,1],[2,3]] # data_shape = [2, 2]

indices = [[0,0],[1,1]] # indices_shape = [2, 2]

output = [0,3] # output_shape = [2]

Example 2

batch_dims = 0

data = [[0,1],[2,3]] # data_shape = [2, 2]

indices = [[1],[0]] # indices_shape = [2, 1]

output = [[2,3],[0,1]] # output_shape = [2, 2]

Example 3

batch_dims = 0

data = [[[0,1],[2,3]],[[4,5],[6,7]]] # data_shape = [2, 2, 2]

indices = [[0,1],[1,0]] # indices_shape = [2, 2]

output = [[2,3],[4,5]] # output_shape = [2, 2]

Example 4

batch_dims = 0

data = [[[0,1],[2,3]],[[4,5],[6,7]]] # data_shape = [2, 2, 2]

indices = [[[0,1]],[[1,0]]] # indices_shape = [2, 1, 2]

output = [[[2,3]],[[4,5]]] # output_shape = [2, 1, 2]

Example 5

batch_dims = 1

data = [[[0,1],[2,3]],[[4,5],[6,7]]] # data_shape = [2, 2, 2]

indices = [[1],[0]] # indices_shape = [2, 1]

output = [[2,3],[4,5]] # output_shape = [2, 2]

Node Inputs

data (T, single) – Tensor of rank r >= 1.
indices (indices_constraint, single) – Tensor of rank q >= 1. All index values are expected to be within bounds [-s, s-1] along axis of size s. It is an error if any of the index values are out of bounds.

Node Outputs

output (T, single) – Tensor of rank q + r - indices_shape[-1] - 1.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128
indices_constraint – int64

Parameters

name – the name of the node.
batch_dims (Optional [int], default=0) – The number of batch dimensions. The gather of indexing starts from dimension of data[batch_dims:]

batch_dims¶: The number of batch dimensions. The gather of indexing starts from dimension of data[batch_dims:]

class ONNXGemm(name, *, alpha=1.0, beta=1.0, transA=0, transB=0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

General Matrix multiplication: https://en.wikipedia.org/wiki/Basic_Linear_Algebra_Subprograms#Level_3

A’ = transpose(A) if transA else A

B’ = transpose(B) if transB else B

Compute Y = alpha * A’ * B’ + beta * C, where input tensor A has shape (M, K) or (K, M), input tensor B has shape (K, N) or (N, K), input tensor C is broadcastable to shape (M, N), and output tensor Y has shape (M, N). A will be transposed before doing the computation if attribute transA is non-zero, same for B and transB. This operator supports unidirectional broadcasting (tensor C should be unidirectional broadcastable to tensor A * B); for more details please check [the doc](Broadcasting.md). This operator has optional inputs/outputs. See [the doc](IR.md) for more details about the representation of optional arguments. An empty string may be used in the place of an actual argument’s name to indicate a missing argument. Trailing optional arguments (those not followed by an argument that is present) may also be simply omitted.

Node Inputs

A (T, single) – Input tensor A. The shape of A should be (M, K) if transA is 0, or (K, M) if transA is non-zero.
B (T, single) – Input tensor B. The shape of B should be (K, N) if transB is 0, or (N, K) if transB is non-zero.
C (T, optional) – Optional input tensor C. If not specified, the computation is done as if C is a scalar 0. The shape of C should be unidirectional broadcastable to (M, N).

Node Outputs

Y (T, single) – Output tensor of shape (M, N).

Type Constraints

T – float16, float32, float64, uint32, uint64, int32, int64

Parameters

name – the name of the node.
alpha (Optional [float], default=1.0) – Scalar multiplier for the product of input tensors A * B.
beta (Optional [float], default=1.0) – Scalar multiplier for input tensor C.
transA (Optional [int], default=0) – Whether A should be transposed
transB (Optional [int], default=0) – Whether B should be transposed

alpha¶: Scalar multiplier for the product of input tensors A * B.

beta¶: Scalar multiplier for input tensor C.

transA¶: Whether A should be transposed

transB¶: Whether B should be transposed

class ONNXGlobalAveragePool(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

GlobalAveragePool consumes an input tensor X and applies average pooling across the values in the same channel. This is equivalent to AveragePool with kernel size equal to the spatial dimension of input tensor.

Node Inputs

X (T, single) – Input data tensor from the previous operator; dimensions for image case are (N x C x H x W), where N is the batch size, C is the number of channels, and H and W are the height and the width of the data. For non image case, the dimensions are in the form of (N x C x D1 x D2 … Dn), where N is the batch size.

Node Outputs

Y (T, single) – Output data tensor from pooling across the input tensor. The output tensor has the same rank as the input. The first two dimensions of output shape are the same as the input (N x C), while the other dimensions are all 1.

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXGlobalLpPool(name, *, p=2)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

GlobalLpPool consumes an input tensor X and applies lp pool pooling across the values in the same channel. This is equivalent to LpPool with kernel size equal to the spatial dimension of input tensor.

Node Inputs

X (T, single) – Input data tensor from the previous operator; dimensions for image case are (N x C x H x W), where N is the batch size, C is the number of channels, and H and W are the height and the width of the data. For non image case, the dimensions are in the form of (N x C x D1 x D2 … Dn), where N is the batch size.

Node Outputs

Y (T, single) – Output data tensor from pooling across the input tensor. The output tensor has the same rank as the input. The first two dimensions of output shape are the same as the input (N x C), while the other dimensions are all 1.

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.
p (Optional [int], default=2) – p value of the Lp norm used to pool over the input data.

p¶: p value of the Lp norm used to pool over the input data.

class ONNXGlobalMaxPool(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

GlobalMaxPool consumes an input tensor X and applies max pooling across the values in the same channel. This is equivalent to MaxPool with kernel size equal to the spatial dimension of input tensor.

Node Inputs

X (T, single) – Input data tensor from the previous operator; dimensions for image case are (N x C x H x W), where N is the batch size, C is the number of channels, and H and W are the height and the width of the data. For non image case, the dimensions are in the form of (N x C x D1 x D2 … Dn), where N is the batch size.

Node Outputs

Y (T, single) – Output data tensor from pooling across the input tensor. The output tensor has the same rank as the input. The first two dimensions of output shape are the same as the input (N x C), while the other dimensions are all 1.

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXGradient(name, *, xs, y, zs=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Gradient operator computes the partial derivatives of a specific tensor w.r.t. some other tensors. This operator is widely used in gradient-based training algorithms. To illustrate its use, let’s consider a computation graph,

``` X —–.

v

W –> Conv –> H –> Gemm –> Y: Z

```

, where W and Z are trainable tensors. Note that operators’ attributes are omitted for the sake of simplicity. Let dY/dW (dY/dZ) be the gradient of Y with respect to W (Z). The user can compute gradient by inserting Gradient operator to form another graph shown below.

``` W –> Conv –> H –> Gemm –> Y | ^ ^ | | | | X Z | | | | | .———-‘ | | | (W/Z/X is the 1st/2nd/3rd input of Gradient as shown in | | | “xs” followed by “zs”) | v v ‘—> Gradient(xs=[“W”, “Z”], zs=[“X”], y=”Y”)

|

‘———————————–> dY/dW (1st output of Gradient)

‘—————————————> dY/dZ (2nd output of Gradient)

```

By definition, the tensor “y” is a function of independent variables in “xs” and “zs”. Since we only compute the gradient of “y” w.r.t. the differentiable variables in “xs”, this Gradient only outputs dY/dW and dY/dZ. Note that “H” cannot appear in “xs” and “zs”. The reason is that “H” can be determined by tensors “W” and “X” and therefore “H” is not an independent variable.

All outputs are optional. If needed, for example, user can assign an empty string to the 1st output name of that Gradient to skip the generation of dY/dW. Note that the concept of optional outputs can also be found in ONNX’s RNN, GRU, and LSTM.

Gradient operator can compute derivative against intermediate tensors. For example, the gradient of Y with respect to H can be done via

``` W –> Conv –> H –> Gemm –> Y

^ | ^ | | | X | Z .——-‘ | | .———-‘ | | (H/Z is the 1st/2nd input of Gradient as shown in “xs”) v v

Gradient(xs=[“H”, “Z”], y=”Y”)

|

‘———————————–> dY/dH (1st output of Gradient)

‘—————————————> dY/dZ (2nd output of Gradient)

```

It is possible to represent high-order differentiation using Gradient operators. For example, given the following linear model:

``` W –> Gemm –> Y –> Loss –> O

^ ^ | | X L

```

To compute the 2nd order derivative of O with respect to W (denoted by d^2O/dW^2), one can do

Gradient)

‘—> d^2O/dW^2 (2nd output of Gradient)

```

The tensors named in attributes “xs”, “zs”, and “y” define the differentiated computation graph, and the inputs to Gradient node define the values at which the gradient is computed. We can feed different tensors to the identified graph. For example, one can compute the gradient of Y with respect to H at a specific value of H, H_1, by providing that value as an input to the Gradient node.

``` W –> Conv –> H –> Gemm –> Y

^ ^ | | X Z

Z_1 (2nd input of Gradient)

v

H_1 –> Gradient(xs=[“H”, “Z”], y=”Y”) —> dY/dH when H = H_1 and Y = Y_1.: ‘——————————> dY/dZ (2nd output of Gradient)

```

When the inputs of Gradient are the tensors named in “xs” and “zs”, the computation can be optimized. More specifically, intermediate variables in forward pass can be reused if the gradient is computed via reverse-mode auto-differentiation.

Node Inputs

Inputs (T1, variadic) – The values fed into graph identified by the attributes. The i-th input is the value of the i-th tensor specified in the concatenated list of the attribute “xs” and the attribute “zs”. For example, if xs=[“A”, “B”] and zs=[“C”], the first input is used as the value of symbol “A” and the 3rd input is substituted for all the occurrences of “C”.

Node Outputs

Outputs (T2, variadic) – The gradient of the tensor specified by the attribute “y” with respect to each of tensors specified in the attribute “xs”. The i-th output is the gradient of “y” with respect to the i-th tensor specified in the attribute “xs”.

Type Constraints

T1 – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128
T2 – float16, float32, float64

Parameters

name – the name of the node.
xs (List [str]) – Input tensor names of the differentiated sub-graph. It contains only the necessary differentiated inputs of a (sub-)graph. Variables (usually called intermediate variables) that can be generated from inputs cannot be included in this attribute.
y (str) – The targeted tensor. It can be viewed as the output of the differentiated function. The attribute “xs” and attribute “zs” are the minimal independent variable set that determines the value of “y”.
zs (Optional [List [str]], default=None) – Input tensor names of the differentiated sub-graph. It contains only the necessary non-differentiated inputs of a (sub-)graph. Variables (usually called intermediate variables) that can be generated from inputs cannot be included in this attribute.

xs¶: Input tensor names of the differentiated sub-graph. It contains only the necessary differentiated inputs of a (sub-)graph. Variables (usually called intermediate variables) that can be generated from inputs cannot be included in this attribute.

y¶: The targeted tensor. It can be viewed as the output of the differentiated function. The attribute “xs” and attribute “zs” are the minimal independent variable set that determines the value of “y”.

zs¶: Input tensor names of the differentiated sub-graph. It contains only the necessary non-differentiated inputs of a (sub-)graph. Variables (usually called intermediate variables) that can be generated from inputs cannot be included in this attribute.

class ONNXGraphCall(name, *, graph_name)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

The GraphCall operator invokes a graph inside TrainingInfoProto’s algorithm field. The GraphCall inputs and outputs are bound to those of invoked graph by position. If a graph input has an initializer, that input is considered optional. All graph outputs are optional.

Below Python syntax is used for describing dictionary and list.

Assume that ModelProto’s graph field has - name: “MyInferenceGraph” - input: [“X”, “W”, “Z”] - initializer: [W] - output: [“Y”]

as visualized below for inference.

``` X —–.

v

W –> Conv –> H –> Gemm –> Y: Z

```

Assume that the training algorithm contains

inputs: [“X_1”, “Z_1”, “C”]
initializer: [T]
outputs: [“W_new”]

with a dictionary

update_binding: {“W”: “W_new”, “T”: “T_new”}

Inside the training algorithm graph, one can invoke the inference graph via adding a GraphCall node with

inputs: [“X_1”, “W”, Z_1”]
outputs: [“Y_1”]
an attribute graph_name=”MyInferenceGraph”,

The initializers, “W” and “T” in this case, in update_binding are considered globally-visible and mutable variables, which can be used as inputs of operators in the training graph.

An example training algorithm graph may look like

` .-------- W (a global and mutable variable from | | the inference graph) | | | .-----'-----------. | | | | | v | | .-- X_1 --> GraphCall(graph_name="MyInferenceGraph") | | | | | | | | | | | | | Z_1 -----' | | | | | V | | | | Y_1 ---> Loss ---> O | | | | ^ | | | | | | | `--. | C | | | | | | | | | .----------------' | | | | | | | v v v | `--> Gradient(xs=["W"], zs=["X_1", "Z_1", "C"], y="O") | | | v | dO_dW (gradient of W) 1 (a scalar one) | | | | V v | Div <--- T ------------> Add ---> T_new | | (T is the number of training iterations. | | T is also globally visible and mutable.) | v `-----> Sub ----> W_new `

where Loss is a dummy node which computes the minimized objective function.

The variable “W” is an optional input in the called graph. If the user omits it, the input list of GraphCall becomes [“X_1”, “”, “Z_1”]. In this case, from the view of computation graph, the Conv operator invoked by GraphCall’s may be still connected the global “W” variable and therefore the structure of the computation graph is unchanged.

Node Inputs

Inputs (T, variadic) – Inputs fed to the invoked graph. The i-th input here goes to the i-th input of the invoked graph. To omit an optional input in this field, the user can drop it or use an empty string.

Node Outputs

Outputs (T, variadic) – The outputs generated by the called graph. Its i-th value is bound to the i-th output of the called graph. Similar to the inputs, all outputs are optional.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128

Parameters

name – the name of the node.
graph_name (str) – The invoked graph’s name. The only allowed value is the name of the inference graph, which is stored in “ModelProto.graph.name” in the ONNX model format.

graph_name¶: The invoked graph’s name. The only allowed value is the name of the inference graph, which is stored in “ModelProto.graph.name” in the ONNX model format.

class ONNXGreater(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Returns the tensor resulted from performing the greater logical operation elementwise on the input tensors A and B (with Numpy-style broadcasting support).

This operator supports multidirectional (i.e., Numpy-style) broadcasting; for more details please check [the doc](Broadcasting.md).

Node Inputs

A (T, single) – First input operand for the logical operator.
B (T, single) – Second input operand for the logical operator.

Node Outputs

C (T1, single) – Result tensor.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64
T1 – bool_

Parameters

name – the name of the node.

class ONNXGreaterOrEqual(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Returns the tensor resulted from performing the greater_equal logical operation elementwise on the input tensors A and B (with Numpy-style broadcasting support).

This operator supports multidirectional (i.e., Numpy-style) broadcasting; for more details please check [the doc](Broadcasting.md).

Node Inputs

A (T, single) – First input operand for the logical operator.
B (T, single) – Second input operand for the logical operator.

Node Outputs

C (T1, single) – Result tensor.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64
T1 – bool_

Parameters

name – the name of the node.

class ONNXHardSigmoid(name, *, alpha=0.20000000298023224, beta=0.5)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

HardSigmoid takes one input data (Tensor<T>) and produces one output data (Tensor<T>) where the HardSigmoid function, y = max(0, min(1, alpha * x + beta)), is applied to the tensor elementwise.

Node Inputs

X (T, single) – Input tensor

Node Outputs

Y (T, single) – Output tensor

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.
alpha (Optional [float], default=0.20000000298023224) – Value of alpha.
beta (Optional [float], default=0.5) – Value of beta.

alpha¶: Value of alpha.

beta¶: Value of beta.

class ONNXHardmax(name, *, axis=1)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

The operator computes the hardmax (1 for the first maximum value, and 0 for all others) values for each layer in the batch: of the given input.

The input does not need to explicitly be a 2D vector; rather, it will be coerced into one. For an arbitrary n-dimensional tensor input in [a_0, a_1, …, a_{k-1}, a_k, …, a_{n-1}] and k is the axis provided, then input will be coerced into a 2-dimensional tensor with dimensions [a_0 * … * a_{k-1}, a_k * … * a_{n-1}]. For the default case where axis=1, this means the input tensor will be coerced into a 2D tensor of dimensions [a_0, a_1 * … * a_{n-1}], where a_0 is often the batch size. In this situation, we must have a_0 = N and a_1 * … * a_{n-1} = D. Each of these dimensions must be matched correctly, or else the operator will throw errors. The output tensor has the same shape and contains the hardmax values of the corresponding input.

Node Inputs

input (T, single) – The input tensor that’s coerced into a 2D matrix of size (NxD) as described above.

Node Outputs

output (T, single) – The output values with the same shape as input tensor (the original size without coercion).

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.
axis (Optional [int], default=1) – Describes the axis of the inputs when coerced to 2D; defaults to one because the 0th axis most likely describes the batch_size. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(input).

axis¶: Describes the axis of the inputs when coerced to 2D; defaults to one because the 0th axis most likely describes the batch_size. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(input).

class ONNXIdentity(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Identity operator

Node Inputs

input (T, single) – Input tensor

Node Outputs

output (T, single) – Tensor to copy input into.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128

Parameters

name – the name of the node.

class ONNXImputer(name, *, imputed_value_floats=None, imputed_value_int64s=None, replaced_value_float=0.0, replaced_value_int64=0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Replaces inputs that equal one value with another, leaving all other elements alone. This operator is typically used to replace missing values in situations where they have a canonical representation, such as -1, 0, NaN, or some extreme value. One and only one of imputed_value_floats or imputed_value_int64s should be defined – floats if the input tensor holds floats, integers if the input tensor holds integers. The imputed values must all fit within the width of the tensor element type. One and only one of the replaced_value_float or replaced_value_int64 should be defined, which one depends on whether floats or integers are being processed. The imputed_value attribute length can be 1 element, or it can have one element per input feature. In other words, if the input tensor has the shape [*,F], then the length of the attribute array may be 1 or F. If it is 1, then it is broadcast along the last dimension and applied to each feature.

Node Inputs

X (T, single) – Data to be processed.

Node Outputs

Y (T, single) – Imputed output data

Type Constraints

T – float32, float64, int64, int32

Parameters

name – the name of the node.
imputed_value_floats (Optional [List [float]], default=None) – Value(s) to change to
imputed_value_int64s (Optional [List [int]], default=None) – Value(s) to change to.
replaced_value_float (Optional [float], default=0.0) – A value that needs replacing.
replaced_value_int64 (Optional [int], default=0) – A value that needs replacing.

imputed_value_floats¶: Value(s) to change to

imputed_value_int64s¶: Value(s) to change to.

replaced_value_float¶: A value that needs replacing.

replaced_value_int64¶: A value that needs replacing.

class ONNXInstanceNormalization(name, *, epsilon=9.999999747378752e-06)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Carries out instance normalization as described in the paper https://arxiv.org/abs/1607.08022.

y = scale * (x - mean) / sqrt(variance + epsilon) + B, where mean and variance are computed per instance per channel.

Node Inputs

input (T, single) – Input data tensor from the previous operator; dimensions for image case are (N x C x H x W), where N is the batch size, C is the number of channels, and H and W are the height and the width of the data. For non image case, the dimensions are in the form of (N x C x D1 x D2 … Dn), where N is the batch size.
scale (T, single) – The input 1-dimensional scale tensor of size C.
B (T, single) – The input 1-dimensional bias tensor of size C.

Node Outputs

output (T, single) – The output tensor of the same shape as input.

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.
epsilon (Optional [float], default=9.999999747378752e-06) – The epsilon value to use to avoid division by zero.

epsilon¶: The epsilon value to use to avoid division by zero.

class ONNXIsInf(name, *, detect_negative=1, detect_positive=1)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Map infinity to true and other values to false.

Node Inputs

X (T1, single) – input

Node Outputs

Y (T2, single) – output

Type Constraints

T1 – float32, float64
T2 – bool_

Parameters

name – the name of the node.
detect_negative (Optional [int], default=1) – (Optional) Whether map negative infinity to true. Default to 1 so that negative infinity induces true. Set this attribute to 0 if negative infinity should be mapped to false.
detect_positive (Optional [int], default=1) – (Optional) Whether map positive infinity to true. Default to 1 so that positive infinity induces true. Set this attribute to 0 if positive infinity should be mapped to false.

detect_negative¶: (Optional) Whether map negative infinity to true. Default to 1 so that negative infinity induces true. Set this attribute to 0 if negative infinity should be mapped to false.

detect_positive¶: (Optional) Whether map positive infinity to true. Default to 1 so that positive infinity induces true. Set this attribute to 0 if positive infinity should be mapped to false.

class ONNXIsNaN(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Returns which elements of the input are NaN.

Node Inputs

X (T1, single) – input

Node Outputs

Y (T2, single) – output

Type Constraints

T1 – float16, float32, float64
T2 – bool_

Parameters

name – the name of the node.

class ONNXLRN(name, *, alpha=9.999999747378752e-05, beta=0.75, bias=1.0, size)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Local Response Normalization proposed in the [AlexNet paper](https://papers.nips.cc/paper/4824-imagenet-classification-with-deep-convolutional-neural-networks.pdf). It normalizes over local input regions. The local region is defined across the channels. For an element X[n, c, d1, …, dk] in a tensor of shape (N x C x D1 x D2, …, Dk), its region is {X[n, i, d1, …, dk] | max(0, c - floor((size - 1) / 2)) <= i <= min(C - 1, c + ceil((size - 1) / 2))}.

square_sum[n, c, d1, …, dk] = sum(X[n, i, d1, …, dk] ^ 2), where max(0, c - floor((size - 1) / 2)) <= i <= min(C - 1, c + ceil((size - 1) / 2)).

Y[n, c, d1, …, dk] = X[n, c, d1, …, dk] / (bias + alpha / size * square_sum[n, c, d1, …, dk] ) ^ beta

Node Inputs

X (T, single) – Input data tensor from the previous operator; dimensions for image case are (N x C x H x W), where N is the batch size, C is the number of channels, and H and W are the height and the width of the data. For non image case, the dimensions are in the form of (N x C x D1 x D2 … Dn), where N is the batch size. Optionally, if dimension denotation is in effect, the operation expects the input data tensor to arrive with the dimension denotation of [DATA_BATCH, DATA_CHANNEL, DATA_FEATURE, DATA_FEATURE …].

Node Outputs

Y (T, single) – Output tensor, which has the shape and type as input tensor

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.
alpha (Optional [float], default=9.999999747378752e-05) – Scaling parameter.
beta (Optional [float], default=0.75) – The exponent.
bias (Optional [float], default=1.0) –
size (int) – The number of channels to sum over

alpha¶: Scaling parameter.

beta¶: The exponent.

bias¶: Object property of type float

size¶: The number of channels to sum over

class ONNXLSTM(name, *, activation_alpha=None, activation_beta=None, activations=None, clip=None, direction='forward', hidden_size=None, input_forget=0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Computes an one-layer LSTM. This operator is usually supported via some custom implementation such as CuDNN.

Notations:

X - input tensor

i - input gate

o - output gate

f - forget gate

c - cell gate

t - time step (t-1 means previous time step)

W[iofc] - W parameter weight matrix for input, output, forget, and cell gates

R[iofc] - R recurrence weight matrix for input, output, forget, and cell gates

Wb[iofc] - W bias vectors for input, output, forget, and cell gates

Rb[iofc] - R bias vectors for input, output, forget, and cell gates

P[iof] - P peephole weight vector for input, output, and forget gates

WB[iofc] - W parameter weight matrix for backward input, output, forget, and cell gates

RB[iofc] - R recurrence weight matrix for backward input, output, forget, and cell gates

WBb[iofc] - W bias vectors for backward input, output, forget, and cell gates

RBb[iofc] - R bias vectors for backward input, output, forget, and cell gates

PB[iof] - P peephole weight vector for backward input, output, and forget gates

H - Hidden state

num_directions - 2 if direction == bidirectional else 1

Activation functions:

Relu(x) - max(0, x)

Tanh(x) - (1 - e^{-2x})/(1 + e^{-2x})

Sigmoid(x) - 1/(1 + e^{-x})

(NOTE: Below are optional)

Affine(x) - alpha*x + beta

LeakyRelu(x) - x if x >= 0 else alpha * x

ThresholdedRelu(x) - x if x >= alpha else 0

ScaledTanh(x) - alpha*Tanh(beta*x)

HardSigmoid(x) - min(max(alpha*x + beta, 0), 1)

Elu(x) - x if x >= 0 else alpha*(e^x - 1)

Softsign(x) - x/(1 + |x|)

Softplus(x) - log(1 + e^x)

Equations (Default: f=Sigmoid, g=Tanh, h=Tanh):

it = f(Xt*(Wi^T) + Ht-1*(Ri^T) + Pi (.) Ct-1 + Wbi + Rbi)

ft = f(Xt*(Wf^T) + Ht-1*(Rf^T) + Pf (.) Ct-1 + Wbf + Rbf)

ct = g(Xt*(Wc^T) + Ht-1*(Rc^T) + Wbc + Rbc)

Ct = ft (.) Ct-1 + it (.) ct

ot = f(Xt*(Wo^T) + Ht-1*(Ro^T) + Po (.) Ct + Wbo + Rbo)

Ht = ot (.) h(Ct)

This operator has optional inputs/outputs. See [the doc](IR.md) for more details about the representation of optional arguments. An empty string may be used in the place of an actual argument’s name to indicate a missing argument. Trailing optional arguments (those not followed by an argument that is present) may also be simply omitted.

Node Inputs

X (T, single) – The input sequences packed (and potentially padded) into one 3-D tensor with the shape of [seq_length, batch_size, input_size].
W (T, single) – The weight tensor for the gates. Concatenation of W[iofc] and WB[iofc] (if bidirectional) along dimension 0. The tensor has shape [num_directions, 4*hidden_size, input_size].
R (T, single) – The recurrence weight tensor. Concatenation of R[iofc] and RB[iofc] (if bidirectional) along dimension 0. This tensor has shape [num_directions, 4*hidden_size, hidden_size].
B (T, optional) – The bias tensor for input gate. Concatenation of [Wb[iofc], Rb[iofc]], and [WBb[iofc], RBb[iofc]] (if bidirectional) along dimension 0. This tensor has shape [num_directions, 8*hidden_size]. Optional: If not specified - assumed to be 0.
sequence_lens (T1, optional) – Optional tensor specifying lengths of the sequences in a batch. If not specified - assumed all sequences in the batch to have length seq_length. It has shape [batch_size].
initial_h (T, optional) – Optional initial value of the hidden. If not specified - assumed to be 0. It has shape [num_directions, batch_size, hidden_size].
initial_c (T, optional) – Optional initial value of the cell. If not specified - assumed to be 0. It has shape [num_directions, batch_size, hidden_size].
P (T, optional) – The weight tensor for peepholes. Concatenation of P[iof] and PB[iof] (if bidirectional) along dimension 0. It has shape [num_directions, 3*hidde_size]. Optional: If not specified - assumed to be 0.

Node Outputs

Y (T, optional) – A tensor that concats all the intermediate output values of the hidden. It has shape [seq_length, num_directions, batch_size, hidden_size].
Y_h (T, optional) – The last output value of the hidden. It has shape [num_directions, batch_size, hidden_size].
Y_c (T, optional) – The last output value of the cell. It has shape [num_directions, batch_size, hidden_size].

Type Constraints

T – float16, float32, float64
T1 – int32

Parameters

name – the name of the node.
activation_alpha (Optional [List [float]], default=None) – Optional scaling values used by some activation functions. The values are consumed in the order of activation functions, for example (f, g, h) in LSTM. Default values are the same as of corresponding ONNX operators.For example with LeakyRelu, the default alpha is 0.01.
activation_beta (Optional [List [float]], default=None) – Optional scaling values used by some activation functions. The values are consumed in the order of activation functions, for example (f, g, h) in LSTM. Default values are the same as of corresponding ONNX operators.
activations (Optional [List [str]], default=None) – A list of 3 (or 6 if bidirectional) activation functions for input, output, forget, cell, and hidden. The activation functions must be one of the activation functions specified above. Optional: See the equations for default if not specified.
clip (Optional [float], default=None) – Cell clip threshold. Clipping bounds the elements of a tensor in the range of [-threshold, +threshold] and is applied to the input of activations. No clip if not specified.
direction (Optional [str], default=’forward’) – Specify if the RNN is forward, reverse, or bidirectional. Must be one of forward (default), reverse, or bidirectional.
hidden_size (Optional [int], default=None) – Number of neurons in the hidden layer
input_forget (Optional [int], default=0) – Couple the input and forget gates if 1.

activation_alpha¶: Optional scaling values used by some activation functions. The values are consumed in the order of activation functions, for example (f, g, h) in LSTM. Default values are the same as of corresponding ONNX operators.For example with LeakyRelu, the default alpha is 0.01.

activation_beta¶: Optional scaling values used by some activation functions. The values are consumed in the order of activation functions, for example (f, g, h) in LSTM. Default values are the same as of corresponding ONNX operators.

activations¶: A list of 3 (or 6 if bidirectional) activation functions for input, output, forget, cell, and hidden. The activation functions must be one of the activation functions specified above. Optional: See the equations for default if not specified.

clip¶: Cell clip threshold. Clipping bounds the elements of a tensor in the range of [-threshold, +threshold] and is applied to the input of activations. No clip if not specified.

direction¶: Specify if the RNN is forward, reverse, or bidirectional. Must be one of forward (default), reverse, or bidirectional.

hidden_size¶: Number of neurons in the hidden layer

input_forget¶: Couple the input and forget gates if 1.

class ONNXLabelEncoder(name, *, default_float=- 0.0, default_int64=- 1, default_string='_Unused', keys_floats=None, keys_int64s=None, keys_strings=None, values_floats=None, values_int64s=None, values_strings=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Maps each element in the input tensor to another value. The mapping is determined by the two parallel attributes, ‘keys_*’ and ‘values_*’ attribute. The i-th value in the specified ‘keys_*’ attribute would be mapped to the i-th value in the specified ‘values_*’ attribute. It implies that input’s element type and the element type of the specified ‘keys_*’ should be identical while the output type is identical to the specified ‘values_*’ attribute. If an input element can not be found in the specified ‘keys_*’ attribute, the ‘default_*’ that matches the specified ‘values_*’ attribute may be used as its output value. Let’s consider an example which maps a string tensor to an integer tensor. Assume and ‘keys_strings’ is [“Amy”, “Sally”], ‘values_int64s’ is [5, 6], and ‘default_int64’ is ‘-1’. The input [“Dori”, “Amy”, “Amy”, “Sally”, “Sally”] would be mapped to [-1, 5, 5, 6, 6]. Since this operator is an one-to-one mapping, its input and output shapes are the same. Notice that only one of ‘keys_*’/’values_*’ can be set. For key look-up, bit-wise comparison is used so even a float NaN can be mapped to a value in ‘values_*’ attribute.

Node Inputs

X (T1, single) – Input data. It can be either tensor or scalar.

Node Outputs

Y (T2, single) – Output data.

Type Constraints

T1 – int64, float32
T2 – int64, float32

Parameters

name – the name of the node.
default_float (Optional [float], default=-0.0) – A float.
default_int64 (Optional [int], default=-1) – An integer.
default_string (Optional [str], default=’_Unused’) – A string.
keys_floats (Optional [List [float]], default=None) – A list of floats.
keys_int64s (Optional [List [int]], default=None) – A list of ints.
keys_strings (Optional [List [str]], default=None) – A list of strings. One and only one of ‘keys_*’s should be set.
values_floats (Optional [List [float]], default=None) – A list of floats.
values_int64s (Optional [List [int]], default=None) – A list of ints.
values_strings (Optional [List [str]], default=None) – A list of strings. One and only one of ‘value_*’s should be set.

default_float¶: A float.

default_int64¶: An integer.

default_string¶: A string.

keys_floats¶: A list of floats.

keys_int64s¶: A list of ints.

keys_strings¶: A list of strings. One and only one of ‘keys_*’s should be set.

values_floats¶: A list of floats.

values_int64s¶: A list of ints.

values_strings¶: A list of strings. One and only one of ‘value_*’s should be set.

class ONNXLeakyRelu(name, *, alpha=0.009999999776482582)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

LeakyRelu takes input data (Tensor<T>) and an argument alpha, and produces one output data (Tensor<T>) where the function f(x) = alpha * x for x < 0, f(x) = x for x >= 0, is applied to the data tensor elementwise.

Node Inputs

X (T, single) – Input tensor

Node Outputs

Y (T, single) – Output tensor

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.
alpha (Optional [float], default=0.009999999776482582) – Coefficient of leakage.

alpha¶: Coefficient of leakage.

class ONNXLess(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Returns the tensor resulted from performing the less logical operation elementwise on the input tensors A and B (with Numpy-style broadcasting support).

This operator supports multidirectional (i.e., Numpy-style) broadcasting; for more details please check [the doc](Broadcasting.md).

Node Inputs

A (T, single) – First input operand for the logical operator.
B (T, single) – Second input operand for the logical operator.

Node Outputs

C (T1, single) – Result tensor.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64
T1 – bool_

Parameters

name – the name of the node.

class ONNXLessOrEqual(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Returns the tensor resulted from performing the less_equal logical operation elementwise on the input tensors A and B (with Numpy-style broadcasting support).

This operator supports multidirectional (i.e., Numpy-style) broadcasting; for more details please check [the doc](Broadcasting.md).

Node Inputs

A (T, single) – First input operand for the logical operator.
B (T, single) – Second input operand for the logical operator.

Node Outputs

C (T1, single) – Result tensor.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64
T1 – bool_

Parameters

name – the name of the node.

class ONNXLinearClassifier(name, *, classlabels_ints=None, classlabels_strings=None, coefficients, intercepts=None, multi_class=0, post_transform='NONE')¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Linear classifier

Node Inputs

X (T1, single) – Data to be classified.

Node Outputs

Y (T2, single) – Classification outputs (one class per example).
Z (Z_constraint, single) – Classification scores ([N,E] - one score for each class and example

Type Constraints

T1 – float32, float64, int64, int32
T2 – int64
Z_constraint – float32

Parameters

name – the name of the node.
classlabels_ints (Optional [List [int]], default=None) – Class labels when using integer labels. One and only one ‘classlabels’ attribute must be defined.
classlabels_strings (Optional [List [str]], default=None) – Class labels when using string labels. One and only one ‘classlabels’ attribute must be defined.
coefficients (List [float]) – A collection of weights of the model(s).
intercepts (Optional [List [float]], default=None) – A collection of intercepts.
multi_class (Optional [int], default=0) – Indicates whether to do OvR or multinomial (0=OvR is the default).
post_transform (Optional [str], default=’NONE’) – Indicates the transform to apply to the scores vector. One of ‘NONE,’ ‘SOFTMAX,’ ‘LOGISTIC,’ ‘SOFTMAX_ZERO,’ or ‘PROBIT’

classlabels_ints¶: Class labels when using integer labels. One and only one ‘classlabels’ attribute must be defined.

classlabels_strings¶: Class labels when using string labels. One and only one ‘classlabels’ attribute must be defined.

coefficients¶: A collection of weights of the model(s).

intercepts¶: A collection of intercepts.

multi_class¶: Indicates whether to do OvR or multinomial (0=OvR is the default).

post_transform¶: Indicates the transform to apply to the scores vector. One of ‘NONE,’ ‘SOFTMAX,’ ‘LOGISTIC,’ ‘SOFTMAX_ZERO,’ or ‘PROBIT’

class ONNXLinearRegressor(name, *, coefficients=None, intercepts=None, post_transform='NONE', targets=1)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Generalized linear regression evaluation. If targets is set to 1 (default) then univariate regression is performed. If targets is set to M then M sets of coefficients must be passed in as a sequence and M results will be output for each input n in N. The coefficients array is of length n, and the coefficients for each target are contiguous. Intercepts are optional but if provided must match the number of targets.

Node Inputs

X (T, single) – Data to be regressed.

Node Outputs

Y (Y_constraint, single) – Regression outputs (one per target, per example).

Type Constraints

T – float32, float64, int64, int32
Y_constraint – float32

Parameters

name – the name of the node.
coefficients (Optional [List [float]], default=None) – Weights of the model(s).
intercepts (Optional [List [float]], default=None) – Weights of the intercepts, if used.
post_transform (Optional [str], default=’NONE’) – Indicates the transform to apply to the regression output vector. One of ‘NONE,’ ‘SOFTMAX,’ ‘LOGISTIC,’ ‘SOFTMAX_ZERO,’ or ‘PROBIT’
targets (Optional [int], default=1) – The total number of regression targets, 1 if not defined.

coefficients¶: Weights of the model(s).

intercepts¶: Weights of the intercepts, if used.

post_transform¶: Indicates the transform to apply to the regression output vector. One of ‘NONE,’ ‘SOFTMAX,’ ‘LOGISTIC,’ ‘SOFTMAX_ZERO,’ or ‘PROBIT’

targets¶: The total number of regression targets, 1 if not defined.

class ONNXLog(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Calculates the natural log of the given input tensor, element-wise.

Node Inputs

input (T, single) – Input tensor

Node Outputs

output (T, single) – The natural log of the input tensor computed element-wise

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXLogSoftmax(name, *, axis=1)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

The operator computes the logsoftmax (log of softmax) values for each layer in the batch: of the given input.

The input does not need to explicitly be a 2D vector; rather, it will be coerced into one. For an arbitrary n-dimensional tensor input in [a_0, a_1, …, a_{k-1}, a_k, …, a_{n-1}] and k is the axis provided, then input will be coerced into a 2-dimensional tensor with dimensions [a_0 * … * a_{k-1}, a_k * … * a_{n-1}]. For the default case where axis=1, this means the input tensor will be coerced into a 2D tensor of dimensions [a_0, a_1 * … * a_{n-1}], where a_0 is often the batch size. In this situation, we must have a_0 = N and a_1 * … * a_{n-1} = D. Each of these dimensions must be matched correctly, or else the operator will throw errors. The output tensor has the same shape and contains the logsoftmax values of the corresponding input.

Node Inputs

input (T, single) – The input tensor that’s coerced into a 2D matrix of size (NxD) as described above.

Node Outputs

output (T, single) – The output values with the same shape as input tensor (the original size without coercion).

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.
axis (Optional [int], default=1) – Describes the axis of the inputs when coerced to 2D; defaults to one because the 0th axis most likely describes the batch_size. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(input).

axis¶: Describes the axis of the inputs when coerced to 2D; defaults to one because the 0th axis most likely describes the batch_size. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(input).

class ONNXLpNormalization(name, *, axis=- 1, p=2)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Given a matrix, apply Lp-normalization along the provided axis.

Node Inputs

input (T, single) – Input matrix

Node Outputs

output (T, single) – Matrix after normalization

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.
axis (Optional [int], default=-1) – The axis on which to apply normalization, -1 mean last axis.
p (Optional [int], default=2) – The order of the normalization, only 1 or 2 are supported.

axis¶: The axis on which to apply normalization, -1 mean last axis.

p¶: The order of the normalization, only 1 or 2 are supported.

class ONNXLpPool(name, *, auto_pad='NOTSET', kernel_shape, p=2, pads=None, strides=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

LpPool consumes an input tensor X and applies Lp pooling across the tensor according to kernel sizes, stride sizes, and pad lengths. Lp pooling consisting of computing the Lp norm on all values of a subset of the input tensor according to the kernel size and downsampling the data into the output tensor Y for further processing.

Node Inputs

X (T, single) – Input data tensor from the previous operator; dimensions for image case are (N x C x H x W), where N is the batch size, C is the number of channels, and H and W are the height and the width of the data. For non image case, the dimensions are in the form of (N x C x D1 x D2 … Dn), where N is the batch size.

Node Outputs

Y (T, single) – Output data tensor from Lp pooling across the input tensor. Dimensions will vary based on various kernel, stride, and pad sizes.

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.
auto_pad (Optional [str], default=’NOTSET’) – auto_pad must be either NOTSET, SAME_UPPER, SAME_LOWER or VALID. Where default value is NOTSET, which means explicit padding is used. SAME_UPPER or SAME_LOWER mean pad the input so that the output spatial size match the input.In case of odd number add the extra padding at the end for SAME_UPPER and at the beginning for SAME_LOWER. VALID mean no padding.
kernel_shape (List [int]) – The size of the kernel along each axis.
p (Optional [int], default=2) – p value of the Lp norm used to pool over the input data.
pads (Optional [List [int]], default=None) – Padding for the beginning and ending along each spatial axis, it can take any value greater than or equal to 0. The value represent the number of pixels added to the beginning and end part of the corresponding axis. pads format should be as follow [x1_begin, x2_begin…x1_end, x2_end,…], where xi_begin the number of pixels added at the beginning of axis i and xi_end, the number of pixels added at the end of axis i. This attribute cannot be used simultaneously with auto_pad attribute. If not present, the padding defaults to 0 along start and end of each spatial axis.
strides (Optional [List [int]], default=None) – Stride along each spatial axis. If not present, the stride defaults to 1 along each spatial axis.

auto_pad¶: auto_pad must be either NOTSET, SAME_UPPER, SAME_LOWER or VALID. Where default value is NOTSET, which means explicit padding is used. SAME_UPPER or SAME_LOWER mean pad the input so that the output spatial size match the input.In case of odd number add the extra padding at the end for SAME_UPPER and at the beginning for SAME_LOWER. VALID mean no padding.

kernel_shape¶: The size of the kernel along each axis.

p¶: p value of the Lp norm used to pool over the input data.

pads¶: Padding for the beginning and ending along each spatial axis, it can take any value greater than or equal to 0. The value represent the number of pixels added to the beginning and end part of the corresponding axis. pads format should be as follow [x1_begin, x2_begin…x1_end, x2_end,…], where xi_begin the number of pixels added at the beginning of axis i and xi_end, the number of pixels added at the end of axis i. This attribute cannot be used simultaneously with auto_pad attribute. If not present, the padding defaults to 0 along start and end of each spatial axis.

strides¶: Stride along each spatial axis. If not present, the stride defaults to 1 along each spatial axis.

class ONNXMatMul(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Matrix product that behaves like numpy.matmul: https://docs.scipy.org/doc/numpy-1.13.0/reference/generated/numpy.matmul.html

Node Inputs

A (T, single) – N-dimensional matrix A
B (T, single) – N-dimensional matrix B

Node Outputs

Y (T, single) – Matrix multiply results from A * B

Type Constraints

T – float16, float32, float64, uint32, uint64, int32, int64

Parameters

name – the name of the node.

class ONNXMatMulInteger(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Matrix product that behaves like numpy.matmul: https://docs.scipy.org/doc/numpy-1.13.0/reference/generated/numpy.matmul.html. The production MUST never overflow. The accumulation may overflow if and only if in 32 bits.

Node Inputs

A (T1, single) – N-dimensional matrix A
B (T2, single) – N-dimensional matrix B
a_zero_point (T1, optional) – Zero point tensor for input ‘A’. It’s optional and default value is 0. It could be a scalar or a 1-D tensor, which means a per-tensor or per-row quantization. If it’s a 1-D tensor, its number of elements should be equal to the number of rows of input ‘A’.
b_zero_point (T2, optional) – Zero point tensor for input ‘B’. It’s optional and default value is 0. It could be a scalar or a 1-D tensor, which means a per-tensor or per-column quantization. If it’s a 1-D tensor, its number of elements should be equal to the number of columns of input ‘B’.

Node Outputs

Y (T3, single) – Matrix multiply results from A * B

Type Constraints

T1 – int8, uint8
T2 – int8, uint8
T3 – int32

Parameters

name – the name of the node.

class ONNXMax(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Element-wise max of each of the input tensors (with Numpy-style broadcasting support). All inputs and outputs must have the same data type. This operator supports multidirectional (i.e., Numpy-style) broadcasting; for more details please check [the doc](Broadcasting.md).

Node Inputs

data_0 (T, variadic) – List of tensors for max.

Node Outputs

max (T, single) – Output tensor.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64

Parameters

name – the name of the node.

class ONNXMaxPool(name, *, auto_pad='NOTSET', ceil_mode=0, dilations=None, kernel_shape, pads=None, storage_order=0, strides=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

MaxPool consumes an input tensor X and applies max pooling across the tensor according to kernel sizes, stride sizes, and pad lengths. max pooling consisting of computing the max on all values of a subset of the input tensor according to the kernel size and downsampling the data into the output tensor Y for further processing. The output spatial shape will be following: ` output_spatial_shape[i] = floor((input_spatial_shape[i] + pad_shape[i] - ((kernel_spatial_shape[i] - 1) * dilations[i] + 1)) / strides_spatial_shape[i] + 1) ` or ` output_spatial_shape[i] = ceil((input_spatial_shape[i] + pad_shape[i] - ((kernel_spatial_shape[i] - 1) * dilations[i] + 1)) / strides_spatial_shape[i] + 1) ` if ceil_mode is enabled

` * pad_shape[i] is sum of pads along axis i `

auto_pad is a DEPRECATED attribute. If you are using them currently, the output spatial shape will be following: ` VALID: output_spatial_shape[i] = ceil((input_spatial_shape[i] - ((kernel_spatial_shape[i] - 1) * dilations[i] + 1) + 1) / strides_spatial_shape[i]) SAME_UPPER or SAME_LOWER: output_spatial_shape[i] = ceil(input_spatial_shape[i] / strides_spatial_shape[i]) ` And pad shape will be following if SAME_UPPER or SAME_LOWER: ` pad_shape[i] = (output_spatial_shape[i] - 1) * strides_spatial_shape[i] + ((kernel_spatial_shape[i] - 1) * dilations[i] + 1) - input_spatial_shape[i] ` The output of each pooling window is maximum number of elements exclude pad.

Node Inputs

X (T, single) – Input data tensor from the previous operator; dimensions for image case are (N x C x H x W), where N is the batch size, C is the number of channels, and H and W are the height and the width of the data. For non image case, the dimensions are in the form of (N x C x D1 x D2 … Dn), where N is the batch size. Optionally, if dimension denotation is in effect, the operation expects the input data tensor to arrive with the dimension denotation of [DATA_BATCH, DATA_CHANNEL, DATA_FEATURE, DATA_FEATURE …].

Node Outputs

Y (T, single) – Output data tensor from average or max pooling across the input tensor. Dimensions will vary based on various kernel, stride, and pad sizes. Floor value of the dimension is used
Indices (I, optional) – Indices tensor from max pooling across the input tensor. The dimensions of indices are the same as output tensor. The values in indices of are the indices of the selected values during pooling. The indices are computed as flatten 1-D tensor, and the indices do not consider padding. So the values in indices are in [0, N x C x D1 x … x Dn).

Type Constraints

T – float16, float32, float64, int8, uint8
I – int64

Parameters

name – the name of the node.
auto_pad (Optional [str], default=’NOTSET’) – auto_pad must be either NOTSET, SAME_UPPER, SAME_LOWER or VALID. Where default value is NOTSET, which means explicit padding is used. SAME_UPPER or SAME_LOWER mean pad the input so that the output spatial size match the input.In case of odd number add the extra padding at the end for SAME_UPPER and at the beginning for SAME_LOWER. VALID mean no padding.
ceil_mode (Optional [int], default=0) – Whether to use ceil or floor (default) to compute the output shape.
dilations (Optional [List [int]], default=None) – Dilation value along each spatial axis of filter. If not present, the dilation defaults to 1 along each spatial axis.
kernel_shape (List [int]) – The size of the kernel along each axis.
pads (Optional [List [int]], default=None) – Padding for the beginning and ending along each spatial axis, it can take any value greater than or equal to 0. The value represent the number of pixels added to the beginning and end part of the corresponding axis. pads format should be as follow [x1_begin, x2_begin…x1_end, x2_end,…], where xi_begin the number of pixels added at the beginning of axis i and xi_end, the number of pixels added at the end of axis i. This attribute cannot be used simultaneously with auto_pad attribute. If not present, the padding defaults to 0 along start and end of each spatial axis.
storage_order (Optional [int], default=0) – The storage order of the tensor. 0 is row major, and 1 is column major.
strides (Optional [List [int]], default=None) – Stride along each spatial axis. If not present, the stride defaults to 1 along each spatial axis.

auto_pad¶: auto_pad must be either NOTSET, SAME_UPPER, SAME_LOWER or VALID. Where default value is NOTSET, which means explicit padding is used. SAME_UPPER or SAME_LOWER mean pad the input so that the output spatial size match the input.In case of odd number add the extra padding at the end for SAME_UPPER and at the beginning for SAME_LOWER. VALID mean no padding.

ceil_mode¶: Whether to use ceil or floor (default) to compute the output shape.

dilations¶: Dilation value along each spatial axis of filter. If not present, the dilation defaults to 1 along each spatial axis.

kernel_shape¶: The size of the kernel along each axis.

pads¶: Padding for the beginning and ending along each spatial axis, it can take any value greater than or equal to 0. The value represent the number of pixels added to the beginning and end part of the corresponding axis. pads format should be as follow [x1_begin, x2_begin…x1_end, x2_end,…], where xi_begin the number of pixels added at the beginning of axis i and xi_end, the number of pixels added at the end of axis i. This attribute cannot be used simultaneously with auto_pad attribute. If not present, the padding defaults to 0 along start and end of each spatial axis.

storage_order¶: The storage order of the tensor. 0 is row major, and 1 is column major.

strides¶: Stride along each spatial axis. If not present, the stride defaults to 1 along each spatial axis.

class ONNXMaxRoiPool(name, *, pooled_shape, spatial_scale=1.0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

ROI max pool consumes an input tensor X and region of interests (RoIs) to apply max pooling across each RoI, to produce output 4-D tensor of shape (num_rois, channels, pooled_shape[0], pooled_shape[1]).

Node Inputs

X (T, single) – Input data tensor from the previous operator; dimensions for image case are (N x C x H x W), where N is the batch size, C is the number of channels, and H and W are the height and the width of the data.
rois (T, single) – RoIs (Regions of Interest) to pool over. Should be a 2-D tensor of shape (num_rois, 5) given as [[batch_id, x1, y1, x2, y2], …].

Node Outputs

Y (T, single) – RoI pooled output 4-D tensor of shape (num_rois, channels, pooled_shape[0], pooled_shape[1]).

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.
pooled_shape (List [int]) – ROI pool output shape (height, width).
spatial_scale (Optional [float], default=1.0) – Multiplicative spatial scale factor to translate ROI coordinates from their input scale to the scale used when pooling.

pooled_shape¶: ROI pool output shape (height, width).

spatial_scale¶: Multiplicative spatial scale factor to translate ROI coordinates from their input scale to the scale used when pooling.

class ONNXMaxUnpool(name, *, kernel_shape, pads=None, strides=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

MaxUnpool essentially computes the partial inverse of the MaxPool op.: The input information to this op is typically the the output information from a MaxPool op. The first input tensor X is the tensor that needs to be unpooled, which is typically the pooled tensor (first output) from MaxPool. The second input tensor, I, contains the indices to the (locally maximal) elements corrsponding to the elements in the first input tensor X. Input tensor I is typically the second output of the MaxPool op. The third (optional) input is a tensor that specifies the output size of the unpooling operation.
MaxUnpool is intended to do ‘partial’ inverse of the MaxPool op. ‘Partial’ because all the non-maximal: values from the original input to MaxPool are set to zero in the output of the MaxUnpool op. Pooling the result of an unpooling operation should give back the original input to the unpooling op.
MaxUnpool can produce the same output size for several input sizes, which makes unpooling op ambiguous.: The third input argument, output_size, is meant to disambiguate the op and produce output tensor of known/predictable size.
In addition to the inputs, MaxUnpool takes three attributes, namely kernel_shape, strides, and pads,: which define the exact unpooling op. The attributes typically have the same values as the corrsponding pooling op that the unpooling op is trying to invert.

Node Inputs

X (T1, single) – Input data tensor that has to be unpooled. This tensor is typically the first output of the MaxPool op.Dimensions for image case are (N x C x H x W), where N is the batch size, C is the number of channels, and H and W are the height and the width of the data. For non-image case, the dimensions are in the form of (N x C x D1 x D2 … Dn), where N is the batch size. Optionally, if dimension denotation is in effect, the operation expects the input data tensor to arrive with the dimension denotation of [DATA_BATCH, DATA_CHANNEL, DATA_FEATURE, DATA_FEATURE …].
I (T2, single) – Input data tensor containing the indices corresponding to elements in the first input tensor X.This tensor is typically the second output of the MaxPool op.Dimensions must be the same as input tensor X. The indices are linear, i.e. computed considering the tensor as flattened 1-D tensor, assuming row-major storage. Also, the linear indices should not consider padding. So the values in indices are in the range [0, N x C x D1 x … x Dn).
output_shape (T2, optional) – The shape of the output can be explicitly set which will cause pads values to be auto generated. If ‘output_shape’ is specified, ‘pads’ values are ignored.

Node Outputs

output (T1, single) – Output data tensor that contains the result of the unpooling.

Type Constraints

T1 – float16, float32, float64
T2 – int64

Parameters

name – the name of the node.
kernel_shape (List [int]) – The size of the kernel along each axis.
pads (Optional [List [int]], default=None) – Padding for the beginning and ending along each spatial axis, it can take any value greater than or equal to 0. The value represent the number of pixels added to the beginning and end part of the corresponding axis. pads format should be as follow [x1_begin, x2_begin…x1_end, x2_end,…], where xi_begin the number of pixels added at the beginning of axis i and xi_end, the number of pixels added at the end of axis i. This attribute cannot be used simultaneously with auto_pad attribute. If not present, the padding defaults to 0 along start and end of each spatial axis.
strides (Optional [List [int]], default=None) – Stride along each spatial axis. If not present, the stride defaults to 1 along each spatial axis.

kernel_shape¶: The size of the kernel along each axis.

pads¶: Padding for the beginning and ending along each spatial axis, it can take any value greater than or equal to 0. The value represent the number of pixels added to the beginning and end part of the corresponding axis. pads format should be as follow [x1_begin, x2_begin…x1_end, x2_end,…], where xi_begin the number of pixels added at the beginning of axis i and xi_end, the number of pixels added at the end of axis i. This attribute cannot be used simultaneously with auto_pad attribute. If not present, the padding defaults to 0 along start and end of each spatial axis.

strides¶: Stride along each spatial axis. If not present, the stride defaults to 1 along each spatial axis.

class ONNXMean(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Element-wise mean of each of the input tensors (with Numpy-style broadcasting support). All inputs and outputs must have the same data type. This operator supports multidirectional (i.e., Numpy-style) broadcasting; for more details please check [the doc](Broadcasting.md).

Node Inputs

data_0 (T, variadic) – List of tensors for mean.

Node Outputs

mean (T, single) – Output tensor.

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXMeanVarianceNormalization(name, *, axes=[0, 2, 3])¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

A MeanVarianceNormalization Function: Perform mean variance normalization on the input tensor X using formula: ` (X-EX)/sqrt(E(X-EX)^2) `

Node Inputs

X (T, single) – Input tensor

Node Outputs

Y (T, single) – Output tensor

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.
axes (Optional [List [int]], default=[0, 2, 3]) – A list of integers, along which to reduce. The default is to caculate along axes [0,2,3] for calculating mean and variance along each channel. Two variables with the same C-coordinate are associated with the same mean and variance.

axes¶: A list of integers, along which to reduce. The default is to caculate along axes [0,2,3] for calculating mean and variance along each channel. Two variables with the same C-coordinate are associated with the same mean and variance.

class ONNXMin(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Element-wise min of each of the input tensors (with Numpy-style broadcasting support). All inputs and outputs must have the same data type. This operator supports multidirectional (i.e., Numpy-style) broadcasting; for more details please check [the doc](Broadcasting.md).

Node Inputs

data_0 (T, variadic) – List of tensors for min.

Node Outputs

min (T, single) – Output tensor.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64

Parameters

name – the name of the node.

class ONNXMod(name, *, fmod=0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Performs element-wise binary modulus (with Numpy-style broadcasting support).
The sign of the remainder is the same as that of the Divisor.

Mod operator can also behave like C fmod() or numpy.fmod. In this case, the sign of the remainder however, will be the same as the Dividend (in contrast to integer mod). To force a behavior like numpy.fmod() an ‘fmod’ Attribute is provided. This attribute is set to 0 by default causing the behavior to be like integer mod. Setting this attribute to 1 causes the remainder to be calculated similar to that of numpy.fmod().

If the input type is floating point, then fmod attribute must be set to 1.

In case of dividend being zero, the results will be platform dependent.

This operator supports multidirectional (i.e., Numpy-style) broadcasting; for more details please check [the doc](Broadcasting.md).

Node Inputs

A (T, single) – Dividend tensor
B (T, single) – Divisor tensor

Node Outputs

C (T, single) – Remainder tensor

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64

Parameters

name – the name of the node.
fmod (Optional [int], default=0) – Whether the operator should behave like fmod (default=0 meaning it will do integer mods); Set this to 1 to force fmod treatment

fmod¶: Whether the operator should behave like fmod (default=0 meaning it will do integer mods); Set this to 1 to force fmod treatment

class ONNXMomentum(name, *, alpha, beta, mode, norm_coefficient)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Compute one iteration of stochastic gradient update with momentum. This operator can conduct the optimization of multiple tensor variables.

Let’s define the behavior of this operator. As you can imagine, SG with momentum requires several parameters:

The learning-rate “R”.

The update count “T”. That is, the number of conducted training iterations. It should be zero in the first training iteration.

A L2-norm regularization coefficient “norm_coefficient”.

A decay coefficient of previous accumulated gradient (i.e., momentum) “alpha”.

The scaling coefficient of current gradient “beta”.

An attribute to choose either standard momentum or Nesterov’s momentum “mode” should be used.

For the sake of simplicity, assume that there is only one tensor (called “X”) to be optimized. Other necessary inputs are “X“‘s gradient (called “G”) and “X“‘s momentum (called “V”). This Momentum operator maps all these inputs to the new value of “X” (called “X_new”) and its new momentum (called “V_new”).

This operator supports two different momentum algorithms. Set the attribute “mode” to “nesterov” if Nesterov’s momentum is desired. Otherwise, set the attribute “model” to “standard” to use standard momentum. Computation details are described subsequently.

Let “+”, “-“, “*”, and “/” are all element-wise operations with numpy-style broadcasting.

Pseudo code for SG with standard momentum:

// Add gradient of 0.5 * norm_coefficient * ||X||^2, where ||X|| is the sum of squared // values of all elements in X. G_regularized = norm_coefficient * X + G

// In the first training iteration, beta should always be 1. beta_adjusted = T > 0 ? beta : 1

// Compute the current momentum based on previous momentum and the current gradient. V_new = alpha * V + beta_adjusted * G_regularized

// Update X. X_new = X - R * V_new

Pseudo code for SG with Nesterov’s momentum:

// Add gradient of 0.5 * norm_coefficient * ||X||^2, where ||X|| is the sum of squared // values of all elements in X. G_regularized = norm_coefficient * X + G;

// In the first training iteration, beta should always be 1. beta_adjusted = T > 0 ? beta : 1

// Compute the current momentum based on previous momentum and the current gradient. V_new = alpha * V + beta_adjusted * G_regularized;

// Compute final update direction and then update X. X_new = X - R * (G_regularized + alpha * V_new)

If one assign this operators to optimize multiple inputs, for example, “X_1” and “X_2”. The same pseudo code would be extended to handle all tensors jointly. More specifically, we can view “X” as a concatenation of “X_1” and “X_2” (of course, their gradient and accumulate gradient should be concatenated too) and then our pseudo code becomes applicable.

Node Inputs

R (T1, single) – The learning rate.
T (T2, single) – Update count of “X”. It should be a scalar.
inputs (T3, variadic) – It sequentially contains the current values of optimized tensors, then their gradient tensors, and finally their momentum tensors. For example, if two tensors “X_1” and “X_2” are optimized, The expected input list would be [“X_1”, “X_2”, gradient of “X_1”, gradient of “X_2”, momentum of “X_1”, momentum of “X_2”].

Node Outputs

outputs (T3, variadic) – It sequentially contains the new values of optimized tensors and then the new values of their momentum tensors. For example, if two tensors “X_1” and “X_2” are optimized, the output list would be [new value of “X_1,” new value of “X_2” new momentum of “X_1”, new momentum of “X_2”].

Type Constraints

T1 – float32, float64
T2 – int64
T3 – float32, float64

Parameters

name – the name of the node.
alpha (float) – The decay factor of momentum. It should be a scalar.
beta (float) – The coefficient of gradient in computing new momentum. It should be a scalar.
mode (str) – Its value should be either “nesterov” or “standard”. The value “nesterov” leads to the use of Nesterov’s momentum while “standard” invokes stochastic gradient method using standard momentum
norm_coefficient (float) – Coefficient of 0.5 * norm_coefficient * ||X||^2.

alpha¶: The decay factor of momentum. It should be a scalar.

beta¶: The coefficient of gradient in computing new momentum. It should be a scalar.

mode¶: Its value should be either “nesterov” or “standard”. The value “nesterov” leads to the use of Nesterov’s momentum while “standard” invokes stochastic gradient method using standard momentum

norm_coefficient¶: Coefficient of 0.5 * norm_coefficient * ||X||^2.

class ONNXMul(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Performs element-wise binary multiplication (with Numpy-style broadcasting support).

This operator supports multidirectional (i.e., Numpy-style) broadcasting; for more details please check [the doc](Broadcasting.md).

Node Inputs

A (T, single) – First operand.
B (T, single) – Second operand.

Node Outputs

C (T, single) – Result, has same element type as two inputs

Type Constraints

T – uint32, uint64, int32, int64, float16, float32, float64

Parameters

name – the name of the node.

class ONNXMultinomial(name, *, dtype=6, sample_size=1, seed=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Generate a tensor of samples from a multinomial distribution according to the probabilities of each of the possible outcomes.

Node Inputs

input (T1, single) – Input tensor with shape [batch_size, class_size], where class_size is the number of all possible outcomes. Each value along the axis zero represents the unnormalized log-probability of each corresponding outcome in a batch.

Node Outputs

output (T2, single) – Output tensor with shape [batch_size, sample_size], where sample_size is the number of times to sample. Each value along the axis zero represents the outcome of the corresponding sample in a batch.

Type Constraints

T1 – float16, float32, float64
T2 – int32, int64

Parameters

name – the name of the node.
dtype (Optional [int], default=6) – (Optional) The data type for the elements of the output tensor, if not specified, we will use int32.
sample_size (Optional [int], default=1) – Number of times to sample.
seed (Optional [float], default=None) – (Optional) Seed to the random generator, if not specified we will auto generate one.

dtype¶: (Optional) The data type for the elements of the output tensor, if not specified, we will use int32.

sample_size¶: Number of times to sample.

seed¶: (Optional) Seed to the random generator, if not specified we will auto generate one.

class ONNXNeg(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Neg takes one input data (Tensor<T>) and produces one output data (Tensor<T>) where each element flipped sign, y = -x, is applied to the tensor elementwise.

Node Inputs

X (T, single) – Input tensor

Node Outputs

Y (T, single) – Output tensor

Type Constraints

T – float32, int32, int8, int16, int64, float16, float64

Parameters

name – the name of the node.

class ONNXNegativeLogLikelihoodLoss(name, *, ignore_index=None, reduction='mean')¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

A NegativeLogLikelihoodLoss operator computes (weighted) negative log likelihood loss. Its “input” tensor has the shape of (N, C, d1, d2, …, dk) where k >= 0. The “input” tensor contains log-probabilities for input[n, :, d_1, d_2,…, d_k] being in a class of [0, C). The operator’s “target” input tensor has the shape of (N, d1, d2, …, dk). It encodes class labels (one of C classes) or it may contain a special value (indicated by an attribute ignore_index) for N x d1 x d2 x … x dk samples. The loss value for input[n, :, d_1, d_2,…d_k] being classified as class c = target[n][d_1][d_2]…[d_k] is computed as:

loss[n][d_1][d_2]…[d_k] = -input[n][c][d_1][d_2]…[d_k].

When an optional “weight” is provided, the sample loss is calculated as:

loss[n][d_1][d_2]…[d_k] = -input[n][c][d_1][d_2]…[d_k] * weight[c].

loss is zero for the case when target-value equals ignore_index.

loss[n][d_1][d_2]…[d_k] = 0, when target[n][d_1][d_2]…[d_k] = ignore_index

If “reduction” attribute is set to “none”, the operator’s output will be the above loss with shape (N, d1, d2, …, dk). If “reduction” attribute is set to “mean” (the default attribute value), the output loss is (weight) averaged:

mean(loss), if “weight” is not provided,

or if weight is provided,

sum(loss) / sum(weight[target[n][d_1][d_2]…[d_k]]]), for all samples.

If “reduction” attribute is set to “sum”, the output is a scalar:: sum(loss).

Example 1:

// negative log likelihood loss, “none” reduction N, C, d1 = 2, 3, 2 input = [[[1.0, 2.0], [2.0, 2.0], [3.0, 2.0]],

[[0.0, 1.0], [2.0, 2.0], [1.0, 2]]]

target = [[2, 1], [0, 2]]

loss = np.zeros((N, d1)) for n in range(N):

for d_1 in range(d1):
c = target[n][d_1] loss[n][d_1] = -input[n][c][d_1]

// print(loss) // [[-3. -2.] // [-0. -2.]]

Example 2:

// weighted negative log likelihood loss, sum reduction N, C, d1 = 2, 3, 2 input = [[[1.0, 2.0], [2.0, 2.0], [3.0, 2.0]],

[[0.0, 1.0], [2.0, 2.0], [1.0, 2]]]

target = [[2, 1], [0, 2]] weight = [0.2, 0.3, 0.1] loss = np.zeros((N, d1)) for n in range(N):

for d_1 in range(d1):
c = target[n][d_1] loss[n][d_1] = -input[n][c][d_1] * weight[c]

loss = np.sum(loss) // print(loss) // -1.1

Example 3:

// weighted negative log likelihood loss, mean reduction N, C, d1 = 2, 3, 2 input = [[[1.0, 2.0], [2.0, 2.0], [3.0, 2.0]],

[[0.0, 1.0], [2.0, 2.0], [1.0, 2]]]

target = [[2, 1], [0, 2]] weight = [0.2, 0.3, 0.1] loss = np.zeros((N, d1)) weight_total = 0 for n in range(N):

for d_1 in range(d1):
c = target[n][d_1] loss[n][d_1] = -input[n][c][d_1] * weight[c] weight_total = weight_total + weight[c]

loss = np.sum(loss) / weight_total // print(loss) // -1.57

Node Inputs

input (T, single) – Input tensor of shape (N, C) or (N, C, d1, d2, …, dk).
target (Tind, single) – Target tensor of shape (N) or (N, d1, d2, …, dk). Target element value shall be in range of [0, C). If ignore_index is specified, it may have a value outside [0, C) and the target values should either be in the range [0, C) or have the value ignore_index.
weight (T, optional) – Optional rescaling weight tensor. If given, it has to be a tensor of size C. Otherwise, it is treated as if having all ones.

Node Outputs

loss (T, single) – The negative log likelihood loss

Type Constraints

T – float16, float32, float64
Tind – int32, int64

Parameters

name – the name of the node.
ignore_index (Optional [int], default=None) – Specifies a target value that is ignored and does not contribute to the input gradient. It’s an optional value.
reduction (Optional [str], default=’mean’) – Type of reduction to apply to loss: none, sum, mean (default). ‘none’: the output is the loss for each sample. ‘sum’: the output will be summed. ‘mean’: the sum of the output will be divided by the sum of applied weights.

ignore_index¶: Specifies a target value that is ignored and does not contribute to the input gradient. It’s an optional value.

reduction¶: Type of reduction to apply to loss: none, sum, mean (default). ‘none’: the output is the loss for each sample. ‘sum’: the output will be summed. ‘mean’: the sum of the output will be divided by the sum of applied weights.

class ONNXNonMaxSuppression(name, *, center_point_box=0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Filter out boxes that have high intersection-over-union (IOU) overlap with previously selected boxes. Bounding boxes with score less than score_threshold are removed. Bounding box format is indicated by attribute center_point_box. Note that this algorithm is agnostic to where the origin is in the coordinate system and more generally is invariant to orthogonal transformations and translations of the coordinate system; thus translating or reflections of the coordinate system result in the same boxes being selected by the algorithm. The selected_indices output is a set of integers indexing into the input collection of bounding boxes representing the selected boxes. The bounding box coordinates corresponding to the selected indices can then be obtained using the Gather or GatherND operation.

Node Inputs

boxes (boxes_constraint, single) – An input tensor with shape [num_batches, spatial_dimension, 4]. The single box data format is indicated by center_point_box.
scores (scores_constraint, single) – An input tensor with shape [num_batches, num_classes, spatial_dimension]
max_output_boxes_per_class (max_output_boxes_per_class_constraint, optional) – Integer representing the maximum number of boxes to be selected per batch per class. It is a scalar. Default to 0, which means no output.
iou_threshold (iou_threshold_constraint, optional) – Float representing the threshold for deciding whether boxes overlap too much with respect to IOU. It is scalar. Value range [0, 1]. Default to 0.
score_threshold (score_threshold_constraint, optional) – Float representing the threshold for deciding when to remove boxes based on score. It is a scalar.

Node Outputs

selected_indices (selected_indices_constraint, single) – selected indices from the boxes tensor. [num_selected_indices, 3], the selected index format is [batch_index, class_index, box_index].

Type Constraints

boxes_constraint – float32
scores_constraint – float32
max_output_boxes_per_class_constraint – int64
iou_threshold_constraint – float32
score_threshold_constraint – float32
selected_indices_constraint – int64

Parameters

name – the name of the node.
center_point_box (Optional [int], default=0) – Integer indicate the format of the box data. The default is 0. 0 - the box data is supplied as [y1, x1, y2, x2] where (y1, x1) and (y2, x2) are the coordinates of any diagonal pair of box corners and the coordinates can be provided as normalized (i.e., lying in the interval [0, 1]) or absolute. Mostly used for TF models. 1 - the box data is supplied as [x_center, y_center, width, height]. Mostly used for Pytorch models.

center_point_box¶: Integer indicate the format of the box data. The default is 0. 0 - the box data is supplied as [y1, x1, y2, x2] where (y1, x1) and (y2, x2) are the coordinates of any diagonal pair of box corners and the coordinates can be provided as normalized (i.e., lying in the interval [0, 1]) or absolute. Mostly used for TF models. 1 - the box data is supplied as [x_center, y_center, width, height]. Mostly used for Pytorch models.

class ONNXNonZero(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Returns the indices of the elements that are non-zero (in row-major order - by dimension). NonZero behaves similar to numpy.nonzero: https://docs.scipy.org/doc/numpy/reference/generated/numpy.nonzero.html

Node Inputs

X (T, single) – input

Node Outputs

Y (Y_constraint, single) – output

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128
Y_constraint – int64

Parameters

name – the name of the node.

class ONNXNormalizer(name, *, norm='MAX')¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Normalize the input. There are three normalization modes, which have the corresponding formulas, defined using element-wise infix operators ‘/’ and ‘^’ and tensor-wide functions ‘max’ and ‘sum’:

 : Max: Y = X / max(X) L1: Y = X / sum(X) L2: Y = sqrt(X^2 / sum(X^2)} In all modes, if the divisor is zero, Y == X.
 : For batches, that is, [N,C] tensors, normalization is done along the C axis. In other words, each row of the batch is normalized independently.

Node Inputs

X (T, single) – Data to be encoded, a tensor of shape [N,C] or [C]

Node Outputs

Y (Y_constraint, single) – Encoded output data

Type Constraints

T – float32, float64, int64, int32
Y_constraint – float32

Parameters

name – the name of the node.
norm (Optional [str], default=’MAX’) – One of ‘MAX,’ ‘L1,’ ‘L2’

norm¶: One of ‘MAX,’ ‘L1,’ ‘L2’

class ONNXNot(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Returns the negation of the input tensor element-wise.

Node Inputs

X (T, single) – Input tensor

Node Outputs

Y (T, single) – Output tensor

Type Constraints

T – bool_

Parameters

name – the name of the node.

class ONNXOneHot(name, *, axis=- 1)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Produces a one-hot tensor based on inputs. The locations represented by the index values in the ‘indices’ input tensor will have ‘on_value’ and the other locations will have ‘off_value’ in the output tensor, where ‘on_value’ and ‘off_value’ are specified as part of required input argument ‘values’, which is a two-element tensor of format [off_value, on_value]. The rank of the output tensor will be one greater than the rank of the input tensor. The additional dimension is for one-hot representation. The additional dimension will be inserted at the position specified by ‘axis’. If ‘axis’ is not specified then then additional dimension will be inserted as the innermost dimension, i.e. axis=-1. The size of the additional dimension is specified by required scalar input ‘depth’. The type of the output tensor is the same as the type of the ‘values’ input. Any entries in the ‘indices’ input tensor with values outside the range [-depth, depth-1] will result in one-hot representation with all ‘off_value’ values in the output tensor.

when axis = 0: output[input[i, j, k], i, j, k] = 1 for all i, j, k and 0 otherwise.

when axis = -1: output[i, j, k, input[i, j, k]] = 1 for all i, j, k and 0 otherwise.

Node Inputs

indices (T1, single) – Input tensor containing indices. Any entries in the ‘indices’ input tensor with values outside the range [-depth, depth-1] will result in one-hot representation with all ‘off_value’ values in the output tensor.In case ‘indices’ is of non-integer type, the values will be casted to int64 before use.
depth (T2, single) – Scalar specifying the number of classes in one-hot tensor. This is also the size of the one-hot dimension (specified by ‘axis’ attribute) added on in the output tensor. The values in the ‘indices’ input tensor are expected to be in the range [-depth, depth-1]. In case ‘depth’ is of non-integer type, it will be casted to int64 before use.
values (T3, single) – Rank 1 tensor containing exactly two elements, in the format [off_value, on_value], where ‘on_value’ is the value used for filling locations specified in ‘indices’ input tensor, and ‘off_value’ is the value used for filling locations other than those specified in ‘indices’ input tensor.

Node Outputs

output (T3, single) – Tensor of rank one greater than input tensor ‘indices’, i.e. rank(output) = rank(indices) + 1. The data type for the elements of the output tensor is the same as the type of input ‘values’ is used.

Type Constraints

T1 – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64
T2 – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64
T3 – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128

Parameters

name – the name of the node.
axis (Optional [int], default=-1) – (Optional) Axis along which one-hot representation in added. Default: axis=-1. axis=-1 means that the additional dimension will be inserted as the innermost/last dimension in the output tensor. Negative value means counting dimensions from the back. Accepted range is [-r-1, r] where r = rank(indices).

axis¶: (Optional) Axis along which one-hot representation in added. Default: axis=-1. axis=-1 means that the additional dimension will be inserted as the innermost/last dimension in the output tensor. Negative value means counting dimensions from the back. Accepted range is [-r-1, r] where r = rank(indices).

class ONNXOneHotEncoder(name, *, cats_int64s=None, cats_strings=None, zeros=1)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Replace each input element with an array of ones and zeros, where a single one is placed at the index of the category that was passed in. The total category count will determine the size of the extra dimension of the output array Y. For example, if we pass a tensor with a single value of 4, and a category count of 8, the output will be a tensor with [0,0,0,0,1,0,0,0]. This operator assumes every input feature is from the same set of categories. If the input is a tensor of float, int32, or double, the data will be cast to integers and the cats_int64s category list will be used for the lookups.

Node Inputs

X (T, single) – Data to be encoded.

Node Outputs

Y (Y_constraint, single) – Encoded output data, having one more dimension than X.

Type Constraints

T – int64, int32, float32, float64
Y_constraint – float32

Parameters

name – the name of the node.
cats_int64s (Optional [List [int]], default=None) – List of categories, ints. One and only one of the ‘cats_*’ attributes must be defined.
cats_strings (Optional [List [str]], default=None) – List of categories, strings. One and only one of the ‘cats_*’ attributes must be defined.
zeros (Optional [int], default=1) – If true and category is not present, will return all zeros; if false and a category if not found, the operator will fail.

cats_int64s¶: List of categories, ints. One and only one of the ‘cats_*’ attributes must be defined.

cats_strings¶: List of categories, strings. One and only one of the ‘cats_*’ attributes must be defined.

zeros¶: If true and category is not present, will return all zeros; if false and a category if not found, the operator will fail.

class ONNXOr(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Returns the tensor resulted from performing the or logical operation elementwise on the input tensors A and B (with Numpy-style broadcasting support).

This operator supports multidirectional (i.e., Numpy-style) broadcasting; for more details please check [the doc](Broadcasting.md).

Node Inputs

A (T, single) – First input operand for the logical operator.
B (T, single) – Second input operand for the logical operator.

Node Outputs

C (T1, single) – Result tensor.

Type Constraints

T – bool_
T1 – bool_

Parameters

name – the name of the node.

class ONNXPRelu(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

PRelu takes input data (Tensor<T>) and slope tensor as input, and produces one output data (Tensor<T>) where the function f(x) = slope * x for x < 0, f(x) = x for x >= 0., is applied to the data tensor elementwise. This operator supports unidirectional broadcasting (tensor slope should be unidirectional broadcastable to input tensor X); for more details please check [the doc](Broadcasting.md).

Node Inputs

X (T, single) – Input tensor
slope (T, single) – Slope tensor. The shape of slope can be smaller then first input X; if so, its shape must be unidirectional broadcastable to X

Node Outputs

Y (T, single) – Output tensor (same size as X)

Type Constraints

T – float16, float32, float64, uint32, uint64, int32, int64

Parameters

name – the name of the node.

class ONNXPad(name, *, mode='constant')¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Given a tensor containing the data to be padded (data), a tensor containing the number of start and end pad values for axis (pads), (optionally) a mode, and (optionally) constant_value, a padded tensor (output) is generated.

The three supported modes are (similar to corresponding modes supported by numpy.pad):

constant`(default) - pads with a given constant value as specified by `constant_value (which defaults to 0)
reflect - pads with the reflection of the vector mirrored on the first and last values of the vector along each axis
edge - pads with the edge values of array

Example 1 (constant mode):

Insert 0 pads to the beginning of the second dimension.

data = [

[1.0, 1.2], [2.3, 3.4], [4.5, 5.7],

]

pads = [0, 2, 0, 0]

mode = ‘constant’

constant_value = 0.0

output = [

[
[0.0, 0.0, 1.0, 1.2], [0.0, 0.0, 2.3, 3.4], [0.0, 0.0, 4.5, 5.7],

],

]

Example 2 (reflect mode):

data = [

[1.0, 1.2], [2.3, 3.4], [4.5, 5.7],

]

pads = [0, 2, 0, 0]

mode = ‘reflect’

output = [

[
[1.0, 1.2, 1.0, 1.2], [2.3, 3.4, 2.3, 3.4], [4.5, 5.7, 4.5, 5.7],

],

]

Example 3 (edge mode):

data = [

[1.0, 1.2], [2.3, 3.4], [4.5, 5.7],

]

pads = [0, 2, 0, 0]

mode = ‘edge’

output = [

[
[1.0, 1.0, 1.0, 1.2], [2.3, 2.3, 2.3, 3.4], [4.5, 4.5, 4.5, 5.7],

],

]

Node Inputs

data (T, single) – Input tensor.
pads (pads_constraint, single) – Tensor of integers indicating the number of padding elements to add or remove (if negative) at the beginning and end of each axis. For 2D input tensor, it is the number of pixels. pads should be a 1D tensor of shape [2 * input_rank]. pads format should be: [x1_begin, x2_begin,…,x1_end, x2_end,…], where xi_begin is the number of pad values added at the beginning of axis i and xi_end, the number of pad values added at the end of axis i.
constant_value (T, optional) – (Optional) A scalar value to be used if the mode chosen is constant (by default it is 0).

Node Outputs

output (T, single) – Tensor after padding.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64
pads_constraint – int64

Parameters

name – the name of the node.
mode (Optional [str], default=’constant’) – Supported modes: constant`(default), `reflect, edge

mode¶: Supported modes: constant`(default), `reflect, edge

class ONNXPow(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Pow takes input data (Tensor<T>) and exponent Tensor, and produces one output data (Tensor<T>) where the function f(x) = x^exponent, is applied to the data tensor elementwise. This operator supports multidirectional (i.e., Numpy-style) broadcasting; for more details please check [the doc](Broadcasting.md).

Node Inputs

X (T, single) – First operand, base of the exponent.
Y (T1, single) – Second operand, power of the exponent.

Node Outputs

Z (T, single) – Output tensor (same size as X)

Type Constraints

T – int32, int64, float16, float32, float64
T1 – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64

Parameters

name – the name of the node.

class ONNXQLinearConv(name, *, auto_pad='NOTSET', dilations=None, group=1, kernel_shape=None, pads=None, strides=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

The convolution operator consumes a quantized input tensor, its scale and zero point, a quantized filter, its scale and zero point, and output’s scale and zero point, and computes the quantized output. Each scale and zero-point pair must have same shape. It means they must be either scalars (per tensor) or 1-D tensors (per output channel). Each input or output and its related zero point must have same type. When bias is present it must be quantized using scale = input scale * weight scale and zero point as 0.

Node Inputs

x (T1, single) – Input data tensor from previous layer; has size (N x C x H x W), where N is the batch size, C is the number of channels, and H and W are the height and width. Note that this is for the 2D image. Otherwise the size is (N x C x D1 x D2 … x Dn). Optionally, if dimension denotation is in effect, the operation expects input data tensor to arrive with the dimension denotation of [DATA_BATCH, DATA_CHANNEL, DATA_FEATURE, DATA_FEATURE …].
x_scale (x_scale_constraint, single) – Scale tensor for input ‘x’. It’s a scalar, which means a per-tensor/layer quantization.
x_zero_point (T1, single) – Zero point tensor for input ‘x’. It’s a scalar, which means a per-tensor/layer quantization.
w (T2, single) – The weight tensor that will be used in the convolutions; has size (M x C/group x kH x kW), where C is the number of channels, and kH and kW are the height and width of the kernel, and M is the number of feature maps. For more than 2 dimensions, the kernel shape will be (M x C/group x k1 x k2 x … x kn), where (k1 x k2 x … kn) is the dimension of the kernel. Optionally, if dimension denotation is in effect, the operation expects the weight tensor to arrive with the dimension denotation of [FILTER_OUT_CHANNEL, FILTER_IN_CHANNEL, FILTER_SPATIAL, FILTER_SPATIAL …]. X.shape[1] == (W.shape[1] * group) == C (assuming zero based indices for the shape array). Or in other words FILTER_IN_CHANNEL should be equal to DATA_CHANNEL.
w_scale (w_scale_constraint, single) – Scale tensor for input ‘w’. It could be a scalar or a 1-D tensor, which means a per-tensor/layer or per output channel quantization. If it’s a 1-D tensor, its number of elements should be equal to the number of output channels (M).
w_zero_point (T2, single) – Zero point tensor for input ‘w’. It could be a scalar or a 1-D tensor, which means a per-tensor/layer or per output channel quantization. If it’s a 1-D tensor, its number of elements should be equal to the number of output channels (M).
y_scale (y_scale_constraint, single) – Scale tensor for output ‘y’. It’s a scalar, which means a per-tensor/layer quantization.
y_zero_point (T3, single) – Zero point tensor for output ‘y’. It’s a scalar, which means a per-tensor/layer quantization.
B (T4, optional) – Optional 1D bias to be added to the convolution, has size of M. Bias must be quantized using scale = x_scale * w_scale and zero_point = 0

Node Outputs

y (T3, single) – Output data tensor that contains the result of the convolution. The output dimensions are functions of the kernel size, stride size, and pad lengths.

Type Constraints

T1 – int8, uint8
T2 – int8, uint8
T3 – int8, uint8
T4 – int32
x_scale_constraint – float32
w_scale_constraint – float32
y_scale_constraint – float32

Parameters

name – the name of the node.
auto_pad (Optional [str], default=’NOTSET’) – auto_pad must be either NOTSET, SAME_UPPER, SAME_LOWER or VALID. Where default value is NOTSET, which means explicit padding is used. SAME_UPPER or SAME_LOWER mean pad the input so that the output spatial size match the input.In case of odd number add the extra padding at the end for SAME_UPPER and at the beginning for SAME_LOWER. VALID mean no padding.
dilations (Optional [List [int]], default=None) – dilation value along each spatial axis of the filter. If not present, the dilation defaults to 1 along each spatial axis.
group (Optional [int], default=1) – number of groups input channels and output channels are divided into. default is 1.
kernel_shape (Optional [List [int]], default=None) – The shape of the convolution kernel. If not present, should be inferred from input ‘w’.
pads (Optional [List [int]], default=None) – Padding for the beginning and ending along each spatial axis, it can take any value greater than or equal to 0.The value represent the number of pixels added to the beginning and end part of the corresponding axis.`pads` format should be as follow [x1_begin, x2_begin…x1_end, x2_end,…], where xi_begin the number ofpixels added at the beginning of axis i and xi_end, the number of pixels added at the end of axis i.This attribute cannot be used simultaneously with auto_pad attribute. If not present, the padding defaultsto 0 along start and end of each spatial axis.
strides (Optional [List [int]], default=None) – Stride along each spatial axis. If not present, the stride defaults to 1 along each spatial axis.

auto_pad¶: auto_pad must be either NOTSET, SAME_UPPER, SAME_LOWER or VALID. Where default value is NOTSET, which means explicit padding is used. SAME_UPPER or SAME_LOWER mean pad the input so that the output spatial size match the input.In case of odd number add the extra padding at the end for SAME_UPPER and at the beginning for SAME_LOWER. VALID mean no padding.

dilations¶: dilation value along each spatial axis of the filter. If not present, the dilation defaults to 1 along each spatial axis.

group¶: number of groups input channels and output channels are divided into. default is 1.

kernel_shape¶: The shape of the convolution kernel. If not present, should be inferred from input ‘w’.

pads¶: Padding for the beginning and ending along each spatial axis, it can take any value greater than or equal to 0.The value represent the number of pixels added to the beginning and end part of the corresponding axis.`pads` format should be as follow [x1_begin, x2_begin…x1_end, x2_end,…], where xi_begin the number ofpixels added at the beginning of axis i and xi_end, the number of pixels added at the end of axis i.This attribute cannot be used simultaneously with auto_pad attribute. If not present, the padding defaultsto 0 along start and end of each spatial axis.

strides¶: Stride along each spatial axis. If not present, the stride defaults to 1 along each spatial axis.

class ONNXQLinearMatMul(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Matrix product that behaves like numpy.matmul: https://docs.scipy.org/doc/numpy-1.13.0/reference/generated/numpy.matmul.html. It consumes two quantized input tensors, their scales and zero points, scale and zero point of output, and computes the quantized output. The quantization formula is y = saturate((x / y_scale) + y_zero_point). For (x / y_scale), it is rounding to nearest ties to even. Refer to https://en.wikipedia.org/wiki/Rounding for details. Scale and zero point must have same shape. They must be either scalar (per tensor) or 1-D tensor (per row for ‘a’ and per column for ‘b’). If scale and zero point are 1-D tensor, the number of elements of scale and zero point tensor of input ‘a’ and output ‘y’ should be equal to the number of rows of input ‘a’, and the number of elements of scale and zero point tensor of input ‘b’ should be equal to the number of columns of input ‘b’. Production must never overflow, and accumulation may overflow if and only if in 32 bits.

Node Inputs

a (T1, single) – N-dimensional quantized matrix a
a_scale (a_scale_constraint, single) – scale of quantized input a
a_zero_point (T1, single) – zero point of quantized input a
b (T2, single) – N-dimensional quantized matrix b
b_scale (b_scale_constraint, single) – scale of quantized input b
b_zero_point (T2, single) – zero point of quantized input b
y_scale (y_scale_constraint, single) – scale of quantized output y
y_zero_point (T3, single) – zero point of quantized output y

Node Outputs

y (T3, single) – Quantized matrix multiply results from a * b

Type Constraints

T1 – int8, uint8
T2 – int8, uint8
T3 – int8, uint8
a_scale_constraint – float32
b_scale_constraint – float32
y_scale_constraint – float32

Parameters

name – the name of the node.

class ONNXQuantizeLinear(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

The linear per-tensor/layer quantization operator. It consumes a high precision tensor, a scale, a zero point to compute the low precision / quantized tensor. The quantization formula is y = saturate ((x / y_scale) + y_zero_point). For saturation, it saturates to [0, 255] if it’s uint8, or [-128, 127] if it’s int8. For (x / y_scale), it’s rounding to nearest ties to even. Refer to https://en.wikipedia.org/wiki/Rounding for details. ‘y_zero_point’ and ‘y’ must have same type.

Node Inputs

x (T1, single) – N-D full precision Input tensor to be quantized.
y_scale (y_scale_constraint, single) – Scale for doing quantization to get ‘y’. It’s a scalar, which means a per-tensor/layer quantization.
y_zero_point (T2, optional) – Zero point for doing quantization to get ‘y’. It’s a scalar, which means a per-tensor/layer quantization. Default value is uint8 typed 0 if it’s not specified.

Node Outputs

y (T2, single) – N-D quantized output tensor. It has same shape as input ‘x’.

Type Constraints

T1 – float32, int32
T2 – int8, uint8
y_scale_constraint – float32

Parameters

name – the name of the node.

class ONNXRNN(name, *, activation_alpha=None, activation_beta=None, activations=['Tanh', 'Tanh'], clip=None, direction='forward', hidden_size=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Computes an one-layer simple RNN. This operator is usually supported via some custom implementation such as CuDNN.

Notations:

X - input tensor

i - input gate

t - time step (t-1 means previous time step)

Wi - W parameter weight matrix for input gate

Ri - R recurrence weight matrix for input gate

Wbi - W parameter bias vector for input gate

Rbi - R parameter bias vector for input gate

WBi - W parameter weight matrix for backward input gate

RBi - R recurrence weight matrix for backward input gate

WBbi - WR bias vectors for backward input gate

RBbi - RR bias vectors for backward input gate

H - Hidden state

num_directions - 2 if direction == bidirectional else 1

Activation functions:

Relu(x) - max(0, x)

Tanh(x) - (1 - e^{-2x})/(1 + e^{-2x})

Sigmoid(x) - 1/(1 + e^{-x})

(NOTE: Below are optional)

Affine(x) - alpha*x + beta

LeakyRelu(x) - x if x >= 0 else alpha * x

ThresholdedRelu(x) - x if x >= alpha else 0

ScaledTanh(x) - alpha*Tanh(beta*x)

HardSigmoid(x) - min(max(alpha*x + beta, 0), 1)

Elu(x) - x if x >= 0 else alpha*(e^x - 1)

Softsign(x) - x/(1 + |x|)

Softplus(x) - log(1 + e^x)

Equations (Default: f=Tanh):

Ht = f(Xt*(Wi^T) + Ht-1*(Ri^T) + Wbi + Rbi)

This operator has optional inputs/outputs. See [the doc](IR.md) for more details about the representation of optional arguments. An empty string may be used in the place of an actual argument’s name to indicate a missing argument. Trailing optional arguments (those not followed by an argument that is present) may also be simply omitted.

Node Inputs

X (T, single) – The input sequences packed (and potentially padded) into one 3-D tensor with the shape of [seq_length, batch_size, input_size].
W (T, single) – The weight tensor for input gate. Concatenation of Wi and WBi (if bidirectional). The tensor has shape [num_directions, hidden_size, input_size].
R (T, single) – The recurrence weight tensor. Concatenation of Ri and RBi (if bidirectional). The tensor has shape [num_directions, hidden_size, hidden_size].
B (T, optional) – The bias tensor for input gate. Concatenation of [Wbi, Rbi] and [WBbi, RBbi] (if bidirectional). The tensor has shape [num_directions, 2*hidden_size]. Optional: If not specified - assumed to be 0.
sequence_lens (T1, optional) – Optional tensor specifying lengths of the sequences in a batch. If not specified - assumed all sequences in the batch to have length seq_length. It has shape [batch_size].
initial_h (T, optional) – Optional initial value of the hidden. If not specified - assumed to be 0. It has shape [num_directions, batch_size, hidden_size].

Node Outputs

Y (T, optional) – A tensor that concats all the intermediate output values of the hidden. It has shape [seq_length, num_directions, batch_size, hidden_size].
Y_h (T, optional) – The last output value of the hidden. It has shape [num_directions, batch_size, hidden_size].

Type Constraints

T – float16, float32, float64
T1 – int32

Parameters

name – the name of the node.
activation_alpha (Optional [List [float]], default=None) – Optional scaling values used by some activation functions. The values are consumed in the order of activation functions, for example (f, g, h) in LSTM. Default values are the same as of corresponding ONNX operators.For example with LeakyRelu, the default alpha is 0.01.
activation_beta (Optional [List [float]], default=None) – Optional scaling values used by some activation functions. The values are consumed in the order of activation functions, for example (f, g, h) in LSTM. Default values are the same as of corresponding ONNX operators.
activations (Optional [List [str]], default=[‘Tanh’, ‘Tanh’]) – One (or two if bidirectional) activation function for input gate. The activation function must be one of the activation functions specified above. Optional: Default Tanh if not specified.
clip (Optional [float], default=None) – Cell clip threshold. Clipping bounds the elements of a tensor in the range of [-threshold, +threshold] and is applied to the input of activations. No clip if not specified.
direction (Optional [str], default=’forward’) – Specify if the RNN is forward, reverse, or bidirectional. Must be one of forward (default), reverse, or bidirectional.
hidden_size (Optional [int], default=None) – Number of neurons in the hidden layer

activation_alpha¶: Optional scaling values used by some activation functions. The values are consumed in the order of activation functions, for example (f, g, h) in LSTM. Default values are the same as of corresponding ONNX operators.For example with LeakyRelu, the default alpha is 0.01.

activation_beta¶: Optional scaling values used by some activation functions. The values are consumed in the order of activation functions, for example (f, g, h) in LSTM. Default values are the same as of corresponding ONNX operators.

activations¶: One (or two if bidirectional) activation function for input gate. The activation function must be one of the activation functions specified above. Optional: Default Tanh if not specified.

clip¶: Cell clip threshold. Clipping bounds the elements of a tensor in the range of [-threshold, +threshold] and is applied to the input of activations. No clip if not specified.

direction¶: Specify if the RNN is forward, reverse, or bidirectional. Must be one of forward (default), reverse, or bidirectional.

hidden_size¶: Number of neurons in the hidden layer

class ONNXRandomNormal(name, *, dtype=1, mean=0.0, scale=1.0, seed=None, shape)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Generate a tensor with random values drawn from a normal distribution. The shape of the tensor is specified by the shape argument and the parameter of the normal distribution specified by mean and scale.

The data type is specified by the ‘dtype’ argument. The ‘dtype’ argument must be one of the data types specified in the ‘DataType’ enum field in the TensorProto message.

Node Inputs

Node Outputs

output (T, single) – Output tensor of random values drawn from normal distribution

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.
dtype (Optional [int], default=1) – The data type for the elements of the output tensor. Default is TensorProto::FLOAT.
mean (Optional [float], default=0.0) – The mean of the normal distribution.
scale (Optional [float], default=1.0) – The standard deviation of the normal distribution.
seed (Optional [float], default=None) – (Optional) Seed to the random generator, if not specified we will auto generate one.
shape (List [int]) – The shape of the output tensor.

dtype¶: The data type for the elements of the output tensor. Default is TensorProto::FLOAT.

mean¶: The mean of the normal distribution.

scale¶: The standard deviation of the normal distribution.

seed¶: (Optional) Seed to the random generator, if not specified we will auto generate one.

shape¶: The shape of the output tensor.

class ONNXRandomNormalLike(name, *, dtype=None, mean=0.0, scale=1.0, seed=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Generate a tensor with random values drawn from a normal distribution. The shape of the output tensor is copied from the shape of the input tensor, and the parameters of the normal distribution are specified by mean and scale.

The data type is specified by the ‘dtype’ argument, or copied from the input tensor if not provided. The ‘dtype’ argument must be one of the data types specified in the ‘DataType’ enum field in the TensorProto message, and be valid as an output type.

Node Inputs

input (T1, single) – Input tensor to copy shape and optionally type information from.

Node Outputs

output (T2, single) – Output tensor of random values drawn from normal distribution

Type Constraints

T1 – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128
T2 – float16, float32, float64

Parameters

name – the name of the node.
dtype (Optional [int], default=None) – (Optional) The data type for the elements of the output tensor, if not specified, we will use the data type of the input tensor.
mean (Optional [float], default=0.0) – The mean of the normal distribution.
scale (Optional [float], default=1.0) – The standard deviation of the normal distribution.
seed (Optional [float], default=None) – (Optional) Seed to the random generator, if not specified we will auto generate one.

dtype¶: (Optional) The data type for the elements of the output tensor, if not specified, we will use the data type of the input tensor.

mean¶: The mean of the normal distribution.

scale¶: The standard deviation of the normal distribution.

seed¶: (Optional) Seed to the random generator, if not specified we will auto generate one.

class ONNXRandomUniform(name, *, dtype=1, high=1.0, low=0.0, seed=None, shape)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Generate a tensor with random values drawn from a uniform distribution. The shape of the tensor is specified by the shape argument and the range by low and high.

The data type is specified by the ‘dtype’ argument. The ‘dtype’ argument must be one of the data types specified in the ‘DataType’ enum field in the TensorProto message.

Node Inputs

Node Outputs

output (T, single) – Output tensor of random values drawn from uniform distribution

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.
dtype (Optional [int], default=1) – The data type for the elements of the output tensor. If not specified, default is TensorProto::FLOAT.
high (Optional [float], default=1.0) – Upper boundary of the output values.
low (Optional [float], default=0.0) – Lower boundary of the output values.
seed (Optional [float], default=None) – (Optional) Seed to the random generator, if not specified we will auto generate one.
shape (List [int]) – The shape of the output tensor.

dtype¶: The data type for the elements of the output tensor. If not specified, default is TensorProto::FLOAT.

high¶: Upper boundary of the output values.

low¶: Lower boundary of the output values.

seed¶: (Optional) Seed to the random generator, if not specified we will auto generate one.

shape¶: The shape of the output tensor.

class ONNXRandomUniformLike(name, *, dtype=None, high=1.0, low=0.0, seed=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Generate a tensor with random values drawn from a uniform distribution. The shape of the output tensor is copied from the shape of the input tensor, and the parameters of the uniform distribution are specified by low and high.

The data type is specified by the ‘dtype’ argument, or copied from the input tensor if not provided. The ‘dtype’ argument must be one of the data types specified in the ‘DataType’ enum field in the TensorProto message and be valid as an output type.

Node Inputs

input (T1, single) – Input tensor to copy shape and optionally type information from.

Node Outputs

output (T2, single) – Output tensor of random values drawn from uniform distribution

Type Constraints

T1 – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128
T2 – float16, float32, float64

Parameters

name – the name of the node.
dtype (Optional [int], default=None) – (Optional) The data type for the elements of the output tensor, if not specified, we will use the data type of the input tensor.
high (Optional [float], default=1.0) – Upper boundary of the output values.
low (Optional [float], default=0.0) – Lower boundary of the output values.
seed (Optional [float], default=None) – (Optional) Seed to the random generator, if not specified we will auto generate one.

dtype¶: (Optional) The data type for the elements of the output tensor, if not specified, we will use the data type of the input tensor.

high¶: Upper boundary of the output values.

low¶: Lower boundary of the output values.

seed¶: (Optional) Seed to the random generator, if not specified we will auto generate one.

class ONNXRange(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Generate a tensor containing a sequence of numbers that begin at start and extends by increments of delta up to limit (exclusive).

The number of elements in the output of range is computed as below-

number_of_elements = max( ceil( (limit - start) / delta ) , 0 )

The pseudocode determining the contents of the output is shown below-

for(int i=0; i<number_of_elements; ++i)

{

` output[i] = start + (i * delta); `

}

Example 1 Inputs: start = 3, limit = 9, delta = 3 Output: [3, 6]

Example 2 Inputs: start = 10, limit = 4, delta = -2 Output: [10, 8, 6]

Node Inputs

start (T, single) – Scalar. First entry for the range of output values.
limit (T, single) – Scalar. Exclusive upper limit for the range of output values.
delta (T, single) – Scalar. Value to step by.

Node Outputs

output (T, single) – A 1-D tensor with same type as the inputs containing generated range of values.

Type Constraints

T – float32, float64, int16, int32, int64

Parameters

name – the name of the node.

class ONNXReciprocal(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Reciprocal takes one input data (Tensor<T>) and produces one output data (Tensor<T>) where the reciprocal is, y = 1/x, is applied to the tensor elementwise.

Node Inputs

X (T, single) – Input tensor

Node Outputs

Y (T, single) – Output tensor

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXReduceL1(name, *, axes=None, keepdims=1)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Computes the L1 norm of the input tensor’s element along the provided axes. The resulted tensor has the same rank as the input if keepdims equal 1. If keepdims equal 0, then the resulted tensor have the reduced dimension pruned.

The above behavior is similar to numpy, with the exception that numpy default keepdims to False instead of True.

Node Inputs

data (T, single) – An input tensor.

Node Outputs

reduced (T, single) – Reduced output tensor.

Type Constraints

T – uint32, uint64, int32, int64, float16, float32, float64

Parameters

name – the name of the node.
axes (Optional [List [int]], default=None) – A list of integers, along which to reduce. The default is to reduce over all the dimensions of the input tensor. Accepted range is [-r, r-1] where r = rank(data).
keepdims (Optional [int], default=1) – Keep the reduced dimension or not, default 1 mean keep reduced dimension.

axes¶: A list of integers, along which to reduce. The default is to reduce over all the dimensions of the input tensor. Accepted range is [-r, r-1] where r = rank(data).

keepdims¶: Keep the reduced dimension or not, default 1 mean keep reduced dimension.

class ONNXReduceL2(name, *, axes=None, keepdims=1)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Computes the L2 norm of the input tensor’s element along the provided axes. The resulted tensor has the same rank as the input if keepdims equal 1. If keepdims equal 0, then the resulted tensor have the reduced dimension pruned.

The above behavior is similar to numpy, with the exception that numpy default keepdims to False instead of True.

Node Inputs

data (T, single) – An input tensor.

Node Outputs

reduced (T, single) – Reduced output tensor.

Type Constraints

T – uint32, uint64, int32, int64, float16, float32, float64

Parameters

name – the name of the node.
axes (Optional [List [int]], default=None) – A list of integers, along which to reduce. The default is to reduce over all the dimensions of the input tensor. Accepted range is [-r, r-1] where r = rank(data).
keepdims (Optional [int], default=1) – Keep the reduced dimension or not, default 1 mean keep reduced dimension.

axes¶: A list of integers, along which to reduce. The default is to reduce over all the dimensions of the input tensor. Accepted range is [-r, r-1] where r = rank(data).

keepdims¶: Keep the reduced dimension or not, default 1 mean keep reduced dimension.

class ONNXReduceLogSum(name, *, axes=None, keepdims=1)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Computes the log sum of the input tensor’s element along the provided axes. The resulted tensor has the same rank as the input if keepdims equal 1. If keepdims equal 0, then the resulted tensor have the reduced dimension pruned.

The above behavior is similar to numpy, with the exception that numpy default keepdims to False instead of True.

Node Inputs

data (T, single) – An input tensor.

Node Outputs

reduced (T, single) – Reduced output tensor.

Type Constraints

T – uint32, uint64, int32, int64, float16, float32, float64

Parameters

name – the name of the node.
axes (Optional [List [int]], default=None) – A list of integers, along which to reduce. The default is to reduce over all the dimensions of the input tensor. Accepted range is [-r, r-1] where r = rank(data).
keepdims (Optional [int], default=1) – Keep the reduced dimension or not, default 1 mean keep reduced dimension.

axes¶: A list of integers, along which to reduce. The default is to reduce over all the dimensions of the input tensor. Accepted range is [-r, r-1] where r = rank(data).

keepdims¶: Keep the reduced dimension or not, default 1 mean keep reduced dimension.

class ONNXReduceLogSumExp(name, *, axes=None, keepdims=1)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Computes the log sum exponent of the input tensor’s element along the provided axes. The resulted tensor has the same rank as the input if keepdims equal 1. If keepdims equal 0, then the resulted tensor have the reduced dimension pruned.

The above behavior is similar to numpy, with the exception that numpy default keepdims to False instead of True.

Node Inputs

data (T, single) – An input tensor.

Node Outputs

reduced (T, single) – Reduced output tensor.

Type Constraints

T – uint32, uint64, int32, int64, float16, float32, float64

Parameters

name – the name of the node.
axes (Optional [List [int]], default=None) – A list of integers, along which to reduce. The default is to reduce over all the dimensions of the input tensor. Accepted range is [-r, r-1] where r = rank(data).
keepdims (Optional [int], default=1) – Keep the reduced dimension or not, default 1 mean keep reduced dimension.

axes¶: A list of integers, along which to reduce. The default is to reduce over all the dimensions of the input tensor. Accepted range is [-r, r-1] where r = rank(data).

keepdims¶: Keep the reduced dimension or not, default 1 mean keep reduced dimension.

class ONNXReduceMax(name, *, axes=None, keepdims=1)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Computes the max of the input tensor’s element along the provided axes. The resulted tensor has the same rank as the input if keepdims equal 1. If keepdims equal 0, then the resulted tensor have the reduced dimension pruned.

The above behavior is similar to numpy, with the exception that numpy default keepdims to False instead of True.

Node Inputs

data (T, single) – An input tensor.

Node Outputs

reduced (T, single) – Reduced output tensor.

Type Constraints

T – uint32, uint64, int32, int64, float16, float32, float64, uint8, int8

Parameters

name – the name of the node.
axes (Optional [List [int]], default=None) – A list of integers, along which to reduce. The default is to reduce over all the dimensions of the input tensor. Accepted range is [-r, r-1] where r = rank(data).
keepdims (Optional [int], default=1) – Keep the reduced dimension or not, default 1 mean keep reduced dimension.

axes¶: A list of integers, along which to reduce. The default is to reduce over all the dimensions of the input tensor. Accepted range is [-r, r-1] where r = rank(data).

keepdims¶: Keep the reduced dimension or not, default 1 mean keep reduced dimension.

class ONNXReduceMean(name, *, axes=None, keepdims=1)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Computes the mean of the input tensor’s element along the provided axes. The resulted tensor has the same rank as the input if keepdims equal 1. If keepdims equal 0, then the resulted tensor have the reduced dimension pruned.

The above behavior is similar to numpy, with the exception that numpy default keepdims to False instead of True.

Node Inputs

data (T, single) – An input tensor.

Node Outputs

reduced (T, single) – Reduced output tensor.

Type Constraints

T – uint32, uint64, int32, int64, float16, float32, float64

Parameters

name – the name of the node.
axes (Optional [List [int]], default=None) – A list of integers, along which to reduce. The default is to reduce over all the dimensions of the input tensor. Accepted range is [-r, r-1] where r = rank(data).
keepdims (Optional [int], default=1) – Keep the reduced dimension or not, default 1 mean keep reduced dimension.

axes¶: A list of integers, along which to reduce. The default is to reduce over all the dimensions of the input tensor. Accepted range is [-r, r-1] where r = rank(data).

keepdims¶: Keep the reduced dimension or not, default 1 mean keep reduced dimension.

class ONNXReduceMin(name, *, axes=None, keepdims=1)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Computes the min of the input tensor’s element along the provided axes. The resulted tensor has the same rank as the input if keepdims equal 1. If keepdims equal 0, then the resulted tensor have the reduced dimension pruned.

The above behavior is similar to numpy, with the exception that numpy default keepdims to False instead of True.

Node Inputs

data (T, single) – An input tensor.

Node Outputs

reduced (T, single) – Reduced output tensor.

Type Constraints

T – uint32, uint64, int32, int64, float16, float32, float64, uint8, int8

Parameters

name – the name of the node.
axes (Optional [List [int]], default=None) – A list of integers, along which to reduce. The default is to reduce over all the dimensions of the input tensor. Accepted range is [-r, r-1] where r = rank(data).
keepdims (Optional [int], default=1) – Keep the reduced dimension or not, default 1 mean keep reduced dimension.

axes¶: A list of integers, along which to reduce. The default is to reduce over all the dimensions of the input tensor. Accepted range is [-r, r-1] where r = rank(data).

keepdims¶: Keep the reduced dimension or not, default 1 mean keep reduced dimension.

class ONNXReduceProd(name, *, axes=None, keepdims=1)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Computes the product of the input tensor’s element along the provided axes. The resulted tensor has the same rank as the input if keepdims equal 1. If keepdims equal 0, then the resulted tensor have the reduced dimension pruned.

The above behavior is similar to numpy, with the exception that numpy default keepdims to False instead of True.

Node Inputs

data (T, single) – An input tensor.

Node Outputs

reduced (T, single) – Reduced output tensor.

Type Constraints

T – uint32, uint64, int32, int64, float16, float32, float64

Parameters

name – the name of the node.
axes (Optional [List [int]], default=None) – A list of integers, along which to reduce. The default is to reduce over all the dimensions of the input tensor. Accepted range is [-r, r-1] where r = rank(data).
keepdims (Optional [int], default=1) – Keep the reduced dimension or not, default 1 mean keep reduced dimension.

axes¶: A list of integers, along which to reduce. The default is to reduce over all the dimensions of the input tensor. Accepted range is [-r, r-1] where r = rank(data).

keepdims¶: Keep the reduced dimension or not, default 1 mean keep reduced dimension.

class ONNXReduceSum(name, *, axes=None, keepdims=1)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Computes the sum of the input tensor’s element along the provided axes. The resulted tensor has the same rank as the input if keepdims equal 1. If keepdims equal 0, then the resulted tensor have the reduced dimension pruned.

The above behavior is similar to numpy, with the exception that numpy default keepdims to False instead of True.

Node Inputs

data (T, single) – An input tensor.

Node Outputs

reduced (T, single) – Reduced output tensor.

Type Constraints

T – uint32, uint64, int32, int64, float16, float32, float64

Parameters

name – the name of the node.
axes (Optional [List [int]], default=None) – A list of integers, along which to reduce. The default is to reduce over all the dimensions of the input tensor. Accepted range is [-r, r-1] where r = rank(data).
keepdims (Optional [int], default=1) – Keep the reduced dimension or not, default 1 mean keep reduced dimension.

axes¶: A list of integers, along which to reduce. The default is to reduce over all the dimensions of the input tensor. Accepted range is [-r, r-1] where r = rank(data).

keepdims¶: Keep the reduced dimension or not, default 1 mean keep reduced dimension.

class ONNXReduceSumSquare(name, *, axes=None, keepdims=1)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Computes the sum square of the input tensor’s element along the provided axes. The resulted tensor has the same rank as the input if keepdims equal 1. If keepdims equal 0, then the resulted tensor have the reduced dimension pruned.

The above behavior is similar to numpy, with the exception that numpy default keepdims to False instead of True.

Node Inputs

data (T, single) – An input tensor.

Node Outputs

reduced (T, single) – Reduced output tensor.

Type Constraints

T – uint32, uint64, int32, int64, float16, float32, float64

Parameters

name – the name of the node.
axes (Optional [List [int]], default=None) – A list of integers, along which to reduce. The default is to reduce over all the dimensions of the input tensor. Accepted range is [-r, r-1] where r = rank(data).
keepdims (Optional [int], default=1) – Keep the reduced dimension or not, default 1 mean keep reduced dimension.

axes¶: A list of integers, along which to reduce. The default is to reduce over all the dimensions of the input tensor. Accepted range is [-r, r-1] where r = rank(data).

keepdims¶: Keep the reduced dimension or not, default 1 mean keep reduced dimension.

class ONNXRelu(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Relu takes one input data (Tensor<T>) and produces one output data (Tensor<T>) where the rectified linear function, y = max(0, x), is applied to the tensor elementwise.

Node Inputs

X (T, single) – Input tensor

Node Outputs

Y (T, single) – Output tensor

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXReshape(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Reshape the input tensor similar to numpy.reshape. First input is the data tensor, second input is a shape tensor which specifies the output shape. It outputs the reshaped tensor. At most one dimension of the new shape can be -1. In this case, the value is inferred from the size of the tensor and the remaining dimensions. A dimension could also be 0, in which case the actual dimension value is unchanged (i.e. taken from the input tensor).

Node Inputs

data (T, single) – An input tensor.
shape (shape_constraint, single) – Specified shape for output.

Node Outputs

reshaped (T, single) – Reshaped data.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128
shape_constraint – int64

Parameters

name – the name of the node.

class ONNXResize(name, *, coordinate_transformation_mode='half_pixel', cubic_coeff_a=- 0.75, exclude_outside=0, extrapolation_value=0.0, mode='nearest', nearest_mode='round_prefer_floor')¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Resize the input tensor. In general, it calculates every value in the output tensor as a weighted average of neighborhood (a.k.a. sampling locations) in the input tensor. Each dimension value of the output tensor is:

output_dimension = floor(input_dimension * (roi_end - roi_start) * scale) if input "sizes" is not specified.

Node Inputs

X (T1, single) – N-D tensor
roi (T2, single) – 1-D tensor given as [start1, …, startN, end1, …, endN], where N is the rank of X. The RoIs’ coordinates are normalized in the coordinate system of the input image. It only takes effect when coordinate_transformation_mode is “tf_crop_and_resize”
scales (scales_constraint, single) – The scale array along each dimension. It takes value greater than 0. If it’s less than 1, it’s sampling down, otherwise, it’s upsampling. The number of elements of ‘scales’ should be the same as the rank of input ‘X’. Only one of ‘scales’ and ‘sizes’ can be specified. If ‘size’ is needed, the user can use an empty string as the name of ‘scales’ in this operator’s input list.
sizes (sizes_constraint, optional) – The size of the output tensor. The number of elements of ‘sizes’ should be the same as the rank of input ‘X’. Only one of ‘scales’ and ‘sizes’ can be specified.

Node Outputs

Y (T1, single) – N-D tensor after resizing

Type Constraints

T1 – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128
T2 – float16, float32, float64
scales_constraint – float32
sizes_constraint – int64

Parameters

name – the name of the node.
coordinate_transformation_mode –

This attribute describes how to transform the coordinate in the resized tensor to the coordinate in the original tensor.

The coordinate of each dimension is transformed individually. Let’s describe a case using axis x as an example. Denote x_resized as the coordinate of axis x in the resized tensor, x_original as the coordinate of axis x in the original tensor, length_original as the length of the original tensor in axis x, length_resized as the length of the resized tensor in axis x, roi_x = (start_x, end_x) of the axis x in input “roi”, scale = length_resized / length_original,

if coordinate_transformation_mode is “half_pixel”, x_original = (x_resized + 0.5) / scale - 0.5,

if coordinate_transformation_mode is “pytorch_half_pixel”, x_original = length_resized > 1 ? (x_resized + 0.5) / scale - 0.5 : 0,

if coordinate_transformation_mode is “align_corners”, x_original = x_resized * (length_original - 1) / (length_resized - 1),

if coordinate_transformation_mode is “asymmetric”, x_original = x_resized / scale,

if coordinate_transformation_mode is “tf_half_pixel_for_nn”, x_original = (x_resized + 0.5) / scale,

if coordinate_transformation_mode is “tf_crop_and_resize”, x_original = length_resized > 1 ? start_x * (length_original - 1) + x_resized * (end_x - start_x) * (length_original - 1) / (length_resized - 1) : 0.5 * (start_x + end_x) * (length_original - 1). :type coordinate_transformation_mode: Optional [str], default=’half_pixel’ :param cubic_coeff_a: The coefficient ‘a’ used in cubic interpolation. Two common choice are -0.5 (in some cases of TensorFlow) and -0.75 (in PyTorch). Check out Equation (4) in https://ieeexplore.ieee.org/document/1163711 for the details. This attribute is valid only if “mode” is “cubic”. :type cubic_coeff_a: Optional [float], default=-0.75 :param exclude_outside: If set to 1, the weight of sampling locations outside the tensor will be set to 0 and the weight will be renormalized so that their sum is 1.0. The default value is 0. :type exclude_outside: Optional [int], default=0 :param extrapolation_value: When coordinate_transformation_mode is “tf_crop_and_resize” and x_original is outside the range [0, length_original - 1], this value is used as the corresponding output value. Default is 0.0f. :type extrapolation_value: Optional [float], default=0.0 :param mode: Three interpolation modes: nearest (default), linear and cubic. The “linear” mode includes linear interpolation for 1D tensor and N-linear interpolation for N-D tensor (for example, bilinear interpolation for 2D tensor). The “cubic” mode includes cubic interpolation for 1D tensor and N-cubic interpolation for N-D tensor (for example, bicubic interpolation for 2D tensor). :type mode: Optional [str], default=’nearest’ :param nearest_mode: Four modes: round_prefer_floor (default, as known as round half down), round_prefer_ceil (as known as round half up), floor, ceil. Only used by nearest interpolation. It indicates how to get “nearest” pixel in input tensor from x_original, so this attribute is valid only if “mode” is “nearest”. :type nearest_mode: Optional [str], default=’round_prefer_floor’

coordinate_transformation_mode¶

This attribute describes how to transform the coordinate in the resized tensor to the coordinate in the original tensor.

The coordinate of each dimension is transformed individually. Let’s describe a case using axis x as an example. Denote x_resized as the coordinate of axis x in the resized tensor, x_original as the coordinate of axis x in the original tensor, length_original as the length of the original tensor in axis x, length_resized as the length of the resized tensor in axis x, roi_x = (start_x, end_x) of the axis x in input “roi”, scale = length_resized / length_original,

if coordinate_transformation_mode is “half_pixel”, x_original = (x_resized + 0.5) / scale - 0.5,

if coordinate_transformation_mode is “pytorch_half_pixel”, x_original = length_resized > 1 ? (x_resized + 0.5) / scale - 0.5 : 0,

if coordinate_transformation_mode is “align_corners”, x_original = x_resized * (length_original - 1) / (length_resized - 1),

if coordinate_transformation_mode is “asymmetric”, x_original = x_resized / scale,

if coordinate_transformation_mode is “tf_half_pixel_for_nn”, x_original = (x_resized + 0.5) / scale,

if coordinate_transformation_mode is “tf_crop_and_resize”, x_original = length_resized > 1 ? start_x * (length_original - 1) + x_resized * (end_x - start_x) * (length_original - 1) / (length_resized - 1) : 0.5 * (start_x + end_x) * (length_original - 1).

cubic_coeff_a¶: The coefficient ‘a’ used in cubic interpolation. Two common choice are -0.5 (in some cases of TensorFlow) and -0.75 (in PyTorch). Check out Equation (4) in https://ieeexplore.ieee.org/document/1163711 for the details. This attribute is valid only if “mode” is “cubic”.

exclude_outside¶: If set to 1, the weight of sampling locations outside the tensor will be set to 0 and the weight will be renormalized so that their sum is 1.0. The default value is 0.

extrapolation_value¶: When coordinate_transformation_mode is “tf_crop_and_resize” and x_original is outside the range [0, length_original - 1], this value is used as the corresponding output value. Default is 0.0f.

mode¶: Three interpolation modes: nearest (default), linear and cubic. The “linear” mode includes linear interpolation for 1D tensor and N-linear interpolation for N-D tensor (for example, bilinear interpolation for 2D tensor). The “cubic” mode includes cubic interpolation for 1D tensor and N-cubic interpolation for N-D tensor (for example, bicubic interpolation for 2D tensor).

nearest_mode¶: Four modes: round_prefer_floor (default, as known as round half down), round_prefer_ceil (as known as round half up), floor, ceil. Only used by nearest interpolation. It indicates how to get “nearest” pixel in input tensor from x_original, so this attribute is valid only if “mode” is “nearest”.

class ONNXReverseSequence(name, *, batch_axis=1, time_axis=0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Reverse batch of sequences having different lengths specified by sequence_lens.

For each slice i iterating on batch axis, the operator reverses the first sequence_lens[i] elements on time axis, and copies elements whose index’s beyond sequence_lens[i] to the output. So the output slice i contains reversed sequences on the first sequence_lens[i] elements, then have original values copied for the other elements.

Example 1:

input = [[0.0, 4.0, 8.0, 12.0],: [1.0, 5.0, 9.0, 13.0], [2.0, 6.0, 10.0, 14.0], [3.0, 7.0, 11.0, 15.0]]

sequence_lens = [4, 3, 2, 1] time_axis = 0 batch_axis = 1

output = [[3.0, 6.0, 9.0, 12.0],: [2.0, 5.0, 8.0, 13.0], [1.0, 4.0, 10.0, 14.0], [0.0, 7.0, 11.0, 15.0]]

Example 2:

input = [[0.0, 1.0, 2.0, 3.0 ],: [4.0, 5.0, 6.0, 7.0 ], [8.0, 9.0, 10.0, 11.0], [12.0, 13.0, 14.0, 15.0]]

sequence_lens = [1, 2, 3, 4] time_axis = 1 batch_axis = 0

output = [[0.0, 1.0, 2.0, 3.0 ],: [5.0, 4.0, 6.0, 7.0 ], [10.0, 9.0, 8.0, 11.0], [15.0, 14.0, 13.0, 12.0]]

Node Inputs

input (T, single) – Tensor of rank r >= 2.
sequence_lens (sequence_lens_constraint, single) – Tensor specifying lengths of the sequences in a batch. It has shape [batch_size].

Node Outputs

Y (T, single) – Tensor with same shape of input.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128
sequence_lens_constraint – int64

Parameters

name – the name of the node.
batch_axis (Optional [int], default=1) – (Optional) Specify which axis is batch axis. Must be one of 1 (default), or 0.
time_axis (Optional [int], default=0) – (Optional) Specify which axis is time axis. Must be one of 0 (default), or 1.

batch_axis¶: (Optional) Specify which axis is batch axis. Must be one of 1 (default), or 0.

time_axis¶: (Optional) Specify which axis is time axis. Must be one of 0 (default), or 1.

class ONNXRoiAlign(name, *, mode='avg', output_height=1, output_width=1, sampling_ratio=0, spatial_scale=1.0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Region of Interest (RoI) align operation described in the [Mask R-CNN paper](https://arxiv.org/abs/1703.06870). RoiAlign consumes an input tensor X and region of interests (rois) to apply pooling across each RoI; it produces a 4-D tensor of shape (num_rois, C, output_height, output_width).

RoiAlign is proposed to avoid the misalignment by removing quantizations while converting from original image into feature map and from feature map into RoI feature; in each ROI bin, the value of the sampled locations are computed directly through bilinear interpolation.

Node Inputs

X (T1, single) – Input data tensor from the previous operator; 4-D feature map of shape (N, C, H, W), where N is the batch size, C is the number of channels, and H and W are the height and the width of the data.
rois (T1, single) – RoIs (Regions of Interest) to pool over; rois is 2-D input of shape (num_rois, 4) given as [[x1, y1, x2, y2], …]. The RoIs’ coordinates are in the coordinate system of the input image. Each coordinate set has a 1:1 correspondence with the ‘batch_indices’ input.
batch_indices (T2, single) – 1-D tensor of shape (num_rois,) with each element denoting the index of the corresponding image in the batch.

Node Outputs

Y (T1, single) – RoI pooled output, 4-D tensor of shape (num_rois, C, output_height, output_width). The r-th batch element Y[r-1] is a pooled feature map corresponding to the r-th RoI X[r-1].

Type Constraints

T1 – float16, float32, float64
T2 – int64

Parameters

name – the name of the node.
mode (Optional [str], default=’avg’) – The pooling method. Two modes are supported: ‘avg’ and ‘max’. Default is ‘avg’.
output_height (Optional [int], default=1) – default 1; Pooled output Y’s height.
output_width (Optional [int], default=1) – default 1; Pooled output Y’s width.
sampling_ratio (Optional [int], default=0) – Number of sampling points in the interpolation grid used to compute the output value of each pooled output bin. If > 0, then exactly sampling_ratio x sampling_ratio grid points are used. If == 0, then an adaptive number of grid points are used (computed as ceil(roi_width / output_width), and likewise for height). Default is 0.
spatial_scale (Optional [float], default=1.0) – Multiplicative spatial scale factor to translate ROI coordinates from their input spatial scale to the scale used when pooling, i.e., spatial scale of the input feature map X relative to the input image. E.g.; default is 1.0f.

mode¶: The pooling method. Two modes are supported: ‘avg’ and ‘max’. Default is ‘avg’.

output_height¶: default 1; Pooled output Y’s height.

output_width¶: default 1; Pooled output Y’s width.

sampling_ratio¶: Number of sampling points in the interpolation grid used to compute the output value of each pooled output bin. If > 0, then exactly sampling_ratio x sampling_ratio grid points are used. If == 0, then an adaptive number of grid points are used (computed as ceil(roi_width / output_width), and likewise for height). Default is 0.

spatial_scale¶: Multiplicative spatial scale factor to translate ROI coordinates from their input spatial scale to the scale used when pooling, i.e., spatial scale of the input feature map X relative to the input image. E.g.; default is 1.0f.

class ONNXRound(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Round takes one input Tensor and rounds the values, element-wise, meaning it finds the nearest integer for each value. In case of halfs, the rule is to round them to the nearest even integer. The output tensor has the same shape and type as the input.

Examples: ` round([0.9]) = [1.0] round([2.5]) = [2.0] round([2.3]) = [2.0] round([1.5]) = [2.0] round([-4.5]) = [-4.0] `

Node Inputs

X (T, single) – Input tensor

Node Outputs

Y (T, single) – Output tensor

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXSVMClassifier(name, *, classlabels_ints=None, classlabels_strings=None, coefficients=None, kernel_params=None, kernel_type='LINEAR', post_transform='NONE', prob_a=None, prob_b=None, rho=None, support_vectors=None, vectors_per_class=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Support Vector Machine classifier

Node Inputs

X (T1, single) – Data to be classified.

Node Outputs

Y (T2, single) – Classification outputs (one class per example).
Z (Z_constraint, single) – Class scores (one per class per example), if prob_a and prob_b are provided they are probabilities for each class, otherwise they are raw scores.

Type Constraints

T1 – float32, float64, int64, int32
T2 – int64
Z_constraint – float32

Parameters

name – the name of the node.
classlabels_ints (Optional [List [int]], default=None) – Class labels if using integer labels. One and only one of the ‘classlabels_*’ attributes must be defined.
classlabels_strings (Optional [List [str]], default=None) – Class labels if using string labels. One and only one of the ‘classlabels_*’ attributes must be defined.
coefficients (Optional [List [float]], default=None) –
kernel_params (Optional [List [float]], default=None) – List of 3 elements containing gamma, coef0, and degree, in that order. Zero if unused for the kernel.
kernel_type (Optional [str], default=’LINEAR’) – The kernel type, one of ‘LINEAR,’ ‘POLY,’ ‘RBF,’ ‘SIGMOID’.
post_transform (Optional [str], default=’NONE’) – Indicates the transform to apply to the score. One of ‘NONE,’ ‘SOFTMAX,’ ‘LOGISTIC,’ ‘SOFTMAX_ZERO,’ or ‘PROBIT’
prob_a (Optional [List [float]], default=None) – First set of probability coefficients.
prob_b (Optional [List [float]], default=None) – Second set of probability coefficients. This array must be same size as prob_a. If these are provided then output Z are probability estimates, otherwise they are raw scores.
rho (Optional [List [float]], default=None) –
support_vectors (Optional [List [float]], default=None) –
vectors_per_class (Optional [List [int]], default=None) –

classlabels_ints¶: Class labels if using integer labels. One and only one of the ‘classlabels_*’ attributes must be defined.

classlabels_strings¶: Class labels if using string labels. One and only one of the ‘classlabels_*’ attributes must be defined.

coefficients¶: Object property of type list

kernel_params¶: List of 3 elements containing gamma, coef0, and degree, in that order. Zero if unused for the kernel.

kernel_type¶: The kernel type, one of ‘LINEAR,’ ‘POLY,’ ‘RBF,’ ‘SIGMOID’.

post_transform¶: Indicates the transform to apply to the score. One of ‘NONE,’ ‘SOFTMAX,’ ‘LOGISTIC,’ ‘SOFTMAX_ZERO,’ or ‘PROBIT’

prob_a¶: First set of probability coefficients.

prob_b¶: Second set of probability coefficients. This array must be same size as prob_a. If these are provided then output Z are probability estimates, otherwise they are raw scores.

rho¶: Object property of type list

support_vectors¶: Object property of type list

vectors_per_class¶: Object property of type list

class ONNXSVMRegressor(name, *, coefficients=None, kernel_params=None, kernel_type='LINEAR', n_supports=0, one_class=0, post_transform='NONE', rho=None, support_vectors=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Support Vector Machine regression prediction and one-class SVM anomaly detection.

Node Inputs

X (T, single) – Data to be regressed.

Node Outputs

Y (Y_constraint, single) – Regression outputs (one score per target per example).

Type Constraints

T – float32, float64, int64, int32
Y_constraint – float32

Parameters

name – the name of the node.
coefficients (Optional [List [float]], default=None) – Support vector coefficients.
kernel_params (Optional [List [float]], default=None) – List of 3 elements containing gamma, coef0, and degree, in that order. Zero if unused for the kernel.
kernel_type (Optional [str], default=’LINEAR’) – The kernel type, one of ‘LINEAR,’ ‘POLY,’ ‘RBF,’ ‘SIGMOID’.
n_supports (Optional [int], default=0) – The number of support vectors.
one_class (Optional [int], default=0) – Flag indicating whether the regression is a one-class SVM or not.
post_transform (Optional [str], default=’NONE’) – Indicates the transform to apply to the score. One of ‘NONE,’ ‘SOFTMAX,’ ‘LOGISTIC,’ ‘SOFTMAX_ZERO,’ or ‘PROBIT.’
rho (Optional [List [float]], default=None) –
support_vectors (Optional [List [float]], default=None) – Chosen support vectors

coefficients¶: Support vector coefficients.

kernel_params¶: List of 3 elements containing gamma, coef0, and degree, in that order. Zero if unused for the kernel.

kernel_type¶: The kernel type, one of ‘LINEAR,’ ‘POLY,’ ‘RBF,’ ‘SIGMOID’.

n_supports¶: The number of support vectors.

one_class¶: Flag indicating whether the regression is a one-class SVM or not.

post_transform¶: Indicates the transform to apply to the score. One of ‘NONE,’ ‘SOFTMAX,’ ‘LOGISTIC,’ ‘SOFTMAX_ZERO,’ or ‘PROBIT.’

rho¶: Object property of type list

support_vectors¶: Chosen support vectors

class ONNXScaler(name, *, offset=None, scale=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Rescale input data, for example to standardize features by removing the mean and scaling to unit variance.

Node Inputs

X (T, single) – Data to be scaled.

Node Outputs

Y (Y_constraint, single) – Scaled output data.

Type Constraints

T – float32, float64, int64, int32
Y_constraint – float32

Parameters

name – the name of the node.
offset (Optional [List [float]], default=None) – First, offset by this. Can be length of features in an [N,F] tensor or length 1, in which case it applies to all features, regardless of dimension count.
scale (Optional [List [float]], default=None) – Second, multiply by this. Can be length of features in an [N,F] tensor or length 1, in which case it applies to all features, regardless of dimension count. Must be same length as ‘offset’

offset¶: First, offset by this. Can be length of features in an [N,F] tensor or length 1, in which case it applies to all features, regardless of dimension count.

scale¶: Second, multiply by this. Can be length of features in an [N,F] tensor or length 1, in which case it applies to all features, regardless of dimension count. Must be same length as ‘offset’

class ONNXScatter(name, *, axis=0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

This operator is deprecated. Please use ScatterElements, which provides the same functionality.

Scatter takes three inputs data, updates, and indices of the same rank r >= 1 and an optional attribute axis that identifies an axis of data (by default, the outer-most axis, that is axis 0). The output of the operation is produced by creating a copy of the input data, and then updating its value to values specified by updates at specific index positions specified by indices. Its output shape is the same as the shape of data.

For each entry in updates, the target index in data is obtained by combining the corresponding entry in indices with the index of the entry itself: the index-value for dimension = axis is obtained from the value of the corresponding entry in indices and the index-value for dimension != axis is obtained from the index of the entry itself.

For instance, in a 2-D tensor case, the update corresponding to the [i][j] entry is performed as below: ```

output[indices[i][j]][j] = updates[i][j] if axis = 0, output[i][indices[i][j]] = updates[i][j] if axis = 1,

```

This operator is the inverse of GatherElements. It is similar to Torch’s Scatter operation.

Example 1: ```

data = [
[0.0, 0.0, 0.0], [0.0, 0.0, 0.0], [0.0, 0.0, 0.0],

] indices = [

[1, 0, 2], [0, 2, 1],

] updates = [

[1.0, 1.1, 1.2], [2.0, 2.1, 2.2],

] output = [

[2.0, 1.1, 0.0] [1.0, 0.0, 2.2] [0.0, 2.1, 1.2]

]

` Example 2: `

data = [[1.0, 2.0, 3.0, 4.0, 5.0]] indices = [[1, 3]] updates = [[1.1, 2.1]] axis = 1 output = [[1.0, 1.1, 3.0, 2.1, 5.0]]

```

Node Inputs

data (T, single) – Tensor of rank r >= 1.
indices (Tind, single) – Tensor of int32/int64 indices, of r >= 1 (same rank as input). All index values are expected to be within bounds [-s, s-1] along axis of size s. It is an error if any of the index values are out of bounds.
updates (T, single) – Tensor of rank r >=1 (same rank and shape as indices)

Node Outputs

output (T, single) – Tensor of rank r >= 1 (same rank as input).

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128
Tind – int32, int64

Parameters

name – the name of the node.
axis (Optional [int], default=0) – Which axis to scatter on. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(data).

axis¶: Which axis to scatter on. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(data).

class ONNXScatterElements(name, *, axis=0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

ScatterElements takes three inputs data, updates, and indices of the same rank r >= 1 and an optional attribute axis that identifies an axis of data (by default, the outer-most axis, that is axis 0). The output of the operation is produced by creating a copy of the input data, and then updating its value to values specified by updates at specific index positions specified by indices. Its output shape is the same as the shape of data.

For each entry in updates, the target index in data is obtained by combining the corresponding entry in indices with the index of the entry itself: the index-value for dimension = axis is obtained from the value of the corresponding entry in indices and the index-value for dimension != axis is obtained from the index of the entry itself.

For instance, in a 2-D tensor case, the update corresponding to the [i][j] entry is performed as below: ```

output[indices[i][j]][j] = updates[i][j] if axis = 0, output[i][indices[i][j]] = updates[i][j] if axis = 1,

```

This operator is the inverse of GatherElements. It is similar to Torch’s Scatter operation.

Example 1: ```

data = [
[0.0, 0.0, 0.0], [0.0, 0.0, 0.0], [0.0, 0.0, 0.0],

] indices = [

[1, 0, 2], [0, 2, 1],

] updates = [

[1.0, 1.1, 1.2], [2.0, 2.1, 2.2],

] output = [

[2.0, 1.1, 0.0] [1.0, 0.0, 2.2] [0.0, 2.1, 1.2]

]

` Example 2: `

data = [[1.0, 2.0, 3.0, 4.0, 5.0]] indices = [[1, 3]] updates = [[1.1, 2.1]] axis = 1 output = [[1.0, 1.1, 3.0, 2.1, 5.0]]

```

Node Inputs

data (T, single) – Tensor of rank r >= 1.
indices (Tind, single) – Tensor of int32/int64 indices, of r >= 1 (same rank as input). All index values are expected to be within bounds [-s, s-1] along axis of size s. It is an error if any of the index values are out of bounds.
updates (T, single) – Tensor of rank r >=1 (same rank and shape as indices)

Node Outputs

output (T, single) – Tensor of rank r >= 1 (same rank as input).

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128
Tind – int32, int64

Parameters

name – the name of the node.
axis (Optional [int], default=0) – Which axis to scatter on. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(data).

axis¶: Which axis to scatter on. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(data).

class ONNXScatterND(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

ScatterND takes three inputs data tensor of rank r >= 1, indices tensor of rank q >= 1, and updates tensor of rank q + r - indices.shape[-1] - 1. The output of the operation is produced by creating a copy of the input data, and then updating its value to values specified by updates at specific index positions specified by indices. Its output shape is the same as the shape of data. Note that indices should not have duplicate entries. That is, two or more updates for the same index-location is not supported.

indices is an integer tensor. Let k denote indices.shape[-1], the last dimension in the shape of indices.: indices is treated as a (q-1)-dimensional tensor of k-tuples, where each k-tuple is a partial-index into data.

Hence, k can be a value at most the rank of data. When k equals rank(data), each update entry specifies an update to a single element of the tensor. When k is less than rank(data) each update entry specifies an update to a slice of the tensor.

updates is treated as a (q-1)-dimensional tensor of replacement-slice-values. Thus, the first (q-1) dimensions of updates.shape must match the first (q-1) dimensions of indices.shape. The remaining dimensions of updates correspond to the dimensions of the replacement-slice-values. Each replacement-slice-value is a (r-k) dimensional tensor, corresponding to the trailing (r-k) dimensions of data. Thus, the shape of updates must equal indices.shape[0:q-1] ++ data.shape[k:r-1], where ++ denotes the concatenation of shapes.

The output is calculated via the following equation:

output = np.copy(data) update_indices = indices.shape[:-1] for idx in np.ndindex(update_indices):

output[indices[idx]] = updates[idx]

The order of iteration in the above loop is not specified. In particular, indices should not have duplicate entries: that is, if idx1 != idx2, then indices[idx1] != indices[idx2]. This ensures that the output value does not depend on the iteration order.

This operator is the inverse of GatherND.

Example 1: ```

data = [1, 2, 3, 4, 5, 6, 7, 8] indices = [[4], [3], [1], [7]] updates = [9, 10, 11, 12] output = [1, 11, 3, 10, 9, 6, 7, 12]

```

Example 2: ```

data = [[[1, 2, 3, 4], [5, 6, 7, 8], [8, 7, 6, 5], [4, 3, 2, 1]],
[[1, 2, 3, 4], [5, 6, 7, 8], [8, 7, 6, 5], [4, 3, 2, 1]], [[8, 7, 6, 5], [4, 3, 2, 1], [1, 2, 3, 4], [5, 6, 7, 8]], [[8, 7, 6, 5], [4, 3, 2, 1], [1, 2, 3, 4], [5, 6, 7, 8]]]

indices = [[0], [2]] updates = [[[5, 5, 5, 5], [6, 6, 6, 6], [7, 7, 7, 7], [8, 8, 8, 8]],

[[1, 1, 1, 1], [2, 2, 2, 2], [3, 3, 3, 3], [4, 4, 4, 4]]]

output = [[[5, 5, 5, 5], [6, 6, 6, 6], [7, 7, 7, 7], [8, 8, 8, 8]],
[[1, 2, 3, 4], [5, 6, 7, 8], [8, 7, 6, 5], [4, 3, 2, 1]], [[1, 1, 1, 1], [2, 2, 2, 2], [3, 3, 3, 3], [4, 4, 4, 4]], [[8, 7, 6, 5], [4, 3, 2, 1], [1, 2, 3, 4], [5, 6, 7, 8]]]

```

Node Inputs

data (T, single) – Tensor of rank r >= 1.
indices (indices_constraint, single) – Tensor of rank q >= 1.
updates (T, single) – Tensor of rank q + r - indices_shape[-1] - 1.

Node Outputs

output (T, single) – Tensor of rank r >= 1.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128
indices_constraint – int64

Parameters

name – the name of the node.

class ONNXSelu(name, *, alpha=1.6732631921768188, gamma=1.0507010221481323)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Selu takes one input data (Tensor<T>) and produces one output data (Tensor<T>) where the scaled exponential linear unit function, y = gamma * (alpha * e^x - alpha) for x <= 0, y = gamma * x for x > 0, is applied to the tensor elementwise.

Node Inputs

X (T, single) – Input tensor

Node Outputs

Y (T, single) – Output tensor

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.
alpha (Optional [float], default=1.6732631921768188) – Coefficient of SELU default to 1.67326319217681884765625 (i.e., float32 approximation of 1.6732632423543772848170429916717).
gamma (Optional [float], default=1.0507010221481323) – Coefficient of SELU default to 1.05070102214813232421875 (i.e., float32 approximation of 1.0507009873554804934193349852946).

alpha¶: Coefficient of SELU default to 1.67326319217681884765625 (i.e., float32 approximation of 1.6732632423543772848170429916717).

gamma¶: Coefficient of SELU default to 1.05070102214813232421875 (i.e., float32 approximation of 1.0507009873554804934193349852946).

class ONNXShape(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Takes a tensor as input and outputs an 1D int64 tensor containing the shape of the input tensor.

Node Inputs

data (T, single) – An input tensor.

Node Outputs

shape (T1, single) – Shape of the input tensor

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128
T1 – int64

Parameters

name – the name of the node.

class ONNXShrink(name, *, bias=0.0, lambd=0.5)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Shrink takes one input data (Tensor<numeric>) and produces one Tensor output, having same datatype and shape with input. It has two attributes, lambd and bias. The formula of this operator is: If x < -lambd, y = x + bias; If x > lambd, y = x - bias; Otherwise, y = 0.

Node Inputs

input (T, single) – The input data as Tensor.

Node Outputs

output (T, single) – The output.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64

Parameters

name – the name of the node.
bias (Optional [float], default=0.0) – The bias value added to output. Default is 0.
lambd (Optional [float], default=0.5) – The lambd value for the Shrink formulation. Default is 0.5.

bias¶: The bias value added to output. Default is 0.

lambd¶: The lambd value for the Shrink formulation. Default is 0.5.

class ONNXSigmoid(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Sigmoid takes one input data (Tensor<T>) and produces one output data (Tensor<T>) where the sigmoid function, y = 1 / (1 + exp(-x)), is applied to the tensor elementwise.

Node Inputs

X (T, single) – Input tensor

Node Outputs

Y (T, single) – Output tensor

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXSign(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Calculate the sign of the given input tensor element-wise. If input > 0, output 1. if input < 0, output -1. if input == 0, output 0.

Node Inputs

input (T, single) – Input tensor

Node Outputs

output (T, single) – The sign of the input tensor computed element-wise. It has the same shape and type of the input.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64

Parameters

name – the name of the node.

class ONNXSin(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Calculates the sine of the given input tensor, element-wise.

Node Inputs

input (T, single) – Input tensor

Node Outputs

output (T, single) – The sine of the input tensor computed element-wise

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXSinh(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Calculates the hyperbolic sine of the given input tensor element-wise.

Node Inputs

input (T, single) – Input tensor

Node Outputs

output (T, single) – The hyperbolic sine values of the input tensor computed element-wise

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXSize(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Takes a tensor as input and outputs a int64 scalar that equals to the total number of elements of the input tensor.

Node Inputs

data (T, single) – An input tensor.

Node Outputs

size (T1, single) – Total number of elements of the input tensor

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128
T1 – int64

Parameters

name – the name of the node.

class ONNXSlice(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Produces a slice of the input tensor along multiple axes. Similar to numpy: https://docs.scipy.org/doc/numpy/reference/arrays.indexing.html Slices uses starts, ends, axes and steps inputs to specify the start and end dimension and step for each axis in the list of axes, it uses this information to slice the input data tensor. If a negative value is passed for any of the start or end indices, it represents number of elements before the end of that dimension. If the value passed to start or end is larger than the n (the number of elements in this dimension), it represents n. For slicing to the end of a dimension with unknown size, it is recommended to pass in INT_MAX when sclicing forward and ‘INT_MIN’ when slicing backward. If a negative value is passed for step, it represents slicing backward. However step value cannot be 0. If axes are omitted, they are set to [0, …, ndim-1]. If steps are omitted, they are set to [1, …, 1] of length len(starts) Example 1:

data = [
[1, 2, 3, 4], [5, 6, 7, 8],

] axes = [0, 1] starts = [1, 0] ends = [2, 3] steps = [1, 2] result = [

[5, 7],

]

Example 2:

data = [: [1, 2, 3, 4], [5, 6, 7, 8],

] starts = [0, 1] ends = [-1, 1000] result = [

[2, 3, 4],

]

Node Inputs

data (T, single) – Tensor of data to extract slices from.
starts (Tind, single) – 1-D tensor of starting indices of corresponding axis in axes
ends (Tind, single) – 1-D tensor of ending indices (exclusive) of corresponding axis in axes
axes (Tind, optional) – 1-D tensor of axes that starts and ends apply to. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(data).
steps (Tind, optional) – 1-D tensor of slice step of corresponding axis in axes. Negative value means slicing backward. ‘steps’ cannot be 0. Defaults to 1.

Node Outputs

output (T, single) – Sliced data tensor.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128
Tind – int32, int64

Parameters

name – the name of the node.

class ONNXSoftmax(name, *, axis=1)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

The operator computes the softmax (normalized exponential) values for each layer in the batch: of the given input.

The input does not need to explicitly be a 2D vector; rather, it will be coerced into one. For an arbitrary n-dimensional tensor input in [a_0, a_1, …, a_{k-1}, a_k, …, a_{n-1}] and k is the axis provided, then input will be coerced into a 2-dimensional tensor with dimensions [a_0 * … * a_{k-1}, a_k * … * a_{n-1}]. For the default case where axis=1, this means the input tensor will be coerced into a 2D tensor of dimensions [a_0, a_1 * … * a_{n-1}], where a_0 is often the batch size. In this situation, we must have a_0 = N and a_1 * … * a_{n-1} = D. Each of these dimensions must be matched correctly, or else the operator will throw errors. The output tensor has the same shape and contains the softmax values of the corresponding input.

Node Inputs

input (T, single) – The input tensor that’s coerced into a 2D matrix of size (NxD) as described above.

Node Outputs

output (T, single) – The output values with the same shape as input tensor (the original size without coercion).

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.
axis (Optional [int], default=1) – Describes the axis of the inputs when coerced to 2D; defaults to one because the 0th axis most likely describes the batch_size. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(input).

axis¶: Describes the axis of the inputs when coerced to 2D; defaults to one because the 0th axis most likely describes the batch_size. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(input).

class ONNXSoftmaxCrossEntropyLoss(name, *, ignore_index=None, reduction='mean')¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Loss function that measures the softmax cross entropy between ‘scores’ and ‘labels’. This operator first computes a loss tensor whose shape is identical to the labels input. If the input is 2-D with shape (N, C), the loss tensor may be a N-element vector L = (l_1, l_2, …, l_N). If the input is N-D tensor with shape (N, C, D1, D2, …, Dk), the loss tensor L may have (N, D1, D2, …, Dk) as its shape and L[i,][j_1][j_2]…[j_k] denotes a scalar element in L. After L is available, this operator can optionally do a reduction operator.

shape(scores): (N, C) where C is the number of classes, or (N, C, D1, D2,…, Dk),: with K >= 1 in case of K-dimensional loss.
shape(labels): (N) where each value is 0 <= labels[i] <= C-1, or (N, D1, D2,…, Dk),: with K >= 1 in case of K-dimensional loss.
The loss for one sample, l_i, can caculated as follows:: l[i][d1][d2]…[dk] = -y[i][c][d1][d2]..[dk], where i is the index of classes.
or: l[i][d1][d2]…[dk] = -y[i][c][d1][d2]..[dk] * weights[c], if ‘weights’ is provided.
loss is zero for the case when label-value equals ignore_index.: l[i][d1][d2]…[dk] = 0, when labels[n][d1][d2]…[dk] = ignore_index
where:: p = Softmax(scores) y = Log(p) c = labels[i][d1][d2]…[dk]

Finally, L is optionally reduced: If reduction = ‘none’, the output is L with shape (N, D1, D2, …, Dk). If reduction = ‘sum’, the output is scalar: Sum(L). If reduction = ‘mean’, the output is scalar: ReduceMean(L), or if weight is provided: ReduceSum(L) / ReduceSum(W), where tensor W is of shape (N, D1, D2, …, Dk) and W[n][d1][d2]…[dk] = weights[labels[i][d1][d2]…[dk]].

Node Inputs

scores (T, single) – The predicted outputs with shape [batch_size, class_size], or [batch_size, class_size, D1, D2 , …, Dk], where K is the number of dimensions.
labels (Tind, single) – The ground truth output tensor, with shape [batch_size], or [batch_size, D1, D2, …, Dk], where K is the number of dimensions. Labels element value shall be in range of [0, C). If ignore_index is specified, it may have a value outside [0, C) and the label values should either be in the range [0, C) or have the value ignore_index.
weights (T, optional) – A manual rescaling weight given to each class. If given, it has to be a 1D Tensor assigning weight to each of the classes. Otherwise, it is treated as if having all ones.

Node Outputs

output (T, single) – Weighted loss float Tensor. If reduction is ‘none’, this has the shape of [batch_size], or [batch_size, D1, D2, …, Dk] in case of K-dimensional loss. Otherwise, it is a scalar.
log_prob (T, optional) – Log probability tensor. If the output of softmax is prob, its value is log(prob).

Type Constraints

T – float16, float32, float64
Tind – int32, int64

Parameters

name – the name of the node.
ignore_index (Optional [int], default=None) – Specifies a target value that is ignored and does not contribute to the input gradient. It’s an optional value.
reduction (Optional [str], default=’mean’) – Type of reduction to apply to loss: none, sum, mean(default). ‘none’: no reduction will be applied, ‘sum’: the output will be summed. ‘mean’: the sum of the output will be divided by the number of elements in the output.

ignore_index¶: Specifies a target value that is ignored and does not contribute to the input gradient. It’s an optional value.

reduction¶: Type of reduction to apply to loss: none, sum, mean(default). ‘none’: no reduction will be applied, ‘sum’: the output will be summed. ‘mean’: the sum of the output will be divided by the number of elements in the output.

class ONNXSoftplus(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Softplus takes one input data (Tensor<T>) and produces one output data (Tensor<T>) where the softplus function, y = ln(exp(x) + 1), is applied to the tensor elementwise.

Node Inputs

X (T, single) – 1D input tensor

Node Outputs

Y (T, single) – 1D input tensor

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXSoftsign(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Calculates the softsign (x/(1+|x|)) of the given input tensor element-wise.

Node Inputs

input (T, single) – Input tensor

Node Outputs

output (T, single) – The softsign (x/(1+|x|)) values of the input tensor computed element-wise

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXSpaceToDepth(name, *, blocksize)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

SpaceToDepth rearranges blocks of spatial data into depth. More specifically, this op outputs a copy of the input tensor where values from the height and width dimensions are moved to the depth dimension.

Node Inputs

input (T, single) – Input tensor of [N,C,H,W], where N is the batch axis, C is the channel or depth, H is the height and W is the width.

Node Outputs

output (T, single) – Output tensor of [N, C * blocksize * blocksize, H/blocksize, W/blocksize].

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128

Parameters

name – the name of the node.
blocksize (int) – Blocks of [blocksize, blocksize] are moved.

blocksize¶: Blocks of [blocksize, blocksize] are moved.

class ONNXSplit(name, *, axis=0, split=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Split a tensor into a list of tensors, along the specified ‘axis’. Lengths of the parts can be specified using argument ‘split’. Otherwise, the tensor is split to equal sized parts.

Node Inputs

input (T, single) – The tensor to split

Node Outputs

outputs (T, variadic) – One or more outputs forming list of tensors after splitting

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128

Parameters

name – the name of the node.
axis (Optional [int], default=0) – Which axis to split on. A negative value means counting dimensions from the back. Accepted range is [-rank, rank-1] where r = rank(input).
split (Optional [List [int]], default=None) – length of each output. Values should be >= 0.

axis¶: Which axis to split on. A negative value means counting dimensions from the back. Accepted range is [-rank, rank-1] where r = rank(input).

split¶: length of each output. Values should be >= 0.

class ONNXSqrt(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Square root takes one input data (Tensor<T>) and produces one output data (Tensor<T>) where the square root is, y = x^0.5, is applied to the tensor elementwise. If x is negative, then it will return NaN.

Node Inputs

X (T, single) – Input tensor

Node Outputs

Y (T, single) – Output tensor

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXSqueeze(name, *, axes=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Remove single-dimensional entries from the shape of a tensor. Takes a parameter axes with a list of axes to squeeze. If axes is not provided, all the single dimensions will be removed from the shape. If an axis is selected with shape entry not equal to one, an error is raised.

Node Inputs

data (T, single) – Tensors with at least max(dims) dimensions.

Node Outputs

squeezed (T, single) – Reshaped tensor with same data as input.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128

Parameters

name – the name of the node.
axes (Optional [List [int]], default=None) – List of integers indicating the dimensions to squeeze. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(data).

axes¶: List of integers indicating the dimensions to squeeze. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(data).

class ONNXSub(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Performs element-wise binary subtraction (with Numpy-style broadcasting support).

This operator supports multidirectional (i.e., Numpy-style) broadcasting; for more details please check [the doc](Broadcasting.md).

Node Inputs

A (T, single) – First operand.
B (T, single) – Second operand.

Node Outputs

C (T, single) – Result, has same element type as two inputs

Type Constraints

T – uint32, uint64, int32, int64, float16, float32, float64

Parameters

name – the name of the node.

class ONNXSum(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Element-wise sum of each of the input tensors (with Numpy-style broadcasting support). All inputs and outputs must have the same data type. This operator supports multidirectional (i.e., Numpy-style) broadcasting; for more details please check [the doc](Broadcasting.md).

Node Inputs

data_0 (T, variadic) – List of tensors for sum.

Node Outputs

sum (T, single) – Output tensor.

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXTan(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Calculates the tangent of the given input tensor, element-wise.

Node Inputs

input (T, single) – Input tensor

Node Outputs

output (T, single) – The tangent of the input tensor computed element-wise

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXTanh(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Calculates the hyperbolic tangent of the given input tensor element-wise.

Node Inputs

input (T, single) – Input tensor

Node Outputs

output (T, single) – The hyperbolic tangent values of the input tensor computed element-wise

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.

class ONNXTfIdfVectorizer(name, *, max_gram_length, max_skip_count, min_gram_length, mode, ngram_counts, ngram_indexes, pool_int64s=None, pool_strings=None, weights=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

This transform extracts n-grams from the input sequence and save them as a vector. Input can be either a 1-D or 2-D tensor. For 1-D input, output is the n-gram representation of that input. For 2-D input, the output is also a 2-D tensor whose i-th row is the n-gram representation of the i-th input row. More specifically, if input shape is [C], the corresponding output shape would be [max(ngram_indexes) + 1]. If input shape is [N, C], this operator produces a [N, max(ngram_indexes) + 1]-tensor.

In contrast to standard n-gram extraction, here, the indexes of extracting an n-gram from the original sequence are not necessarily consecutive numbers. The discontinuity between indexes are controlled by the number of skips. If the number of skips is 2, we should skip two tokens when scanning through the original sequence. Let’s consider an example. Assume that input sequence is [94, 17, 36, 12, 28] and the number of skips is 2. The associated 2-grams are [94, 12] and [17, 28] respectively indexed by [0, 3] and [1, 4]. If the number of skips becomes 0, the 2-grams generated are [94, 17], [17, 36], [36, 12], [12, 28] indexed by [0, 1], [1, 2], [2, 3], [3, 4], respectively.

The output vector (denoted by Y) stores the count of each n-gram; Y[ngram_indexes[i]] indicates the times that the i-th n-gram is found. The attribute ngram_indexes is used to determine the mapping between index i and the corresponding n-gram’s output coordinate. If pool_int64s is [94, 17, 17, 36], ngram_indexes is [1, 0], ngram_counts=[0, 0], then the Y[0] (first element in Y) and Y[1] (second element in Y) are the counts of [17, 36] and [94, 17], respectively. An n-gram which cannot be found in pool_strings/pool_int64s should be ignored and has no effect on the output. Note that we may consider all skips up to S when generating the n-grams.

The examples used above are true if mode is “TF”. If mode is “IDF”, all the counts larger than 1 would be truncated to 1 and the i-th element in weights would be used to scale (by multiplication) the count of the i-th n-gram in pool. If mode is “TFIDF”, this operator first computes the counts of all n-grams and then scale them by the associated values in the weights attribute.

Only one of pool_strings and pool_int64s can be set. If pool_int64s is set, the input should be an integer tensor. If pool_strings is set, the input must be a string tensor.

Node Inputs

X (T, single) – Input for n-gram extraction

Node Outputs

Y (T1, single) – Ngram results

Type Constraints

T – int32, int64
T1 – float32

Parameters

name – the name of the node.
max_gram_length (int) – Maximum n-gram length. If this value is 3, 3-grams will be used to generate the output.
max_skip_count (int) – Maximum number of items (integers/strings) to be skipped when constructing an n-gram from X. If max_skip_count=1, min_gram_length=2, max_gram_length=3, this operator may generate 2-grams with skip_count=0 and skip_count=1, and 3-grams with skip_count=0 and skip_count=1
min_gram_length (int) – Minimum n-gram length. If this value is 2 and max_gram_length is 3, output may contain counts of 2-grams and 3-grams.
mode (str) – The weighting criteria. It can be one of “TF” (term frequency), “IDF” (inverse document frequency), and “TFIDF” (the combination of TF and IDF)
ngram_counts (List [int]) – The starting indexes of 1-grams, 2-grams, and so on in pool. It is useful when determining the boundary between two consecutive collections of n-grams. For example, if ngram_counts is [0, 17, 36], the first index (zero-based) of 1-gram/2-gram/3-gram in pool are 0/17/36. This format is essentially identical to CSR (or CSC) sparse matrix format, and we choose to use this due to its popularity.
ngram_indexes (List [int]) – list of int64s (type: AttributeProto::INTS). This list is parallel to the specified ‘pool_*’ attribute. The i-th element in ngram_indexes indicate the coordinate of the i-th n-gram in the output tensor.
pool_int64s (Optional [List [int]], default=None) – List of int64 n-grams learned from the training set. Either this or pool_strings attributes must be present but not both. It’s an 1-D tensor starting with the collections of all 1-grams and ending with the collections of n-grams. The i-th element in pool stores the n-gram that should be mapped to coordinate ngram_indexes[i] in the output vector.
pool_strings (Optional [List [str]], default=None) – List of strings n-grams learned from the training set. Either this or pool_int64s attributes must be present but not both. It’s an 1-D tensor starting with the collections of all 1-grams and ending with the collections of n-grams. The i-th element in pool stores the n-gram that should be mapped to coordinate ngram_indexes[i] in the output vector.
weights (Optional [List [float]], default=None) – list of floats. This attribute stores the weight of each n-gram in pool. The i-th element in weights is the weight of the i-th n-gram in pool. Its length equals to the size of ngram_indexes. By default, weights is an all-one tensor.This attribute is used when mode is “IDF” or “TFIDF” to scale the associated word counts.

max_gram_length¶: Maximum n-gram length. If this value is 3, 3-grams will be used to generate the output.

max_skip_count¶: Maximum number of items (integers/strings) to be skipped when constructing an n-gram from X. If max_skip_count=1, min_gram_length=2, max_gram_length=3, this operator may generate 2-grams with skip_count=0 and skip_count=1, and 3-grams with skip_count=0 and skip_count=1

min_gram_length¶: Minimum n-gram length. If this value is 2 and max_gram_length is 3, output may contain counts of 2-grams and 3-grams.

mode¶: The weighting criteria. It can be one of “TF” (term frequency), “IDF” (inverse document frequency), and “TFIDF” (the combination of TF and IDF)

ngram_counts¶: The starting indexes of 1-grams, 2-grams, and so on in pool. It is useful when determining the boundary between two consecutive collections of n-grams. For example, if ngram_counts is [0, 17, 36], the first index (zero-based) of 1-gram/2-gram/3-gram in pool are 0/17/36. This format is essentially identical to CSR (or CSC) sparse matrix format, and we choose to use this due to its popularity.

ngram_indexes¶: list of int64s (type: AttributeProto::INTS). This list is parallel to the specified ‘pool_*’ attribute. The i-th element in ngram_indexes indicate the coordinate of the i-th n-gram in the output tensor.

pool_int64s¶: List of int64 n-grams learned from the training set. Either this or pool_strings attributes must be present but not both. It’s an 1-D tensor starting with the collections of all 1-grams and ending with the collections of n-grams. The i-th element in pool stores the n-gram that should be mapped to coordinate ngram_indexes[i] in the output vector.

pool_strings¶: List of strings n-grams learned from the training set. Either this or pool_int64s attributes must be present but not both. It’s an 1-D tensor starting with the collections of all 1-grams and ending with the collections of n-grams. The i-th element in pool stores the n-gram that should be mapped to coordinate ngram_indexes[i] in the output vector.

weights¶: list of floats. This attribute stores the weight of each n-gram in pool. The i-th element in weights is the weight of the i-th n-gram in pool. Its length equals to the size of ngram_indexes. By default, weights is an all-one tensor.This attribute is used when mode is “IDF” or “TFIDF” to scale the associated word counts.

class ONNXThresholdedRelu(name, *, alpha=1.0)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

ThresholdedRelu takes one input data (Tensor<T>) and produces one output data (Tensor<T>) where the rectified linear function, y = x for x > alpha, y = 0 otherwise, is applied to the tensor elementwise.

Node Inputs

X (T, single) – Input tensor

Node Outputs

Y (T, single) – Output tensor

Type Constraints

T – float16, float32, float64

Parameters

name – the name of the node.
alpha (Optional [float], default=1.0) – Threshold value

alpha¶: Threshold value

class ONNXTile(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Constructs a tensor by tiling a given tensor. This is the same as function tile in Numpy, but no broadcast. For example A = [[1, 2], [3, 4]], B = [1, 2], tile(A, B) = [[1, 2, 1, 2], [3, 4, 3, 4]]

Node Inputs

input (T, single) – Input tensor of any shape.
repeats (T1, single) – 1D int64 tensor of the same length as input’s dimension number, includes numbers of repeated copies along input’s dimensions.

Node Outputs

output (T, single) – Output tensor of the same dimension and type as tensor input. output_dim[i] = input_dim[i] * repeats[i]

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128
T1 – int64

Parameters

name – the name of the node.

class ONNXTopK(name, *, axis=- 1, largest=1, sorted=1)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Retrieve the top-K largest or smallest elements along a specified axis. Given an input tensor of shape [a_1, a_2, …, a_n, r] and integer argument k, return two outputs:

-Value tensor of shape [a_1, a_2, …, a_{axis-1}, k, a_{axis+1}, … a_n]
which contains the values of the top k elements along the specified axis

-Index tensor of shape [a_1, a_2, …, a_{axis-1}, k, a_{axis+1}, … a_n] which
contains the indices of the top k elements (original indices from the input tensor).

If “largest” is 1 (the default value) then the k largest elements are returned. If “sorted” is 1 (the default value) then the resulting k elements will be sorted. If “sorted” is 0, order of returned ‘Values’ and ‘Indices’ are undefined.

Given two equivalent values, this operator uses the indices along the axis as: a tiebreaker. That is, the element with the lower index will appear first.

Node Inputs

X (T, single) – Tensor of shape [a_1, a_2, …, a_n, r]
K (K_constraint, single) – A 1-D tensor containing a single positive value corresponding to the number of top elements to retrieve

Node Outputs

Values (T, single) – Tensor of shape [a_1, a_2, …, a_{axis-1}, k, a_{axis+1}, … a_n] containing top K values from the input tensor
Indices (I, single) – Tensor of shape [a_1, a_2, …, a_{axis-1}, k, a_{axis+1}, … a_n] containing the corresponding input tensor indices for the top K values.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64
I – int64
K_constraint – int64

Parameters

name – the name of the node.
axis (Optional [int], default=-1) – Dimension on which to do the sort. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(input).
largest (Optional [int], default=1) – Whether to return the top-K largest or smallest elements.
sorted (Optional [int], default=1) – Whether to return the elements in sorted order.

axis¶: Dimension on which to do the sort. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(input).

largest¶: Whether to return the top-K largest or smallest elements.

sorted¶: Whether to return the elements in sorted order.

class ONNXTranspose(name, *, perm=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Transpose the input tensor similar to numpy.transpose. For example, when perm=(1, 0, 2), given an input tensor of shape (1, 2, 3), the output shape will be (2, 1, 3).

Node Inputs

data (T, single) – An input tensor.

Node Outputs

transposed (T, single) – Transposed output.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128

Parameters

name – the name of the node.
perm (Optional [List [int]], default=None) – A list of integers. By default, reverse the dimensions, otherwise permute the axes according to the values given.

perm¶: A list of integers. By default, reverse the dimensions, otherwise permute the axes according to the values given.

class ONNXTreeEnsembleClassifier(name, *, base_values=None, class_ids=None, class_nodeids=None, class_treeids=None, class_weights=None, classlabels_int64s=None, classlabels_strings=None, nodes_falsenodeids=None, nodes_featureids=None, nodes_hitrates=None, nodes_missing_value_tracks_true=None, nodes_modes=None, nodes_nodeids=None, nodes_treeids=None, nodes_truenodeids=None, nodes_values=None, post_transform='NONE')¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Tree Ensemble classifier. Returns the top class for each of N inputs. The attributes named ‘nodes_X’ form a sequence of tuples, associated by index into the sequences, which must all be of equal length. These tuples define the nodes. Similarly, all fields prefixed with ‘class_’ are tuples of votes at the leaves. A leaf may have multiple votes, where each vote is weighted by the associated class_weights index. One and only one of classlabels_strings or classlabels_int64s will be defined. The class_ids are indices into this list.

Node Inputs

X (T1, single) – Input of shape [N,F]

Node Outputs

Y (T2, single) – N, Top class for each point
Z (Z_constraint, single) – The class score for each class, for each point, a tensor of shape [N,E].

Type Constraints

T1 – float32, float64, int64, int32
T2 – int64
Z_constraint – float32

Parameters

name – the name of the node.
base_values (Optional [List [float]], default=None) – Base values for classification, added to final class score; the size must be the same as the classes or can be left unassigned (assumed 0)
class_ids (Optional [List [int]], default=None) – The index of the class list that each weight is for.
class_nodeids (Optional [List [int]], default=None) – node id that this weight is for.
class_treeids (Optional [List [int]], default=None) – The id of the tree that this node is in.
class_weights (Optional [List [float]], default=None) – The weight for the class in class_id.
classlabels_int64s (Optional [List [int]], default=None) – Class labels if using integer labels. One and only one of the ‘classlabels_*’ attributes must be defined.
classlabels_strings (Optional [List [str]], default=None) – Class labels if using string labels. One and only one of the ‘classlabels_*’ attributes must be defined.
nodes_falsenodeids (Optional [List [int]], default=None) – Child node if expression is false.
nodes_featureids (Optional [List [int]], default=None) – Feature id for each node.
nodes_hitrates (Optional [List [float]], default=None) – Popularity of each node, used for performance and may be omitted.
nodes_missing_value_tracks_true (Optional [List [int]], default=None) – For each node, define what to do in the presence of a missing value: if a value is missing (NaN), use the ‘true’ or ‘false’ branch based on the value in this array. This attribute may be left undefined, and the defalt value is false (0) for all nodes.
nodes_modes (Optional [List [str]], default=None) – The node kind, that is, the comparison to make at the node. There is no comparison to make at a leaf node. One of ‘BRANCH_LEQ’, ‘BRANCH_LT’, ‘BRANCH_GTE’, ‘BRANCH_GT’, ‘BRANCH_EQ’, ‘BRANCH_NEQ’, ‘LEAF’
nodes_nodeids (Optional [List [int]], default=None) – Node id for each node. Ids may restart at zero for each tree, but it not required to.
nodes_treeids (Optional [List [int]], default=None) – Tree id for each node.
nodes_truenodeids (Optional [List [int]], default=None) – Child node if expression is true.
nodes_values (Optional [List [float]], default=None) – Thresholds to do the splitting on for each node.
post_transform (Optional [str], default=’NONE’) – Indicates the transform to apply to the score. One of ‘NONE,’ ‘SOFTMAX,’ ‘LOGISTIC,’ ‘SOFTMAX_ZERO,’ or ‘PROBIT.’

base_values¶: Base values for classification, added to final class score; the size must be the same as the classes or can be left unassigned (assumed 0)

class_ids¶: The index of the class list that each weight is for.

class_nodeids¶: node id that this weight is for.

class_treeids¶: The id of the tree that this node is in.

class_weights¶: The weight for the class in class_id.

classlabels_int64s¶: Class labels if using integer labels. One and only one of the ‘classlabels_*’ attributes must be defined.

classlabels_strings¶: Class labels if using string labels. One and only one of the ‘classlabels_*’ attributes must be defined.

nodes_falsenodeids¶: Child node if expression is false.

nodes_featureids¶: Feature id for each node.

nodes_hitrates¶: Popularity of each node, used for performance and may be omitted.

nodes_missing_value_tracks_true¶: For each node, define what to do in the presence of a missing value: if a value is missing (NaN), use the ‘true’ or ‘false’ branch based on the value in this array. This attribute may be left undefined, and the defalt value is false (0) for all nodes.

nodes_modes¶: The node kind, that is, the comparison to make at the node. There is no comparison to make at a leaf node. One of ‘BRANCH_LEQ’, ‘BRANCH_LT’, ‘BRANCH_GTE’, ‘BRANCH_GT’, ‘BRANCH_EQ’, ‘BRANCH_NEQ’, ‘LEAF’

nodes_nodeids¶: Node id for each node. Ids may restart at zero for each tree, but it not required to.

nodes_treeids¶: Tree id for each node.

nodes_truenodeids¶: Child node if expression is true.

nodes_values¶: Thresholds to do the splitting on for each node.

post_transform¶: Indicates the transform to apply to the score. One of ‘NONE,’ ‘SOFTMAX,’ ‘LOGISTIC,’ ‘SOFTMAX_ZERO,’ or ‘PROBIT.’

class ONNXTreeEnsembleRegressor(name, *, aggregate_function='SUM', base_values=None, n_targets=None, nodes_falsenodeids=None, nodes_featureids=None, nodes_hitrates=None, nodes_missing_value_tracks_true=None, nodes_modes=None, nodes_nodeids=None, nodes_treeids=None, nodes_truenodeids=None, nodes_values=None, post_transform='NONE', target_ids=None, target_nodeids=None, target_treeids=None, target_weights=None)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Tree Ensemble regressor. Returns the regressed values for each input in N. All args with nodes_ are fields of a tuple of tree nodes, and it is assumed they are the same length, and an index i will decode the tuple across these inputs. Each node id can appear only once for each tree id. All fields prefixed with target_ are tuples of votes at the leaves. A leaf may have multiple votes, where each vote is weighted by the associated target_weights index. All trees must have their node ids start at 0 and increment by 1. Mode enum is BRANCH_LEQ, BRANCH_LT, BRANCH_GTE, BRANCH_GT, BRANCH_EQ, BRANCH_NEQ, LEAF

Node Inputs

X (T, single) – Input of shape [N,F]

Node Outputs

Y (Y_constraint, single) – N classes

Type Constraints

T – float32, float64, int64, int32
Y_constraint – float32

Parameters

name – the name of the node.
aggregate_function (Optional [str], default=’SUM’) – Defines how to aggregate leaf values within a target. One of ‘AVERAGE,’ ‘SUM,’ ‘MIN,’ ‘MAX.’
base_values (Optional [List [float]], default=None) – Base values for classification, added to final class score; the size must be the same as the classes or can be left unassigned (assumed 0)
n_targets (Optional [int], default=None) – The total number of targets.
nodes_falsenodeids (Optional [List [int]], default=None) – Child node if expression is false
nodes_featureids (Optional [List [int]], default=None) – Feature id for each node.
nodes_hitrates (Optional [List [float]], default=None) – Popularity of each node, used for performance and may be omitted.
nodes_missing_value_tracks_true (Optional [List [int]], default=None) – For each node, define what to do in the presence of a NaN: use the ‘true’ (if the attribute value is 1) or ‘false’ (if the attribute value is 0) branch based on the value in this array. This attribute may be left undefined and the defalt value is false (0) for all nodes.
nodes_modes (Optional [List [str]], default=None) – The node kind, that is, the comparison to make at the node. There is no comparison to make at a leaf node. One of ‘BRANCH_LEQ’, ‘BRANCH_LT’, ‘BRANCH_GTE’, ‘BRANCH_GT’, ‘BRANCH_EQ’, ‘BRANCH_NEQ’, ‘LEAF’
nodes_nodeids (Optional [List [int]], default=None) – Node id for each node. Node ids must restart at zero for each tree and increase sequentially.
nodes_treeids (Optional [List [int]], default=None) – Tree id for each node.
nodes_truenodeids (Optional [List [int]], default=None) – Child node if expression is true
nodes_values (Optional [List [float]], default=None) – Thresholds to do the splitting on for each node.
post_transform (Optional [str], default=’NONE’) – Indicates the transform to apply to the score. One of ‘NONE,’ ‘SOFTMAX,’ ‘LOGISTIC,’ ‘SOFTMAX_ZERO,’ or ‘PROBIT’
target_ids (Optional [List [int]], default=None) – The index of the target that each weight is for
target_nodeids (Optional [List [int]], default=None) – The node id of each weight
target_treeids (Optional [List [int]], default=None) – The id of the tree that each node is in.
target_weights (Optional [List [float]], default=None) – The weight for each target

aggregate_function¶: Defines how to aggregate leaf values within a target. One of ‘AVERAGE,’ ‘SUM,’ ‘MIN,’ ‘MAX.’

base_values¶: Base values for classification, added to final class score; the size must be the same as the classes or can be left unassigned (assumed 0)

n_targets¶: The total number of targets.

nodes_falsenodeids¶: Child node if expression is false

nodes_featureids¶: Feature id for each node.

nodes_hitrates¶: Popularity of each node, used for performance and may be omitted.

nodes_missing_value_tracks_true¶: For each node, define what to do in the presence of a NaN: use the ‘true’ (if the attribute value is 1) or ‘false’ (if the attribute value is 0) branch based on the value in this array. This attribute may be left undefined and the defalt value is false (0) for all nodes.

nodes_modes¶: The node kind, that is, the comparison to make at the node. There is no comparison to make at a leaf node. One of ‘BRANCH_LEQ’, ‘BRANCH_LT’, ‘BRANCH_GTE’, ‘BRANCH_GT’, ‘BRANCH_EQ’, ‘BRANCH_NEQ’, ‘LEAF’

nodes_nodeids¶: Node id for each node. Node ids must restart at zero for each tree and increase sequentially.

nodes_treeids¶: Tree id for each node.

nodes_truenodeids¶: Child node if expression is true

nodes_values¶: Thresholds to do the splitting on for each node.

post_transform¶: Indicates the transform to apply to the score. One of ‘NONE,’ ‘SOFTMAX,’ ‘LOGISTIC,’ ‘SOFTMAX_ZERO,’ or ‘PROBIT’

target_ids¶: The index of the target that each weight is for

target_nodeids¶: The node id of each weight

target_treeids¶: The id of the tree that each node is in.

target_weights¶: The weight for each target

class ONNXUnique(name, *, axis=None, sorted=1)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Find the unique elements of a tensor. When an optional attribute ‘axis’ is provided, unique subtensors sliced along the ‘axis’ are returned. Otherwise the input tensor is flattened and unique values of the flattened tensor are returned.

This operator returns the unique values or sliced unique subtensors of the input tensor and three optional outputs. The first output tensor ‘Y’ contains all unique values or subtensors of the input. The second optional output tensor ‘indices’ contains indices of ‘Y’ elements’ first occurance in ‘X’.. The third optional output tensor ‘inverse_indices’ contains, for elements of ‘X’, its corresponding indices in ‘Y’. “. The fourth optional output tensor ‘counts’ contains the count of each element of ‘Y’ in the input.

Outputs are either sorted in ascending order or optionally in the order of the first occurrence of the values in the input.

https://docs.scipy.org/doc/numpy/reference/generated/numpy.unique.html

Example 1:

input_X = [2, 1, 1, 3, 4, 3] attribute_sorted = 0 attribute_axis = None output_Y = [2, 1, 3, 4] output_indices = [0, 1, 3, 4] output_inverse_indices = [0, 1, 1, 2, 3, 2] output_counts = [1, 2, 2, 1]

Example 2:

input_X = [[1, 3], [2, 3]] attribute_sorted = 1 attribute_axis = None output_Y = [1, 2, 3] output_indices = [0, 2, 1] output_inverse_indices = [0, 2, 1, 2] output_counts = [1, 1, 2]

Example 3:

input_X = [[1, 0, 0], [1, 0, 0], [2, 3, 4]] attribute_sorted = 1 attribute_axis = 0 output_Y = [[1, 0, 0], [2, 3, 4]] output_indices = [0, 2] output_inverse_indices = [0, 0, 1] output_counts = [2, 1]

Example 4:

input_x = [[[1., 1.], [0., 1.], [2., 1.], [0., 1.]],: [[1., 1.], [0., 1.], [2., 1.], [0., 1.]]]

attribute_sorted = 1 attribute_axis = 1

intermediate data are presented below for better understanding:

there are 4 subtensors sliced along axis 1 of input_x (shape = (2, 4, 2)): A: [[1, 1], [1, 1]],

[[0, 1], [0, 1]], [[2, 1], [2, 1]], [[0, 1], [0, 1]].

there are 3 unique subtensors: [[1, 1], [1, 1]], [[0, 1], [0, 1]], [[2, 1], [2, 1]].

sorted unique subtensors: B: [[0, 1], [0, 1]],

[[1, 1], [1, 1]], [[2, 1], [2, 1]].

output_Y is constructed from B: [[[0. 1.], [1. 1.], [2. 1.]],

[[0. 1.], [1. 1.], [2. 1.]]]

output_indices is to map from B to A: [1, 0, 2]

output_inverse_indices is to map from A to B: [1, 0, 2, 0]

output_counts = [2 1 1]

Node Inputs

X (T, single) – A N-D input tensor that is to be processed.

Node Outputs

Y (T, single) – A tensor of the same type as ‘X’ containing all the unique values or subtensors sliced along a provided ‘axis’ in ‘X’, either sorted or maintained in the same order they occur in input ‘X’
indices (indices_constraint, optional) – A 1-D INT64 tensor containing indices of ‘Y’ elements’ first occurance in ‘X’. When ‘axis’ is provided, it contains indices to subtensors in input ‘X’ on the ‘axis’. When ‘axis’ is not provided, it contains indices to values in the flattened input tensor.
inverse_indices (inverse_indices_constraint, optional) – A 1-D INT64 tensor containing, for elements of ‘X’, its corresponding indices in ‘Y’. When ‘axis’ is provided, it contains indices to subtensors in output ‘Y’ on the ‘axis’. When ‘axis’ is not provided, it contains indices to values in output ‘Y’.
counts (counts_constraint, optional) – A 1-D INT64 tensor containing the count of each element of ‘Y’ in input ‘X’

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128
indices_constraint – int64
inverse_indices_constraint – int64
counts_constraint – int64

Parameters

name – the name of the node.
axis (Optional [int], default=None) – (Optional) The dimension to apply unique. If not specified, the unique elements of the flattened input are returned. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(input).
sorted (Optional [int], default=1) – (Optional) Whether to sort the unique elements in ascending order before returning as output. Must be one of 0, or 1 (default).

axis¶: (Optional) The dimension to apply unique. If not specified, the unique elements of the flattened input are returned. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(input).

sorted¶: (Optional) Whether to sort the unique elements in ascending order before returning as output. Must be one of 0, or 1 (default).

class ONNXUnsqueeze(name, *, axes)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Insert single-dimensional entries to the shape of an input tensor (data). Takes one required argument axes - which contains a list of dimension indices and this operator will insert a dimension of value 1 into the corresponding index of the output tensor (expanded).

For example:: Given an input tensor (data) of shape [3, 4, 5], then Unsqueeze(data, axes=[0, 4]) outputs a tensor (expanded) containing same data as data but with shape [1, 3, 4, 5, 1].

The attribute axes should not contain any duplicate entries. It is an error if it contains duplicates. The rank of the output tensor (output_rank) is the rank of the input tensor (data) plus the number of values in axes. Each value in axes should be within the (inclusive) range [-output_rank , output_rank - 1]. The order of values in axes does not matter and can come in any order.

Node Inputs

data (T, single) – Original tensor

Node Outputs

expanded (T, single) – Reshaped tensor with same data as input.

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128

Parameters

name – the name of the node.
axes (List [int]) – List of integers indicating the dimensions to be inserted. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(expanded).

axes¶: List of integers indicating the dimensions to be inserted. Negative value means counting dimensions from the back. Accepted range is [-r, r-1] where r = rank(expanded).

class ONNXUpsample(name, *, mode='nearest')¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Upsample the input tensor. Each dimension value of the output tensor is:

output_dimension = floor(input_dimension * scale).

Node Inputs

X (T, single) – N-D tensor
scales (scales_constraint, single) – The scale array along each dimension. It takes value greater than or equal to 1. The number of elements of ‘scales’ should be the same as the rank of input ‘X’.

Node Outputs

Y (T, single) – N-D tensor after resizing

Type Constraints

T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128
scales_constraint – float32

Parameters

name – the name of the node.
mode (Optional [str], default=’nearest’) – Two interpolation modes: nearest (default), and linear (including bilinear, trilinear, etc)

mode¶: Two interpolation modes: nearest (default), and linear (including bilinear, trilinear, etc)

class ONNXWhere(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Return elements, either from X or Y, depending on condition (with Numpy-style broadcasting support). Where behaves like numpy.where with three parameters: https://docs.scipy.org/doc/numpy/reference/generated/numpy.where.html

Node Inputs

condition (B, single) – When True (nonzero), yield X, otherwise yield Y
X (T, single) – values selected at indices where condition is True
Y (T, single) – values selected at indices where condition is False

Node Outputs

output (T, single) – Tensor of shape equal to the broadcasted shape of condition, X, and Y.

Type Constraints

B – bool_
T – uint8, uint16, uint32, uint64, int8, int16, int32, int64, float16, float32, float64, bool_, complex64, complex128

Parameters

name – the name of the node.

class ONNXXor(name, *)¶

Bases: daceml.onnx.nodes.onnx_op.ONNXOp

Returns the tensor resulted from performing the xor logical operation elementwise on the input tensors A and B (with Numpy-style broadcasting support).

This operator supports multidirectional (i.e., Numpy-style) broadcasting; for more details please check [the doc](Broadcasting.md).

Node Inputs

A (T, single) – First input operand for the logical operator.
B (T, single) – Second input operand for the logical operator.

Node Outputs

C (T1, single) – Result tensor.

Type Constraints

T – bool_
T1 – bool_

Parameters

name – the name of the node.