concrete.ml.quantization.quantized_ops
module concrete.ml.quantization.quantized_ops
concrete.ml.quantization.quantized_ops
Quantized versions of the ONNX operators for post training quantization.
class QuantizedSigmoid
QuantizedSigmoid
Quantized sigmoid op.
class QuantizedHardSigmoid
QuantizedHardSigmoid
Quantized HardSigmoid op.
class QuantizedRelu
QuantizedRelu
Quantized Relu op.
class QuantizedPRelu
QuantizedPRelu
Quantized PRelu op.
class QuantizedLeakyRelu
QuantizedLeakyRelu
Quantized LeakyRelu op.
class QuantizedHardSwish
QuantizedHardSwish
Quantized Hardswish op.
class QuantizedElu
QuantizedElu
Quantized Elu op.
class QuantizedSelu
QuantizedSelu
Quantized Selu op.
class QuantizedCelu
QuantizedCelu
Quantized Celu op.
class QuantizedClip
QuantizedClip
Quantized clip op.
class QuantizedRound
QuantizedRound
Quantized round op.
class QuantizedPow
QuantizedPow
Quantized pow op.
Only works for a float constant power. This operation will be fused to a (potentially larger) TLU.
method __init__
__init__
method can_fuse
can_fuse
Determine if this op can be fused.
Power raising can be fused and computed in float when a single integer tensor generates both the operands. For example in the formula: f(x) = x ** (x + 1) where x is an integer tensor.
Returns:
bool
: Can fuse
class QuantizedGemm
QuantizedGemm
Quantized Gemm op.
method __init__
__init__
method can_fuse
can_fuse
Determine if this op can be fused.
Gemm operation can not be fused since it must be performed over integer tensors and it combines different values of the input tensors.
Returns:
bool
: False, this operation can not be fused as it adds different encrypted integers
method q_impl
q_impl
class QuantizedMatMul
QuantizedMatMul
Quantized MatMul op.
method __init__
__init__
method can_fuse
can_fuse
Determine if this op can be fused.
Gemm operation can not be fused since it must be performed over integer tensors and it combines different values of the input tensors.
Returns:
bool
: False, this operation can not be fused as it adds different encrypted integers
method q_impl
q_impl
class QuantizedAdd
QuantizedAdd
Quantized Addition operator.
Can add either two variables (both encrypted) or a variable and a constant
method can_fuse
can_fuse
Determine if this op can be fused.
Add operation can be computed in float and fused if it operates over inputs produced by a single integer tensor. For example the expression x + x * 1.75, where x is an encrypted tensor, can be computed with a single TLU.
Returns:
bool
: Whether the number of integer input tensors allows computing this op as a TLU
method q_impl
q_impl
class QuantizedTanh
QuantizedTanh
Quantized Tanh op.
class QuantizedSoftplus
QuantizedSoftplus
Quantized Softplus op.
class QuantizedExp
QuantizedExp
Quantized Exp op.
class QuantizedLog
QuantizedLog
Quantized Log op.
class QuantizedAbs
QuantizedAbs
Quantized Abs op.
class QuantizedIdentity
QuantizedIdentity
Quantized Identity op.
method q_impl
q_impl
class QuantizedReshape
QuantizedReshape
Quantized Reshape op.
method q_impl
q_impl
Reshape the input integer encrypted tensor.
Args:
q_inputs
: an encrypted integer tensor at index 0 and one constant shape at index 1attrs
: additional optional reshape options
Returns:
result
(QuantizedArray): reshaped encrypted integer tensor
class QuantizedConv
QuantizedConv
Quantized Conv op.
method __init__
__init__
Construct the quantized convolution operator and retrieve parameters.
Args:
n_bits_output
: number of bits for the quantization of the outputs of this operatorint_input_names
: names of integer tensors that are taken as input for this operationconstant_inputs
: the weights and activationsinput_quant_opts
: options for the input quantizerattrs
: convolution optionsdilations
(Tuple[int]): dilation of the kernel, default 1 on all dimensions.group
(int): number of convolution groups, default 1kernel_shape
(Tuple[int]): shape of the kernel. Should have 2 elements for 2d convpads
(Tuple[int]): padding in ONNX format (begin, end) on each axisstrides
(Tuple[int]): stride of the convolution on each axis
method can_fuse
can_fuse
Determine if this op can be fused.
Conv operation can not be fused since it must be performed over integer tensors and it combines different elements of the input tensors.
Returns:
bool
: False, this operation can not be fused as it adds different encrypted integers
method q_impl
q_impl
Compute the quantized convolution between two quantized tensors.
Allows an optional quantized bias.
Args:
q_inputs
: input tuple, containsx
(numpy.ndarray): input data. Shape is N x C x H x W for 2dw
(numpy.ndarray): weights tensor. Shape is (O x I x Kh x Kw) for 2db
(numpy.ndarray, Optional): bias tensor, Shape is (O,)attrs
: convolution options handled in constructor
Returns:
res
(QuantizedArray): result of the quantized integer convolution
class QuantizedAvgPool
QuantizedAvgPool
Quantized Average Pooling op.
method __init__
__init__
method can_fuse
can_fuse
Determine if this op can be fused.
Avg Pooling operation can not be fused since it must be performed over integer tensors and it combines different elements of the input tensors.
Returns:
bool
: False, this operation can not be fused as it adds different encrypted integers
method q_impl
q_impl
class QuantizedPad
QuantizedPad
Quantized Padding op.
method __init__
__init__
method can_fuse
can_fuse
Determine if this op can be fused.
Pad operation can not be fused since it must be performed over integer tensors.
Returns:
bool
: False, this operation can not be fused as it is manipulates integer tensors
class QuantizedWhere
QuantizedWhere
Where operator on quantized arrays.
Supports only constants for the results produced on the True/False branches.
method __init__
__init__
class QuantizedCast
QuantizedCast
Cast the input to the required data type.
In FHE we only support a limited number of output types. Booleans are cast to integers.
class QuantizedGreater
QuantizedGreater
Comparison operator >.
Only supports comparison with a constant.
method __init__
__init__
class QuantizedGreaterOrEqual
QuantizedGreaterOrEqual
Comparison operator >=.
Only supports comparison with a constant.
method __init__
__init__
class QuantizedLess
QuantizedLess
Comparison operator <.
Only supports comparison with a constant.
method __init__
__init__
class QuantizedLessOrEqual
QuantizedLessOrEqual
Comparison operator <=.
Only supports comparison with a constant.
method __init__
__init__
class QuantizedOr
QuantizedOr
Or operator ||.
This operation is not really working as a quantized operation. It just works when things got fused, as in e.g. Act(x) = x || (x + 42))
method __init__
__init__
method can_fuse
can_fuse
Determine if this op can be fused.
Or can be fused and computed in float when a single integer tensor generates both the operands. For example in the formula: f(x) = x || (x + 1) where x is an integer tensor.
Returns:
bool
: Can fuse
class QuantizedDiv
QuantizedDiv
Div operator /.
This operation is not really working as a quantized operation. It just works when things got fused, as in e.g. Act(x) = 1000 / (x + 42))
method __init__
__init__
method can_fuse
can_fuse
Determine if this op can be fused.
Div can be fused and computed in float when a single integer tensor generates both the operands. For example in the formula: f(x) = x / (x + 1) where x is an integer tensor.
Returns:
bool
: Can fuse
class QuantizedMul
QuantizedMul
Multiplication operator.
Only multiplies an encrypted tensor with a float constant for now. This operation will be fused to a (potentially larger) TLU.
method __init__
__init__
method can_fuse
can_fuse
Determine if this op can be fused.
Multiplication can be fused and computed in float when a single integer tensor generates both the operands. For example in the formula: f(x) = x * (x + 1) where x is an integer tensor.
Returns:
bool
: Can fuse
class QuantizedSub
QuantizedSub
Subtraction operator.
This works the same as addition on both encrypted - encrypted and on encrypted - constant.
method can_fuse
can_fuse
Determine if this op can be fused.
Add operation can be computed in float and fused if it operates over inputs produced by a single integer tensor. For example the expression x + x * 1.75, where x is an encrypted tensor, can be computed with a single TLU.
Returns:
bool
: Whether the number of integer input tensors allows computing this op as a TLU
method q_impl
q_impl
class QuantizedBatchNormalization
QuantizedBatchNormalization
Quantized Batch normalization with encrypted input and in-the-clear normalization params.
class QuantizedFlatten
QuantizedFlatten
Quantized flatten for encrypted inputs.
method can_fuse
can_fuse
Determine if this op can be fused.
Flatten operation can not be fused since it must be performed over integer tensors.
Returns:
bool
: False, this operation can not be fused as it is manipulates integer tensors.
method q_impl
q_impl
Flatten the input integer encrypted tensor.
Args:
q_inputs
: an encrypted integer tensor at index 0attrs
: contains axis attribute
Returns:
result
(QuantizedArray): reshaped encrypted integer tensor
class QuantizedReduceSum
QuantizedReduceSum
ReduceSum with encrypted input.
This operator is currently an experimental feature.
method __init__
__init__
Construct the quantized ReduceSum operator and retrieve parameters.
Args:
n_bits_output
(int): Number of bits for the operator's quantization of outputs.int_input_names
(Optional[Set[str]]): Names of input integer tensors. Default to None.constant_inputs
(Optional[Dict]): Input constant tensor.axes
(Optional[numpy.ndarray]): Array of integers along which to reduce. The default is to reduce over all the dimensions of the input tensor if 'noop_with_empty_axes' is false, else act as an Identity op when 'noop_with_empty_axes' is true. Accepted range is [-r, r-1] where r = rank(data). Default to None.input_quant_opts
(Optional[QuantizationOptions]): Options for the input quantizer. Default to None.attrs
(dict): RecuseSum options.keepdims
(int): Keep the reduced dimension or not, 1 means keeping the input dimension, 0 will reduce it along the given axis. Default to 1.noop_with_empty_axes
(int): Defines behavior if 'axes' is empty or set to None. Default behavior with 0 is to reduce all axes. When axes is empty and this attribute is set to true 1, input tensor will not be reduced, and the output tensor would be equivalent to input tensor. Default to 0.
method calibrate
calibrate
Create corresponding QuantizedArray for the output of the activation function.
Args:
*inputs (numpy.ndarray)
: Calibration sample inputs.
Returns:
numpy.ndarray
: the output values for the provided calibration samples.
method q_impl
q_impl
Sum the encrypted tensor's values over axis 1.
Args:
q_inputs
(QuantizedArray): An encrypted integer tensor at index 0.attrs
(Dict): Contains axis attribute.
Returns:
(QuantizedArray)
: The sum of all values along axis 1 as an encrypted integer tensor.
method tree_sum
tree_sum
Large sum without overflow (only MSB remains).
Args:
input_qarray
: Enctyped integer tensor.is_calibration
: Whether we are calibrating the tree sum. If so, it will create all the quantizers for the downscaling.
Returns:
(numpy.ndarray)
: The MSB (based on the precision self.n_bits) of the integers sum.
class QuantizedErf
QuantizedErf
Quantized erf op.
class QuantizedNot
QuantizedNot
Quantized Not op.
class QuantizedBrevitasQuant
QuantizedBrevitasQuant
Brevitas uniform quantization with encrypted input.
method __init__
__init__
Construct the Brevitas quantization operator.
Args:
n_bits_output
(int): Number of bits for the operator's quantization of outputs. Not used, will be overridden by the bit_width in ONNXint_input_names
(Optional[Set[str]]): Names of input integer tensors. Default to None.constant_inputs
(Optional[Dict]): Input constant tensor.scale
(float): Quantizer scalezero_point
(float): Quantizer zero-pointbit_width
(int): Number of bits of the integer representationinput_quant_opts
(Optional[QuantizationOptions]): Options for the input quantizer. Default to None. attrs (dict):rounding_mode
(str): Rounding mode (default and only accepted option is "ROUND")signed
(int): Whether this op quantizes to signed integers (default 1),narrow
(int): Whether this op quantizes to a narrow range of integers e.g. [-2n_bits-1 .. 2n_bits-1] (default 0),
method q_impl
q_impl
Quantize values.
Args:
q_inputs
: an encrypted integer tensor at index 0 and one constant shape at index 1attrs
: additional optional reshape options
Returns:
result
(QuantizedArray): reshaped encrypted integer tensor
class QuantizedTranspose
QuantizedTranspose
Transpose operator for quantized inputs.
This operator performs quantization, transposes the encrypted data, then dequantizes again.
method q_impl
q_impl
Reshape the input integer encrypted tensor.
Args:
q_inputs
: an encrypted integer tensor at index 0 and one constant shape at index 1attrs
: additional optional reshape options
Returns:
result
(QuantizedArray): reshaped encrypted integer tensor
Last updated