mirror of
https://github.com/zama-ai/concrete.git
synced 2026-02-09 03:55:04 -05:00
7901d8868c
ONNX spec:
"kernel_shape : list of ints
The shape of the convolution kernel. If not present, should be
inferred from input W."
We were taking the number of input/output feature maps into account,
which we realized we should not.
484 lines
12 KiB
Python
484 lines
12 KiB
Python
"""
|
|
Tests of execution of convolution operation.
|
|
"""
|
|
|
|
import numpy as np
|
|
import pytest
|
|
|
|
import concrete.numpy as cnp
|
|
import concrete.onnx as connx
|
|
from concrete.numpy.representation.node import Node
|
|
from concrete.numpy.tracing.tracer import Tracer
|
|
|
|
|
|
@pytest.mark.parametrize(
|
|
"input_shape,weight_shape, group",
|
|
[
|
|
pytest.param(
|
|
(1, 1, 4, 4),
|
|
(1, 1, 2, 2),
|
|
1,
|
|
),
|
|
pytest.param(
|
|
(4, 3, 4, 4),
|
|
(2, 3, 2, 2),
|
|
1,
|
|
),
|
|
pytest.param(
|
|
(1, 6, 4, 4),
|
|
(6, 1, 2, 2),
|
|
6,
|
|
),
|
|
],
|
|
)
|
|
@pytest.mark.parametrize(
|
|
"strides",
|
|
[
|
|
(2, 2),
|
|
],
|
|
)
|
|
@pytest.mark.parametrize(
|
|
"dilations",
|
|
[
|
|
(1, 1),
|
|
],
|
|
)
|
|
@pytest.mark.parametrize(
|
|
"has_bias",
|
|
[
|
|
True,
|
|
False,
|
|
],
|
|
)
|
|
def test_conv2d(input_shape, weight_shape, group, strides, dilations, has_bias, helpers):
|
|
"""
|
|
Test conv2d.
|
|
"""
|
|
|
|
configuration = helpers.configuration()
|
|
|
|
weight = np.random.randint(0, 4, size=weight_shape)
|
|
|
|
if has_bias:
|
|
bias = np.random.randint(0, 4, size=(weight_shape[0],))
|
|
else:
|
|
bias = None
|
|
|
|
@cnp.compiler({"x": "encrypted"})
|
|
def function(x):
|
|
return connx.conv(x, weight, bias, strides=strides, dilations=dilations, group=group)
|
|
|
|
inputset = [np.random.randint(0, 4, size=input_shape) for i in range(100)]
|
|
circuit = function.compile(inputset, configuration)
|
|
|
|
sample = np.random.randint(0, 4, size=input_shape)
|
|
helpers.check_execution(circuit, function, sample)
|
|
|
|
|
|
@pytest.mark.parametrize(
|
|
"input_shape,weight_shape,bias_shape,pads,strides,dilations,kernel_shape,group,auto_pad,"
|
|
"expected_error,expected_message",
|
|
[
|
|
pytest.param(
|
|
(1, 1, 4, 4),
|
|
(1, 1, 2, 2),
|
|
(1,),
|
|
(0, 0, 0, 0),
|
|
(1, 1),
|
|
(1, 1),
|
|
None,
|
|
1,
|
|
"VALID",
|
|
ValueError,
|
|
"auto_pad should be in {'NOTSET'}, but got 'VALID'",
|
|
),
|
|
pytest.param(
|
|
(1, 1, 4),
|
|
(1, 1, 2),
|
|
(1,),
|
|
(),
|
|
(1,),
|
|
(1,),
|
|
None,
|
|
1,
|
|
"NOTSET",
|
|
ValueError,
|
|
"pads should be of form (D_begin_pad, D_end_pad) when performing 1D convolution, "
|
|
"but it's ()",
|
|
),
|
|
pytest.param(
|
|
(1, 1, 4),
|
|
(1, 1, 2),
|
|
(1,),
|
|
(0, 0),
|
|
(),
|
|
(1,),
|
|
None,
|
|
1,
|
|
"NOTSET",
|
|
ValueError,
|
|
"strides should be of form (D_stride,) when performing 1D " "convolution, but it's ()",
|
|
),
|
|
pytest.param(
|
|
(1, 1, 4),
|
|
(1, 1, 2),
|
|
(1,),
|
|
(0, 0),
|
|
(1,),
|
|
(),
|
|
None,
|
|
1,
|
|
"NOTSET",
|
|
ValueError,
|
|
"dilations should be of form (D_dilation,) when performing 1D "
|
|
"convolution, but it's ()",
|
|
),
|
|
pytest.param(
|
|
(1, 1, 4, 4),
|
|
(1, 1, 2, 2),
|
|
(1,),
|
|
(),
|
|
(1, 1),
|
|
(1, 1),
|
|
None,
|
|
1,
|
|
"NOTSET",
|
|
ValueError,
|
|
"pads should be of form (height_begin_pad, width_begin_pad, "
|
|
"height_end_pad, width_end_pad) when performing 2D convolution, "
|
|
"but it's ()",
|
|
),
|
|
pytest.param(
|
|
(1, 1, 4, 4),
|
|
(1, 1, 2, 2),
|
|
(1,),
|
|
(0, 0, 0, 0),
|
|
(),
|
|
(1, 1),
|
|
None,
|
|
1,
|
|
"NOTSET",
|
|
ValueError,
|
|
"strides should be of form (height_stride, width_stride) when performing 2D "
|
|
"convolution, but it's ()",
|
|
),
|
|
pytest.param(
|
|
(1, 1, 4, 4),
|
|
(1, 1, 2, 2),
|
|
(1,),
|
|
(0, 0, 0, 0),
|
|
(1, 1),
|
|
(),
|
|
None,
|
|
1,
|
|
"NOTSET",
|
|
ValueError,
|
|
"dilations should be of form (height_dilation, width_dilation) when performing 2D "
|
|
"convolution, but it's ()",
|
|
),
|
|
pytest.param(
|
|
(1, 1, 4, 4, 4),
|
|
(1, 1, 2, 2, 4),
|
|
(1,),
|
|
(),
|
|
(1, 1, 1),
|
|
(1, 1, 1),
|
|
None,
|
|
1,
|
|
"NOTSET",
|
|
ValueError,
|
|
"pads should be of form (D_begin_pad, height_begin_pad, width_begin_pad, "
|
|
"D_end_pad, height_end_pad, width_end_pad) when performing 3D convolution, "
|
|
"but it's ()",
|
|
),
|
|
pytest.param(
|
|
(1, 1, 4, 4, 4),
|
|
(1, 1, 2, 2, 2),
|
|
(1,),
|
|
(0, 0, 0, 0, 0, 0),
|
|
(),
|
|
(1, 1, 1),
|
|
None,
|
|
1,
|
|
"NOTSET",
|
|
ValueError,
|
|
"strides should be of form (D_stride, height_stride, width_stride) when performing 3D "
|
|
"convolution, but it's ()",
|
|
),
|
|
pytest.param(
|
|
(1, 1, 4, 4, 4),
|
|
(1, 1, 2, 2, 2),
|
|
(1,),
|
|
(0, 0, 0, 0, 0, 0),
|
|
(1, 1, 1),
|
|
(),
|
|
None,
|
|
1,
|
|
"NOTSET",
|
|
ValueError,
|
|
"dilations should be of form (D_dilation, height_dilation, width_dilation) when "
|
|
"performing 3D convolution, but it's ()",
|
|
),
|
|
pytest.param(
|
|
(),
|
|
(1, 1, 2, 2),
|
|
(1,),
|
|
(0, 0, 0, 0),
|
|
(1, 1),
|
|
(1, 1),
|
|
None,
|
|
1,
|
|
"NOTSET",
|
|
ValueError,
|
|
"expected input x to have at least 3 dimensions (N, C, D1, ...), but got 0",
|
|
),
|
|
pytest.param(
|
|
(1, 1, 4, 4),
|
|
(),
|
|
(1,),
|
|
(0, 0, 0, 0),
|
|
(1, 1),
|
|
(1, 1),
|
|
None,
|
|
1,
|
|
"NOTSET",
|
|
ValueError,
|
|
"expected weight to have at least 3 dimensions (F, C / group, K1, ...), but got 0",
|
|
),
|
|
pytest.param(
|
|
(1, 1, 4, 4),
|
|
(1, 1, 2, 2),
|
|
(),
|
|
(0, 0, 0, 0),
|
|
(1, 1),
|
|
(1, 1),
|
|
None,
|
|
1,
|
|
"NOTSET",
|
|
ValueError,
|
|
"expected bias to have a single dimension (F,), but got 0",
|
|
),
|
|
pytest.param(
|
|
(1, 1, 4, 4),
|
|
(1, 1, 2, 2),
|
|
(1,),
|
|
(0, 0, 0, 0),
|
|
(1, 1),
|
|
(1, 1),
|
|
(1, 2),
|
|
1,
|
|
"NOTSET",
|
|
ValueError,
|
|
"expected kernel_shape to be (2, 2), but got (1, 2)",
|
|
),
|
|
pytest.param(
|
|
(1, 1, 4, 4),
|
|
(1, 1, 2, 2),
|
|
(1,),
|
|
(0, 0, 0, 0),
|
|
(1, 1),
|
|
(1, 1),
|
|
None,
|
|
None,
|
|
"NOTSET",
|
|
ValueError,
|
|
"expected group to be an integer > 0, but got None",
|
|
),
|
|
pytest.param(
|
|
(1, 1, 4, 4),
|
|
(1, 2, 2, 2),
|
|
(1,),
|
|
(0, 0, 0, 0),
|
|
(1, 1),
|
|
(1, 1),
|
|
None,
|
|
1,
|
|
"NOTSET",
|
|
ValueError,
|
|
"expected number of channel in weight to be 1.0 (C / group), but got 2",
|
|
),
|
|
pytest.param(
|
|
(1, 1, 4),
|
|
(1, 1, 2),
|
|
(1,),
|
|
(0, 0),
|
|
(1,),
|
|
(1,),
|
|
None,
|
|
1,
|
|
"NOTSET",
|
|
NotImplementedError,
|
|
"conv1d conversion to MLIR is not yet implemented",
|
|
),
|
|
pytest.param(
|
|
(1, 1, 4, 4, 4),
|
|
(1, 1, 2, 2, 2),
|
|
(1,),
|
|
(0, 0, 0, 0, 0, 0),
|
|
(1, 1, 1),
|
|
(1, 1, 1),
|
|
None,
|
|
1,
|
|
"NOTSET",
|
|
NotImplementedError,
|
|
"conv3d conversion to MLIR is not yet implemented",
|
|
),
|
|
pytest.param(
|
|
(1, 1, 4, 4, 4, 4),
|
|
(1, 1, 2, 2, 2, 2),
|
|
(1,),
|
|
(0, 0, 0, 0, 0, 0, 0, 0),
|
|
(1, 1, 1, 1),
|
|
(1, 1, 1, 1),
|
|
None,
|
|
1,
|
|
"NOTSET",
|
|
NotImplementedError,
|
|
"only 1D, 2D, and 3D convolutions are supported",
|
|
),
|
|
pytest.param(
|
|
(1, 2, 4, 4),
|
|
(1, 1, 2, 2),
|
|
(1,),
|
|
(0, 0, 0, 0),
|
|
(1, 1),
|
|
(1, 1),
|
|
None,
|
|
2,
|
|
"NOTSET",
|
|
ValueError,
|
|
"expected number of feature maps (1) to be a multiple of group (2)",
|
|
),
|
|
],
|
|
)
|
|
# pylint: disable=too-many-arguments
|
|
def test_bad_conv_compilation(
|
|
input_shape,
|
|
weight_shape,
|
|
bias_shape,
|
|
pads,
|
|
strides,
|
|
dilations,
|
|
kernel_shape,
|
|
group,
|
|
auto_pad,
|
|
expected_error,
|
|
expected_message,
|
|
helpers,
|
|
):
|
|
# pylint: enable=too-many-arguments
|
|
"""
|
|
Test conv with bad parameters.
|
|
"""
|
|
|
|
configuration = helpers.configuration()
|
|
|
|
weight = np.random.randint(0, 4, size=weight_shape)
|
|
|
|
if bias_shape is not None:
|
|
bias = np.random.randint(0, 4, size=bias_shape)
|
|
else:
|
|
bias = None
|
|
|
|
@cnp.compiler({"x": "encrypted"})
|
|
def function(x):
|
|
return connx.conv(
|
|
x,
|
|
weight,
|
|
bias=bias,
|
|
pads=pads,
|
|
strides=strides,
|
|
dilations=dilations,
|
|
kernel_shape=kernel_shape,
|
|
group=group,
|
|
auto_pad=auto_pad,
|
|
)
|
|
|
|
inputset = [np.random.randint(0, 4, size=input_shape) for i in range(100)]
|
|
with pytest.raises(expected_error) as excinfo:
|
|
function.compile(inputset, configuration)
|
|
|
|
assert str(excinfo.value) == expected_message
|
|
|
|
|
|
@pytest.mark.parametrize(
|
|
"conv_func_name",
|
|
[
|
|
"conv",
|
|
223,
|
|
None,
|
|
],
|
|
)
|
|
@pytest.mark.parametrize(
|
|
"func",
|
|
[
|
|
# pylint: disable=protected-access
|
|
connx.convolution._evaluate_conv,
|
|
connx.convolution._trace_conv,
|
|
# pylint: enable=protected-access
|
|
],
|
|
)
|
|
def test_bad_conv_func_name(conv_func_name, func):
|
|
"""
|
|
Test invalid conv function name.
|
|
"""
|
|
with pytest.raises(AssertionError) as excinfo:
|
|
func(None, None, None, None, None, None, None, conv_func_name)
|
|
assert (
|
|
str(excinfo.value) == f"expected conv_func to be one of ['conv1d', 'conv2d', 'conv3d'], "
|
|
f"but got {conv_func_name}"
|
|
)
|
|
|
|
|
|
@pytest.mark.parametrize(
|
|
"x,weight,bias,expected_error,expected_message",
|
|
[
|
|
pytest.param(
|
|
np.array([1, 2, 3]),
|
|
"not same type as x",
|
|
None,
|
|
TypeError,
|
|
"expected weight to be of same type as x",
|
|
),
|
|
pytest.param(
|
|
np.array([1, 2, 3]),
|
|
np.array([1, 2, 3]),
|
|
"not same type as x",
|
|
TypeError,
|
|
"expected bias to be of same type as x",
|
|
),
|
|
pytest.param(
|
|
Tracer(Node.constant(np.array([1, 2, 3])), []),
|
|
"not same type as x",
|
|
None,
|
|
TypeError,
|
|
"expected weight to be of type Tracer or ndarray",
|
|
),
|
|
pytest.param(
|
|
Tracer(Node.constant(np.array([1, 2, 3])), []),
|
|
np.array([1, 2, 3]),
|
|
"not same type as x",
|
|
TypeError,
|
|
"expected bias to be of type Tracer or ndarray",
|
|
),
|
|
],
|
|
)
|
|
def test_inconsistent_input_types(
|
|
x,
|
|
weight,
|
|
bias,
|
|
expected_error,
|
|
expected_message,
|
|
):
|
|
"""
|
|
Test conv with inconsistent input types.
|
|
"""
|
|
with pytest.raises(expected_error) as excinfo:
|
|
connx.conv(
|
|
x,
|
|
weight,
|
|
bias=bias,
|
|
)
|
|
|
|
assert str(excinfo.value) == expected_message
|