Skip to content

N
ngraph
Commit 10e383bc authored by Adam Procter's avatar Adam Procter
Browse files
Options

Refactor dyn slice test to improve compile time

parent dc4a2143
Hide whitespace changes
Inline Side-by-side
Showing with 141 additions and 85 deletions
test/dyn_slice_test.in.cpp
View file @ 10e383bc
This source diff could not be displayed because it is too large. You can view the blob instead.
test/ref_generators/generate_dyn_slice_ref.py
View file @ 10e383bc
......@@ -256,92 +256,55 @@ class SliceTestWriter:
try:
data_out = data_in.__getitem__(slices)
except Exception as e:
self._stream.write(' {\n')
self._stream.write(' auto arg = std::make_shared<op::Parameter>(%s, %s);\n' % (np_dt_to_ng(self._dtype), print_shape(self._shape)))
self._stream.write(' auto lb = std::make_shared<op::Parameter>(element::i64, %s);\n' % print_shape((n_slices,)))
self._stream.write(' auto ub = std::make_shared<op::Parameter>(element::i64, %s);\n' % print_shape((n_slices,)))
self._stream.write(' auto strides = std::make_shared<op::Parameter>(element::i64, %s);\n' % print_shape((n_slices,)))
self._stream.write('\n')
self._stream.write(' std::vector<%s> input_values(%s);\n' % (np_dt_to_c(self._dtype), np.prod(self._shape)))
self._stream.write(' std::iota(input_values.begin(), input_values.end(), static_cast<%s>(0));\n' % np_dt_to_c(self._dtype))
self._stream.write(' std::vector<int64_t> lb_values{%s};\n' % print_lb_values(slices))
self._stream.write(' std::vector<int64_t> ub_values{%s};\n' % print_ub_values(slices))
self._stream.write(' std::vector<int64_t> strides_values{%s};\n' % print_stride_values(slices))
self._stream.write(' AxisSet lb_mask{%s};\n' % print_lb_mask_axes(slices))
self._stream.write(' AxisSet ub_mask{%s};\n' % print_ub_mask_axes(slices))
self._stream.write(' AxisSet new_mask{%s};\n' % print_new_mask_axes(slices))
self._stream.write(' AxisSet shrink_mask{%s};\n' % print_shrink_mask_axes(slices))
self._stream.write(' AxisSet ellipsis_mask{%s};\n' % print_ellipsis_mask_axes(slices))
self._stream.write('\n')
self._stream.write(' // numpy threw: %s\n' % type(e))
self._stream.write(' EXPECT_ANY_THROW({\n');
self._stream.write(' auto slice = std::make_shared<op::DynSlice>(arg, lb, ub, strides, lb_mask, ub_mask, new_mask, shrink_mask, ellipsis_mask);\n')
self._stream.write('\n')
self._stream.write(' auto f = std::make_shared<Function>(NodeVector{slice}, ParameterVector{arg, lb, ub, strides});\n')
self._stream.write('\n')
self._stream.write(' auto backend = runtime::Backend::create("${BACKEND_NAME}",true);\n')
self._stream.write(' auto ex = backend->compile(f);\n')
self._stream.write('\n')
self._stream.write(' auto input_arg = backend->create_tensor(%s, %s);\n' % (np_dt_to_ng(self._dtype), print_shape(self._shape)))
self._stream.write(' auto input_lb = backend->create_tensor(element::i64, %s);\n' % print_shape((n_slices,)))
self._stream.write(' auto input_ub = backend->create_tensor(element::i64, %s);\n' % print_shape((n_slices,)))
self._stream.write(' auto input_strides = backend->create_tensor(element::i64, %s);\n' % print_shape((n_slices,)))
self._stream.write(' copy_data(input_arg, input_values);\n')
self._stream.write(' copy_data(input_lb, lb_values);\n')
self._stream.write(' copy_data(input_ub, ub_values);\n')
self._stream.write(' copy_data(input_strides, strides_values);\n')
self._stream.write('\n')
self._stream.write(' auto output = backend->create_dynamic_tensor(%s, PartialShape::dynamic());\n' % np_dt_to_ng(self._dtype))
self._stream.write('\n')
self._stream.write(' ex->call({output}, {input_arg, input_lb, input_ub, input_strides});\n')
self._stream.write(' });\n')
self._stream.write(' }\n')
self._stream.write(' check_failure<%s>\n'
' (%s,\n'
' %s,\n'
' std::vector<int64_t>{%s},\n'
' std::vector<int64_t>{%s},\n'
' std::vector<int64_t>{%s},\n'
' AxisSet{%s},\n'
' AxisSet{%s},\n'
' AxisSet{%s},\n'
' AxisSet{%s},\n'
' AxisSet{%s});\n'
% (np_dt_to_c(self._dtype),
np_dt_to_ng(self._dtype),
print_shape(data_in.shape),
print_lb_values(slices),
print_ub_values(slices),
print_stride_values(slices),
print_lb_mask_axes(slices),
print_ub_mask_axes(slices),
print_new_mask_axes(slices),
print_shrink_mask_axes(slices),
print_ellipsis_mask_axes(slices)))
else:
self._stream.write(' {\n')
self._stream.write(' auto arg = std::make_shared<op::Parameter>(%s, %s);\n' % (np_dt_to_ng(self._dtype), print_shape(self._shape)))
self._stream.write(' auto lb = std::make_shared<op::Parameter>(element::i64, %s);\n' % print_shape((n_slices,)))
self._stream.write(' auto ub = std::make_shared<op::Parameter>(element::i64, %s);\n' % print_shape((n_slices,)))
self._stream.write(' auto strides = std::make_shared<op::Parameter>(element::i64, %s);\n' % print_shape((n_slices,)))
self._stream.write('\n')
self._stream.write(' std::vector<%s> input_values(%s);\n' % (np_dt_to_c(self._dtype), np.prod(self._shape)))
self._stream.write(' std::iota(input_values.begin(), input_values.end(), static_cast<%s>(0));\n' % np_dt_to_c(self._dtype))
self._stream.write(' std::vector<int64_t> lb_values{%s};\n' % print_lb_values(slices))
self._stream.write(' std::vector<int64_t> ub_values{%s};\n' % print_ub_values(slices))
self._stream.write(' std::vector<int64_t> strides_values{%s};\n' % print_stride_values(slices))
self._stream.write(' AxisSet lb_mask{%s};\n' % print_lb_mask_axes(slices))
self._stream.write(' AxisSet ub_mask{%s};\n' % print_ub_mask_axes(slices))
self._stream.write(' AxisSet new_mask{%s};\n' % print_new_mask_axes(slices))
self._stream.write(' AxisSet shrink_mask{%s};\n' % print_shrink_mask_axes(slices))
self._stream.write(' AxisSet ellipsis_mask{%s};\n' % print_ellipsis_mask_axes(slices))
self._stream.write('\n')
self._stream.write(' auto slice = std::make_shared<op::DynSlice>(arg, lb, ub, strides, lb_mask, ub_mask, new_mask, shrink_mask, ellipsis_mask);\n')
self._stream.write('\n')
self._stream.write(' auto f = std::make_shared<Function>(NodeVector{slice}, ParameterVector{arg, lb, ub, strides});\n')
self._stream.write('\n')
self._stream.write(' auto backend = runtime::Backend::create("${BACKEND_NAME}",true);\n')
self._stream.write(' auto ex = backend->compile(f);\n')
self._stream.write('\n')
self._stream.write(' auto input_arg = backend->create_tensor(%s, %s);\n' % (np_dt_to_ng(self._dtype), print_shape(self._shape)))
self._stream.write(' auto input_lb = backend->create_tensor(element::i64, %s);\n' % print_shape((n_slices,)))
self._stream.write(' auto input_ub = backend->create_tensor(element::i64, %s);\n' % print_shape((n_slices,)))
self._stream.write(' auto input_strides = backend->create_tensor(element::i64, %s);\n' % print_shape((n_slices,)))
self._stream.write(' copy_data(input_arg, input_values);\n')
self._stream.write(' copy_data(input_lb, lb_values);\n')
self._stream.write(' copy_data(input_ub, ub_values);\n')
self._stream.write(' copy_data(input_strides, strides_values);\n')
self._stream.write('\n')
self._stream.write(' auto output = backend->create_dynamic_tensor(%s, PartialShape::dynamic());\n' % np_dt_to_ng(self._dtype))
self._stream.write('\n')
self._stream.write(' ex->call({output}, {input_arg, input_lb, input_ub, input_strides});\n')
self._stream.write('\n')
self._stream.write(' EXPECT_EQ(output->get_element_type(), (%s));\n' % np_dt_to_ng(self._dtype))
self._stream.write(' EXPECT_EQ(output->get_shape(), (%s));\n' % print_shape(data_out.shape))
self._stream.write('\n')
self._stream.write(' auto output_values = read_vector<%s>(output);\n' % np_dt_to_c(self._dtype))
self._stream.write('\n')
self._stream.write(' std::vector<%s> expected_values{%s};\n' % (np_dt_to_c(self._dtype), print_values(data_out.reshape(-1))))
self._stream.write(' EXPECT_EQ(output_values, expected_values);\n')
self._stream.write(' }\n')
self._stream.write(' check_success<%s>\n'
' (%s,\n'
' %s,\n'
' std::vector<int64_t>{%s},\n'
' std::vector<int64_t>{%s},\n'
' std::vector<int64_t>{%s},\n'
' AxisSet{%s},\n'
' AxisSet{%s},\n'
' AxisSet{%s},\n'
' AxisSet{%s},\n'
' AxisSet{%s},\n'
' %s,\n'
' std::vector<%s>{%s});\n'
% (np_dt_to_c(self._dtype),
np_dt_to_ng(self._dtype),
print_shape(data_in.shape),
print_lb_values(slices),
print_ub_values(slices),
print_stride_values(slices),
print_lb_mask_axes(slices),
print_ub_mask_axes(slices),
print_new_mask_axes(slices),
print_shrink_mask_axes(slices),
print_ellipsis_mask_axes(slices),
print_shape(data_out.shape),
np_dt_to_c(self._dtype), print_values(data_out.reshape(-1))))
def set_shape(self,shape):
self._shape = shape
......@@ -402,6 +365,99 @@ using namespace ngraph;
static string s_manifest = "${MANIFEST}";
template <typename T>
void check_failure(const element::Type& input_element_type,
const Shape& input_shape,
const std::vector<int64_t>& lb_values,
const std::vector<int64_t>& ub_values,
const std::vector<int64_t>& strides_values,
const AxisSet& lb_mask,
const AxisSet& ub_mask,
const AxisSet& new_mask,
const AxisSet& shrink_mask,
const AxisSet& ellipsis_mask)
{
auto arg = std::make_shared<op::Parameter>(input_element_type, input_shape);
auto lb = std::make_shared<op::Parameter>(element::i64, Shape{lb_values.size()});
auto ub = std::make_shared<op::Parameter>(element::i64, Shape{ub_values.size()});
auto strides = std::make_shared<op::Parameter>(element::i64, Shape{strides_values.size()});
std::vector<T> input_values(shape_size(input_shape));
std::iota(input_values.begin(), input_values.end(), static_cast<T>(0));
EXPECT_ANY_THROW({
auto slice = std::make_shared<op::DynSlice>(arg, lb, ub, strides, lb_mask, ub_mask, new_mask, shrink_mask, ellipsis_mask);
auto f = std::make_shared<Function>(NodeVector{slice}, ParameterVector{arg, lb, ub, strides});
auto backend = runtime::Backend::create("${BACKEND_NAME}",true);
auto ex = backend->compile(f);
auto input_arg = backend->create_tensor(input_element_type, input_shape);
auto input_lb = backend->create_tensor(element::i64, Shape{lb_values.size()});
auto input_ub = backend->create_tensor(element::i64, Shape{ub_values.size()});
auto input_strides = backend->create_tensor(element::i64, Shape{strides_values.size()});
copy_data(input_arg, input_values);
copy_data(input_lb, lb_values);
copy_data(input_ub, ub_values);
copy_data(input_strides, strides_values);
auto output = backend->create_dynamic_tensor(input_element_type, PartialShape::dynamic());
ex->call_with_validate({output}, {input_arg, input_lb, input_ub, input_strides});
});
}
template <typename T>
void check_success(const element::Type& input_element_type,
const Shape& input_shape,
const std::vector<int64_t>& lb_values,
const std::vector<int64_t>& ub_values,
const std::vector<int64_t>& strides_values,
const AxisSet& lb_mask,
const AxisSet& ub_mask,
const AxisSet& new_mask,
const AxisSet& shrink_mask,
const AxisSet& ellipsis_mask,
const Shape& expected_output_shape,
const std::vector<T>& expected_values)
{
auto arg = std::make_shared<op::Parameter>(input_element_type, input_shape);
auto lb = std::make_shared<op::Parameter>(element::i64, Shape{lb_values.size()});
auto ub = std::make_shared<op::Parameter>(element::i64, Shape{ub_values.size()});
auto strides = std::make_shared<op::Parameter>(element::i64, Shape{strides_values.size()});
std::vector<T> input_values(shape_size(input_shape));
std::iota(input_values.begin(), input_values.end(), static_cast<T>(0));
auto slice = std::make_shared<op::DynSlice>(arg, lb, ub, strides, lb_mask, ub_mask, new_mask, shrink_mask, ellipsis_mask);
auto f = std::make_shared<Function>(NodeVector{slice}, ParameterVector{arg, lb, ub, strides});
auto backend = runtime::Backend::create("${BACKEND_NAME}",true);
auto ex = backend->compile(f);
auto input_arg = backend->create_tensor(input_element_type, input_shape);
auto input_lb = backend->create_tensor(element::i64, Shape{lb_values.size()});
auto input_ub = backend->create_tensor(element::i64, Shape{ub_values.size()});
auto input_strides = backend->create_tensor(element::i64, Shape{strides_values.size()});
copy_data(input_arg, input_values);
copy_data(input_lb, lb_values);
copy_data(input_ub, ub_values);
copy_data(input_strides, strides_values);
auto output = backend->create_dynamic_tensor(input_element_type, PartialShape::dynamic());
ex->call_with_validate({output}, {input_arg, input_lb, input_ub, input_strides});
EXPECT_EQ(output->get_element_type(), input_element_type);
EXPECT_EQ(output->get_shape(), expected_output_shape);
auto output_values = read_vector<T>(output);
EXPECT_EQ(output_values, expected_values);
}
NGRAPH_TEST(${BACKEND_NAME}, dyn_slice)
{
''')
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment