wehub-resource-sync
493 lines
17 KiB
Python
493 lines
17 KiB
Python
# Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved.
|
|
#
|
|
# Licensed under the Apache License, Version 2.0 (the "License");
|
|
# you may not use this file except in compliance with the License.
|
|
# You may obtain a copy of the License at
|
|
#
|
|
# http://www.apache.org/licenses/LICENSE-2.0
|
|
#
|
|
# Unless required by applicable law or agreed to in writing, software
|
|
# distributed under the License is distributed on an "AS IS" BASIS,
|
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
# See the License for the specific language governing permissions and
|
|
# limitations under the License.
|
|
|
|
import unittest
|
|
|
|
import numpy as np
|
|
from op_test import (
|
|
OpTest,
|
|
check_cudnn_version_and_compute_capability,
|
|
convert_float_to_uint16,
|
|
skip_check_grad_ci,
|
|
)
|
|
|
|
import paddle
|
|
|
|
|
|
class TestElementwiseOp(OpTest):
|
|
def init_data(self):
|
|
# If x and y have the same value, the max() is not differentiable.
|
|
# So we generate test data by the following method
|
|
# to avoid them being too close to each other.
|
|
self.x = np.random.uniform(0.1, 1, [13, 17]).astype("float64")
|
|
sgn = np.random.choice([-1, 1], [13, 17]).astype("float64")
|
|
self.y = self.x + sgn * np.random.uniform(0.1, 1, [13, 17]).astype(
|
|
"float64"
|
|
)
|
|
|
|
def setUp(self):
|
|
self.init_data()
|
|
self.op_type = "elementwise_max"
|
|
self.prim_op_type = "prim"
|
|
self.if_enable_cinn()
|
|
self.python_api = paddle.maximum
|
|
self.public_python_api = paddle.maximum
|
|
self.inputs = {'X': self.x, 'Y': self.y}
|
|
self.outputs = {'Out': np.maximum(self.inputs['X'], self.inputs['Y'])}
|
|
|
|
def test_check_output(self):
|
|
if hasattr(self, 'attrs'):
|
|
self.check_output(check_dygraph=False)
|
|
else:
|
|
self.check_output()
|
|
|
|
def test_check_grad_normal(self):
|
|
if hasattr(self, 'attrs'):
|
|
if self.attrs['axis'] == -1:
|
|
self.check_grad(
|
|
['X', 'Y'],
|
|
'Out',
|
|
check_dygraph=False,
|
|
check_prim=False,
|
|
check_prim_pir=True,
|
|
)
|
|
else:
|
|
self.check_grad(['X', 'Y'], 'Out', check_dygraph=False)
|
|
else:
|
|
self.check_grad(
|
|
['X', 'Y'], 'Out', check_prim=False, check_prim_pir=True
|
|
)
|
|
|
|
def test_check_grad_ignore_x(self):
|
|
if hasattr(self, 'attrs') and self.attrs['axis'] != -1:
|
|
self.check_grad(
|
|
['Y'],
|
|
'Out',
|
|
max_relative_error=0.005,
|
|
no_grad_set=set("X"),
|
|
check_dygraph=False,
|
|
)
|
|
else:
|
|
self.check_grad(
|
|
['Y'],
|
|
'Out',
|
|
max_relative_error=0.005,
|
|
no_grad_set=set("X"),
|
|
check_prim=False,
|
|
check_prim_pir=True,
|
|
)
|
|
|
|
def test_check_grad_ignore_y(self):
|
|
if hasattr(self, 'attrs') and self.attrs['axis'] != -1:
|
|
self.check_grad(
|
|
['X'],
|
|
'Out',
|
|
max_relative_error=0.005,
|
|
no_grad_set=set('Y'),
|
|
check_dygraph=False,
|
|
)
|
|
else:
|
|
self.check_grad(
|
|
['X'],
|
|
'Out',
|
|
max_relative_error=0.005,
|
|
no_grad_set=set('Y'),
|
|
check_prim=False,
|
|
check_prim_pir=True,
|
|
)
|
|
|
|
def if_enable_cinn(self):
|
|
pass
|
|
|
|
|
|
class TestElementwiseFP16Op(TestElementwiseOp):
|
|
def init_data(self):
|
|
self.x = np.random.uniform(0.1, 1, [13, 17]).astype(np.float16)
|
|
sgn = np.random.choice([-1, 1], [13, 17]).astype(np.float16)
|
|
self.y = self.x + sgn * np.random.uniform(0.1, 1, [13, 17]).astype(
|
|
np.float16
|
|
)
|
|
|
|
def setUp(self):
|
|
self.init_data()
|
|
self.op_type = "elementwise_max"
|
|
self.prim_op_type = "prim"
|
|
self.if_enable_cinn()
|
|
self.python_api = paddle.maximum
|
|
self.dtype = np.float16
|
|
self.public_python_api = paddle.maximum
|
|
self.inputs = {'X': self.x, 'Y': self.y}
|
|
self.outputs = {'Out': np.maximum(self.inputs['X'], self.inputs['Y'])}
|
|
|
|
|
|
class TestElementwiseMaxOp_ZeroDim1(TestElementwiseOp):
|
|
def init_data(self):
|
|
self.x = np.random.uniform(0.1, 1, []).astype("float64")
|
|
self.y = np.random.uniform(0.1, 1, []).astype("float64")
|
|
|
|
|
|
class TestElementwiseMaxFP16Op_ZeroDim1(TestElementwiseFP16Op):
|
|
def init_data(self):
|
|
self.x = np.random.uniform(0.1, 1, []).astype(np.float16)
|
|
self.y = np.random.uniform(0.1, 1, []).astype(np.float16)
|
|
|
|
|
|
class TestElementwiseMaxOp_ZeroDim2(TestElementwiseOp):
|
|
def init_data(self):
|
|
self.x = np.random.uniform(0.1, 1, [13, 17]).astype("float64")
|
|
self.y = np.random.uniform(0.1, 1, []).astype("float64")
|
|
|
|
|
|
class TestElementwiseMaxFP16Op_ZeroDim2(TestElementwiseFP16Op):
|
|
def init_data(self):
|
|
self.x = np.random.uniform(0.1, 1, [13, 17]).astype(np.float16)
|
|
self.y = np.random.uniform(0.1, 1, []).astype(np.float16)
|
|
|
|
|
|
class TestElementwiseMaxOp_ZeroDim3(TestElementwiseOp):
|
|
def init_data(self):
|
|
self.x = np.random.uniform(0.1, 1, []).astype("float64")
|
|
self.y = np.random.uniform(0.1, 1, [13, 17]).astype("float64")
|
|
|
|
|
|
class TestElementwiseMaxFP16Op_ZeroDim3(TestElementwiseFP16Op):
|
|
def init_data(self):
|
|
self.x = np.random.uniform(0.1, 1, []).astype(np.float16)
|
|
self.y = np.random.uniform(0.1, 1, [13, 17]).astype(np.float16)
|
|
|
|
|
|
@unittest.skipIf(
|
|
not check_cudnn_version_and_compute_capability(8100, 8),
|
|
"only support compiled with CUDA or custom device, and for CUDA cudnn version need larger than 8.1.0 and device's compute capability is at least 8.0",
|
|
)
|
|
class TestElementwiseBF16Op(OpTest):
|
|
def init_data(self):
|
|
# If x and y have the same value, the max() is not differentiable.
|
|
# So we generate test data by the following method
|
|
# to avoid them being too close to each other.
|
|
self.x = np.random.uniform(0.1, 1, [13, 17]).astype(np.float32)
|
|
sgn = np.random.choice([-1, 1], [13, 17]).astype(np.float32)
|
|
self.y = self.x + sgn * np.random.uniform(0.1, 1, [13, 17]).astype(
|
|
np.float32
|
|
)
|
|
|
|
def setUp(self):
|
|
self.init_data()
|
|
self.op_type = "elementwise_max"
|
|
self.python_api = paddle.maximum
|
|
self.public_python_api = paddle.maximum
|
|
self.prim_op_type = "prim"
|
|
self.dtype = np.uint16
|
|
self.inputs = {
|
|
'X': convert_float_to_uint16(self.x),
|
|
'Y': convert_float_to_uint16(self.y),
|
|
}
|
|
self.outputs = {
|
|
'Out': convert_float_to_uint16(np.maximum(self.x, self.y))
|
|
}
|
|
self.if_enable_cinn()
|
|
|
|
def test_check_output(self):
|
|
if hasattr(self, 'attrs'):
|
|
self.check_output(check_dygraph=False)
|
|
else:
|
|
self.check_output(check_dygraph=True)
|
|
|
|
def if_enable_cinn(self):
|
|
pass
|
|
|
|
def test_check_grad_normal(self):
|
|
if hasattr(self, 'attrs'):
|
|
# check_prim=False, bfloat16 is not supported in `less_equal`
|
|
self.check_grad(
|
|
['X', 'Y'], 'Out', numeric_grad_delta=0.05, check_dygraph=False
|
|
)
|
|
else:
|
|
self.check_grad(
|
|
['X', 'Y'],
|
|
'Out',
|
|
numeric_grad_delta=0.05,
|
|
check_prim=False,
|
|
check_prim_pir=True,
|
|
)
|
|
|
|
def test_check_grad_ignore_x(self):
|
|
self.check_grad(
|
|
['Y'],
|
|
'Out',
|
|
numeric_grad_delta=0.05,
|
|
no_grad_set=set("X"),
|
|
check_prim=False,
|
|
check_prim_pir=True,
|
|
)
|
|
|
|
def test_check_grad_ignore_y(self):
|
|
self.check_grad(
|
|
['X'],
|
|
'Out',
|
|
numeric_grad_delta=0.05,
|
|
no_grad_set=set('Y'),
|
|
check_prim=False,
|
|
check_prim_pir=True,
|
|
)
|
|
|
|
|
|
class TestElementwiseMaxBF16Op_ZeroDim1(TestElementwiseBF16Op):
|
|
def init_data(self):
|
|
self.x = np.random.uniform(0.1, 1, []).astype("float32")
|
|
self.y = np.random.uniform(0.1, 1, []).astype("float32")
|
|
|
|
|
|
class TestElementwiseMaxBF16Op_scalar(TestElementwiseBF16Op):
|
|
def init_data(self):
|
|
self.x = np.random.random_integers(-5, 5, [2, 3, 20]).astype("float32")
|
|
self.y = np.array([0.5]).astype("float32")
|
|
self.__class__.no_need_check_grad = True
|
|
|
|
|
|
@skip_check_grad_ci(
|
|
reason="[skip shape check] Use y_shape(1) to test broadcast."
|
|
)
|
|
class TestElementwiseMaxOp_scalar(TestElementwiseOp):
|
|
def init_data(self):
|
|
self.x = np.random.random_integers(-5, 5, [2, 3, 20]).astype("float64")
|
|
self.y = np.array([0.5]).astype("float64")
|
|
|
|
|
|
@skip_check_grad_ci(
|
|
reason="[skip shape check] Use y_shape(1) to test broadcast."
|
|
)
|
|
class TestElementwiseMaxFP16Op_scalar(TestElementwiseFP16Op):
|
|
def init_data(self):
|
|
self.x = np.random.random_integers(-5, 5, [2, 3, 20]).astype(np.float16)
|
|
self.y = np.array([0.5]).astype(np.float16)
|
|
|
|
|
|
class TestElementwiseMaxOp_Vector(TestElementwiseOp):
|
|
def init_data(self):
|
|
self.x = np.random.random((100,)).astype("float64")
|
|
sgn = np.random.choice([-1, 1], (100,)).astype("float64")
|
|
self.y = self.x + sgn * np.random.uniform(0.1, 1, (100,)).astype(
|
|
"float64"
|
|
)
|
|
|
|
|
|
class TestElementwiseMaxFP16Op_Vector(TestElementwiseFP16Op):
|
|
def init_data(self):
|
|
self.x = np.random.random((100,)).astype(np.float16)
|
|
sgn = np.random.choice([-1, 1], (100,)).astype(np.float16)
|
|
self.y = self.x + sgn * np.random.uniform(0.1, 1, (100,)).astype(
|
|
np.float16
|
|
)
|
|
|
|
|
|
class TestElementwiseMaxBF16Op_Vector(TestElementwiseBF16Op):
|
|
def init_data(self):
|
|
self.x = np.random.random((100,)).astype("float32")
|
|
sgn = np.random.choice([-1, 1], (100,)).astype("float32")
|
|
self.y = self.x + sgn * np.random.uniform(0.1, 1, (100,)).astype(
|
|
"float32"
|
|
)
|
|
|
|
|
|
class TestElementwiseMaxOp_broadcast_2(TestElementwiseOp):
|
|
def setUp(self):
|
|
self.op_type = "elementwise_max"
|
|
self.python_api = paddle.maximum
|
|
self.public_python_api = paddle.maximum
|
|
self.prim_op_type = "prim"
|
|
x = np.random.uniform(0.5, 1, (1, 3, 100)).astype(np.float64)
|
|
sgn = np.random.choice([-1, 1], (100,)).astype(np.float64)
|
|
y = x[0, 0, :] + sgn * np.random.uniform(1, 2, (100,)).astype(
|
|
np.float64
|
|
)
|
|
self.inputs = {'X': x, 'Y': y}
|
|
|
|
self.outputs = {
|
|
'Out': np.maximum(
|
|
self.inputs['X'], self.inputs['Y'].reshape(1, 1, 100)
|
|
)
|
|
}
|
|
|
|
|
|
class TestElementwiseMaxFP16Op_broadcast_2(TestElementwiseFP16Op):
|
|
def setUp(self):
|
|
self.op_type = "elementwise_max"
|
|
self.python_api = paddle.maximum
|
|
self.public_python_api = paddle.maximum
|
|
self.prim_op_type = "prim"
|
|
self.dtype = np.float16
|
|
x = np.random.uniform(0.5, 1, (1, 3, 100)).astype(np.float16)
|
|
sgn = np.random.choice([-1, 1], (100,)).astype(np.float16)
|
|
y = x[0, 0, :] + sgn * np.random.uniform(1, 2, (100,)).astype(
|
|
np.float16
|
|
)
|
|
self.inputs = {'X': x, 'Y': y}
|
|
|
|
self.outputs = {
|
|
'Out': np.maximum(
|
|
self.inputs['X'], self.inputs['Y'].reshape(1, 1, 100)
|
|
)
|
|
}
|
|
|
|
|
|
class TestElementwiseMaxOp_broadcast_4(TestElementwiseOp):
|
|
def setUp(self):
|
|
self.op_type = "elementwise_max"
|
|
self.python_api = paddle.maximum
|
|
self.public_python_api = paddle.maximum
|
|
self.prim_op_type = "prim"
|
|
x = np.random.uniform(0.5, 1, (2, 3, 4, 5)).astype(np.float64)
|
|
sgn = np.random.choice([-1, 1], (2, 3, 1, 5)).astype(np.float64)
|
|
y = x + sgn * np.random.uniform(1, 2, (2, 3, 1, 5)).astype(np.float64)
|
|
self.inputs = {'X': x, 'Y': y}
|
|
|
|
self.outputs = {'Out': np.maximum(self.inputs['X'], self.inputs['Y'])}
|
|
|
|
|
|
class TestElementwiseFP16Op_broadcast_4(TestElementwiseFP16Op):
|
|
def setUp(self):
|
|
self.op_type = "elementwise_max"
|
|
self.python_api = paddle.maximum
|
|
self.public_python_api = paddle.maximum
|
|
self.prim_op_type = "prim"
|
|
self.dtype = np.float16
|
|
x = np.random.uniform(0.5, 1, (2, 3, 4, 5)).astype(np.float16)
|
|
sgn = np.random.choice([-1, 1], (2, 3, 1, 5)).astype(np.float16)
|
|
y = x + sgn * np.random.uniform(1, 2, (2, 3, 1, 5)).astype(np.float16)
|
|
self.inputs = {'X': x, 'Y': y}
|
|
self.outputs = {'Out': np.maximum(self.inputs['X'], self.inputs['Y'])}
|
|
|
|
|
|
class TestElementwiseOpEqualInput(TestElementwiseOp):
|
|
def init_data(self):
|
|
self.x = np.ones([13, 17]).astype(np.float32)
|
|
self.y = np.ones([13, 17]).astype(np.float32)
|
|
|
|
def setUp(self):
|
|
self.init_data()
|
|
self.op_type = "elementwise_max"
|
|
self.prim_op_type = "prim"
|
|
self.if_enable_cinn()
|
|
self.python_api = paddle.maximum
|
|
self.dtype = np.float32
|
|
self.public_python_api = paddle.maximum
|
|
self.inputs = {'X': self.x, 'Y': self.y}
|
|
self.outputs = {'Out': np.maximum(self.inputs['X'], self.inputs['Y'])}
|
|
|
|
|
|
class TestElementwiseOp0SizeInput(TestElementwiseOp):
|
|
def init_data(self):
|
|
self.x = np.ones([0, 1, 2]).astype(np.float32)
|
|
self.y = np.ones([1, 3598, 2]).astype(np.float32)
|
|
|
|
def setUp(self):
|
|
self.init_data()
|
|
self.op_type = "elementwise_max"
|
|
self.prim_op_type = "prim"
|
|
self.if_enable_cinn()
|
|
self.python_api = paddle.maximum
|
|
self.dtype = np.float32
|
|
self.public_python_api = paddle.maximum
|
|
self.inputs = {'X': self.x, 'Y': self.y}
|
|
self.outputs = {'Out': np.maximum(self.inputs['X'], self.inputs['Y'])}
|
|
|
|
|
|
class TestMaximumOutAndAlias(unittest.TestCase):
|
|
def test_dygraph(self):
|
|
with paddle.base.dygraph.guard():
|
|
np.random.seed(2024)
|
|
x = paddle.to_tensor(
|
|
np.random.randn(5, 7).astype('float32'), stop_gradient=False
|
|
)
|
|
# shift y to avoid ties for stable gradient routing
|
|
y = paddle.to_tensor(
|
|
(np.random.randn(5, 7) + 0.1).astype('float32'),
|
|
stop_gradient=False,
|
|
)
|
|
|
|
def run_case(case_type):
|
|
out_buf = paddle.zeros_like(x)
|
|
out_buf.stop_gradient = False
|
|
|
|
if case_type == 'return':
|
|
z = paddle.maximum(x, y)
|
|
elif case_type == 'input_out':
|
|
paddle.maximum(x, y, out=out_buf)
|
|
z = out_buf
|
|
elif case_type == 'both_return':
|
|
z = paddle.maximum(input=x, other=y, out=out_buf)
|
|
elif case_type == 'both_input_out':
|
|
_ = paddle.maximum(input=x, other=y, out=out_buf)
|
|
z = out_buf
|
|
else:
|
|
raise AssertionError
|
|
|
|
ref = paddle._C_ops.maximum(x, y)
|
|
np.testing.assert_allclose(
|
|
z.numpy(), ref.numpy(), rtol=1e-6, atol=1e-6
|
|
)
|
|
|
|
loss = (z * 2).mean()
|
|
loss.backward()
|
|
return z.numpy(), x.grad.numpy(), y.grad.numpy()
|
|
|
|
z1, gx1, gy1 = run_case('return')
|
|
x.clear_gradient()
|
|
y.clear_gradient()
|
|
z2, gx2, gy2 = run_case('input_out')
|
|
x.clear_gradient()
|
|
y.clear_gradient()
|
|
z3, gx3, gy3 = run_case('both_return')
|
|
x.clear_gradient()
|
|
y.clear_gradient()
|
|
z4, gx4, gy4 = run_case('both_input_out')
|
|
|
|
np.testing.assert_allclose(z1, z2, rtol=1e-6, atol=1e-6)
|
|
np.testing.assert_allclose(z1, z3, rtol=1e-6, atol=1e-6)
|
|
np.testing.assert_allclose(z1, z4, rtol=1e-6, atol=1e-6)
|
|
np.testing.assert_allclose(gx1, gx2, rtol=1e-6, atol=1e-6)
|
|
np.testing.assert_allclose(gx1, gx3, rtol=1e-6, atol=1e-6)
|
|
np.testing.assert_allclose(gx1, gx4, rtol=1e-6, atol=1e-6)
|
|
np.testing.assert_allclose(gy1, gy2, rtol=1e-6, atol=1e-6)
|
|
np.testing.assert_allclose(gy1, gy3, rtol=1e-6, atol=1e-6)
|
|
np.testing.assert_allclose(gy1, gy4, rtol=1e-6, atol=1e-6)
|
|
|
|
def test_static(self):
|
|
paddle.enable_static()
|
|
startup_prog = paddle.static.Program()
|
|
main_prog = paddle.static.Program()
|
|
|
|
with paddle.static.program_guard(main_prog, startup_prog):
|
|
x = paddle.static.data('X', [5, 7], 'float32')
|
|
y = paddle.static.data('Y', [5, 7], 'float32')
|
|
z = paddle.maximum(input=x, other=y)
|
|
|
|
x_data = np.random.random([5, 7]).astype('float32')
|
|
y_data = np.random.random([5, 7]).astype('float32')
|
|
ref = np.maximum(x_data, y_data)
|
|
|
|
exe = paddle.static.Executor(paddle.CPUPlace())
|
|
exe.run(startup_prog)
|
|
out = exe.run(
|
|
main_prog,
|
|
feed={'X': x_data, 'Y': y_data},
|
|
fetch_list=[z],
|
|
)
|
|
np.testing.assert_allclose(out[0], ref, rtol=1e-6, atol=1e-6)
|
|
|
|
|
|
if __name__ == '__main__':
|
|
unittest.main()
|