wehub-resource-sync
466 lines
15 KiB
Python
466 lines
15 KiB
Python
# Copyright (c) 2023 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
|
|
|
|
import paddle
|
|
|
|
# NOTE(Pan Zhaowu): Using legacy_linear to fulfill promise of high-level grad,
|
|
# with no side-effects to other ops.
|
|
# linear_v2's decomposed grad is fully tested in test_gradname_parse.py
|
|
paddle.set_flags({"FLAGS_use_legacy_linear": True})
|
|
from paddle.autograd.backward_utils import ValueDict, ValueSet
|
|
from paddle.autograd.ir_backward import grad
|
|
from paddle.base.wrapped_decorator import signature_safe_contextmanager
|
|
|
|
paddle.enable_static()
|
|
|
|
|
|
@signature_safe_contextmanager
|
|
def dygraph_guard():
|
|
in_dygraph_outside = paddle.base.framework.in_dygraph_mode()
|
|
try:
|
|
if not in_dygraph_outside:
|
|
paddle.disable_static()
|
|
yield
|
|
finally:
|
|
if not in_dygraph_outside:
|
|
paddle.enable_static()
|
|
|
|
|
|
def get_ir_program_0():
|
|
paddle.enable_static()
|
|
x = paddle.randn([4, 4])
|
|
main_program, start_program = (
|
|
paddle.static.Program(),
|
|
paddle.static.Program(),
|
|
)
|
|
with paddle.static.program_guard(main_program, start_program):
|
|
x_s = paddle.static.data('x', [4, 4], x.dtype)
|
|
x_s.stop_gradient = False
|
|
k_s = paddle.tanh(x_s)
|
|
return main_program
|
|
|
|
|
|
class TesBackward_1(unittest.TestCase):
|
|
def test_grad(self):
|
|
pir_program = get_ir_program_0()
|
|
input = pir_program.global_block().ops[-1].operand(0).source()
|
|
tanh_out = pir_program.global_block().ops[-1].result(0)
|
|
with (
|
|
paddle.pir_utils.IrGuard(),
|
|
paddle.pir.core.program_guard(pir_program),
|
|
):
|
|
out = paddle.mean(tanh_out)
|
|
out2 = paddle.mean(tanh_out)
|
|
input_grad = grad(out, input, out2)
|
|
|
|
self.assertEqual(out.get_defining_op().name(), "pd_op.mean")
|
|
self.assertEqual(
|
|
input_grad[0].get_defining_op().name(), "pd_op.tanh_grad"
|
|
)
|
|
self.assertEqual(
|
|
out.get_defining_op()
|
|
.operands()[0]
|
|
.source()
|
|
.get_defining_op()
|
|
.name(),
|
|
"pd_op.tanh",
|
|
)
|
|
|
|
def test_full(self):
|
|
# test create output_grad in backward use full op
|
|
pir_program = get_ir_program_0()
|
|
input = pir_program.global_block().ops[-1].operand(0).source()
|
|
tanh_out = pir_program.global_block().ops[-1].result(0)
|
|
with (
|
|
paddle.pir_utils.IrGuard(),
|
|
paddle.pir.core.program_guard(pir_program),
|
|
):
|
|
out = paddle.mean(tanh_out)
|
|
input_grad = grad(out, input)
|
|
self.assertEqual(
|
|
pir_program.global_block().ops[-3].name(), "pd_op.full_like"
|
|
)
|
|
self.assertEqual(
|
|
input_grad[0].get_defining_op().name(), "pd_op.tanh_grad"
|
|
)
|
|
self.assertEqual(
|
|
input_grad[0]
|
|
.get_defining_op()
|
|
.operands()[1]
|
|
.source()
|
|
.get_defining_op()
|
|
.name(),
|
|
"pd_op.mean_grad",
|
|
)
|
|
|
|
def test_no_grad_set(self):
|
|
# test create output_grad in backward use full op
|
|
pir_program = get_ir_program_0()
|
|
input = pir_program.global_block().ops[-1].operand(0).source()
|
|
tanh_out = pir_program.global_block().ops[-1].result(0)
|
|
with (
|
|
paddle.pir_utils.IrGuard(),
|
|
paddle.pir.core.program_guard(pir_program),
|
|
):
|
|
out = paddle.mean(tanh_out)
|
|
input_grad = grad(out, input, no_grad_vars=[input])
|
|
self.assertEqual(
|
|
pir_program.global_block().ops[-3].name(), "pd_op.mean"
|
|
)
|
|
|
|
def test_split(self):
|
|
# test create output_grad in backward use full op
|
|
pir_program = get_ir_program_0()
|
|
input = pir_program.global_block().ops[-1].operand(0).source()
|
|
tanh_out = pir_program.global_block().ops[-1].result(0)
|
|
with (
|
|
paddle.pir_utils.IrGuard(),
|
|
paddle.pir.core.program_guard(pir_program),
|
|
):
|
|
out = paddle.split(tanh_out, [2, 2], 0)
|
|
input_grad = grad(out, input)
|
|
ops_name = [
|
|
"pd_op.data",
|
|
"pd_op.tanh",
|
|
"pd_op.full_int_array",
|
|
"pd_op.full",
|
|
"pd_op.split",
|
|
"builtin.split",
|
|
"pd_op.full",
|
|
"pd_op.full_like",
|
|
"pd_op.full",
|
|
"pd_op.full_like",
|
|
"builtin.combine",
|
|
"pd_op.concat",
|
|
"pd_op.tanh_grad",
|
|
]
|
|
for i, op in enumerate(pir_program.global_block().ops):
|
|
self.assertEqual(op.name(), ops_name[i])
|
|
|
|
|
|
def get_ir_program_1():
|
|
paddle.enable_static()
|
|
x = paddle.randn([2, 2])
|
|
main_program, start_program = (
|
|
paddle.static.Program(),
|
|
paddle.static.Program(),
|
|
)
|
|
with paddle.static.program_guard(main_program, start_program):
|
|
x_s = paddle.static.data('x', [4, 4], x.dtype)
|
|
x_s.stop_gradient = False
|
|
|
|
k_s = paddle.tanh(x_s)
|
|
z_x = paddle.tanh(x_s)
|
|
out = paddle.add(z_x, k_s)
|
|
return main_program
|
|
|
|
|
|
class TesBackward_2(unittest.TestCase):
|
|
def test_add_n(self):
|
|
pir_program = get_ir_program_1()
|
|
input_x = pir_program.global_block().ops[-3].operand(0).source()
|
|
|
|
add_out = pir_program.global_block().ops[-1].result(0)
|
|
with (
|
|
paddle.pir_utils.IrGuard(),
|
|
paddle.pir.core.program_guard(pir_program),
|
|
):
|
|
out = paddle.mean(add_out)
|
|
input_grad = grad(out, input_x)
|
|
|
|
self.assertEqual(
|
|
pir_program.global_block().ops[-1].name(), "pd_op.add_n"
|
|
)
|
|
self.assertEqual(
|
|
pir_program.global_block().ops[-1].name(), "pd_op.add_n"
|
|
)
|
|
self.assertEqual(
|
|
pir_program.global_block().ops[-2].name(), "builtin.combine"
|
|
)
|
|
|
|
def test_concat(self):
|
|
pir_program = get_ir_program_1()
|
|
input_x = pir_program.global_block().ops[-3].operand(0).source()
|
|
|
|
add_out = pir_program.global_block().ops[-1].result(0)
|
|
with (
|
|
paddle.pir_utils.IrGuard(),
|
|
paddle.pir.core.program_guard(pir_program),
|
|
):
|
|
out = paddle.concat([add_out, add_out])
|
|
input_grad = grad(out, input_x)
|
|
ops_name = [
|
|
"pd_op.data",
|
|
"pd_op.tanh",
|
|
"pd_op.tanh",
|
|
"pd_op.add",
|
|
"pd_op.full",
|
|
"builtin.combine",
|
|
"pd_op.concat",
|
|
"pd_op.full",
|
|
"pd_op.full_like",
|
|
"builtin.combine",
|
|
"pd_op.concat_grad",
|
|
"builtin.split",
|
|
"builtin.combine",
|
|
"pd_op.add_n",
|
|
"pd_op.add_grad",
|
|
"pd_op.tanh_grad",
|
|
"pd_op.tanh_grad",
|
|
"builtin.combine",
|
|
"pd_op.add_n",
|
|
]
|
|
for i, op in enumerate(pir_program.global_block().ops):
|
|
self.assertEqual(op.name(), ops_name[i])
|
|
|
|
|
|
def get_ir_program_2():
|
|
paddle.enable_static()
|
|
x = paddle.randn([2, 2])
|
|
main_program, start_program = (
|
|
paddle.static.Program(),
|
|
paddle.static.Program(),
|
|
)
|
|
with paddle.static.program_guard(main_program, start_program):
|
|
x_s = paddle.static.data('x', [4, 4], x.dtype)
|
|
x_s.stop_gradient = False
|
|
k_s = paddle.sum(x_s, axis=(-1,), keepdim=False)
|
|
return main_program
|
|
|
|
|
|
class TestBackward_3(unittest.TestCase):
|
|
def test_basic_network(self):
|
|
pir_program = get_ir_program_2()
|
|
x = pir_program.global_block().ops[-1].operand(0).source()
|
|
sum_x = pir_program.global_block().ops[-1].result(0)
|
|
with (
|
|
paddle.pir_utils.IrGuard(),
|
|
paddle.pir.core.program_guard(pir_program),
|
|
):
|
|
norm = paddle.tensor.fill_constant(
|
|
shape=[],
|
|
value=1.0,
|
|
dtype=sum_x.dtype,
|
|
)
|
|
res = paddle.divide(sum_x, norm)
|
|
input_grad = grad(res, x)
|
|
|
|
|
|
class TestBackward_4(unittest.TestCase):
|
|
def test_basic_network(self):
|
|
if not paddle.framework.in_pir_mode():
|
|
return
|
|
program = paddle.static.default_main_program()
|
|
with paddle.static.program_guard(program):
|
|
x = paddle.randn((2, 2))
|
|
x.stop_gradient = False
|
|
b = paddle.to_tensor([12])
|
|
grad_x = 0
|
|
double_x = x * 2
|
|
pred = b > 0
|
|
|
|
def true_func():
|
|
y = double_x * 3
|
|
out = grad(y, x)
|
|
filtered_dx = [dxi for dxi in out if dxi is not None]
|
|
grad_x = filtered_dx
|
|
return grad_x
|
|
|
|
def false_func():
|
|
y = double_x * 4
|
|
out = grad(y, x)
|
|
filtered_dx = [dxi for dxi in out if dxi is not None]
|
|
grad_x = filtered_dx
|
|
return grad_x
|
|
|
|
out = paddle.static.nn.cond(pred, true_func, false_func)
|
|
|
|
place = (
|
|
paddle.base.CUDAPlace(0)
|
|
if paddle.base.core.is_compiled_with_cuda()
|
|
else paddle.base.CPUPlace()
|
|
)
|
|
exe = paddle.static.Executor(place)
|
|
(grad_x,) = exe.run(program, fetch_list=[out])
|
|
res = np.full([2, 2], 6.0, dtype='float32')
|
|
self.assertEqual((grad_x == res).all(), True)
|
|
|
|
|
|
class TestBackward_5(unittest.TestCase):
|
|
def test_skip_vjp(self):
|
|
if not paddle.framework.in_pir_mode():
|
|
return
|
|
program = paddle.static.Program()
|
|
with paddle.static.program_guard(program):
|
|
x = paddle.static.data('x', [4, 4], 'float32')
|
|
x.stop_gradient = True
|
|
y = paddle.nn.functional.relu(x)
|
|
y.stop_gradient = False
|
|
z = paddle.nn.functional.relu(y)
|
|
loss = paddle.mean(z)
|
|
|
|
paddle.autograd.ir_backward.append_backward(loss)
|
|
relu_grad_number = 0
|
|
for op in program.global_block().ops:
|
|
if op.name() == "pd_op.relu_grad":
|
|
relu_grad_number += 1
|
|
|
|
self.assertEqual(relu_grad_number, 1)
|
|
|
|
|
|
class TestBackward_6(unittest.TestCase):
|
|
def test_negative_shape(self):
|
|
with dygraph_guard():
|
|
model = paddle.nn.Linear(2, 3)
|
|
|
|
def f(x):
|
|
y = model(x)
|
|
y = paddle.tanh(y)
|
|
return paddle.grad(
|
|
y, x, create_graph=True, grad_outputs=paddle.randn_like(y)
|
|
)[0]
|
|
|
|
f = paddle.jit.to_static(
|
|
f,
|
|
full_graph=True,
|
|
backend=None,
|
|
input_spec=[paddle.static.InputSpec([-1, -1], dtype="float32")],
|
|
)
|
|
x = paddle.randn(4, 2, requires_grad=True)
|
|
y = f(x)
|
|
self.assertEqual(x.shape, y.shape)
|
|
|
|
def test_negative_shape_error1(self):
|
|
with dygraph_guard():
|
|
model = paddle.nn.Linear(2, 3)
|
|
|
|
def f(x):
|
|
y = model(x)
|
|
y = paddle.tanh(y)
|
|
return paddle.grad(
|
|
y, x, create_graph=True, grad_outputs=paddle.randn(1, 3)
|
|
)[0]
|
|
|
|
with self.assertRaisesRegex(
|
|
ValueError,
|
|
r"The shape of grad_output\[0\] paddle.Size\(\[1, 3\]\) should be the same as the shape of output\[0\] paddle.Size\(\[4, 3\]\)",
|
|
):
|
|
x = paddle.randn(4, 2, requires_grad=True)
|
|
f = paddle.jit.to_static(
|
|
f,
|
|
full_graph=True,
|
|
backend=None,
|
|
input_spec=[
|
|
paddle.static.InputSpec(x.shape, dtype="float32")
|
|
],
|
|
)
|
|
y = f(x)
|
|
|
|
def test_negative_shape_error2(self):
|
|
with dygraph_guard():
|
|
model = paddle.nn.Linear(2, 3)
|
|
|
|
def f(x):
|
|
y = model(x)
|
|
y = paddle.tanh(y)
|
|
return paddle.grad(
|
|
y, x, create_graph=True, grad_outputs=paddle.randn(4)
|
|
)[0]
|
|
|
|
with self.assertRaisesRegex(
|
|
ValueError,
|
|
r"The shape of grad_output\[0\] paddle.Size\(\[4\]\) should be the same as the shape of output\[0\] paddle.Size\(\[4, 3\]\)",
|
|
):
|
|
x = paddle.randn(4, 2, requires_grad=True)
|
|
f = paddle.jit.to_static(
|
|
f,
|
|
full_graph=True,
|
|
backend=None,
|
|
input_spec=[
|
|
paddle.static.InputSpec(x.shape, dtype="float32")
|
|
],
|
|
)
|
|
y = f(x)
|
|
|
|
|
|
class TestValueSet(unittest.TestCase):
|
|
def setUp(self) -> None:
|
|
with paddle.pir_utils.IrGuard():
|
|
main = paddle.static.Program()
|
|
startup = paddle.static.Program()
|
|
with paddle.static.program_guard(main, startup):
|
|
self.x = paddle.static.data('x', [2, 3])
|
|
self.y = paddle.static.data('y', [4, 5])
|
|
|
|
def test_copy(self):
|
|
a = ValueSet([self.x, self.y])
|
|
b = a.copy()
|
|
self.assertNotEqual(id(a), id(b))
|
|
self.assertTrue(len(a) == len(b))
|
|
|
|
def test_or(self):
|
|
a = ValueSet([self.x])
|
|
b = ValueSet([self.y])
|
|
c = a | b
|
|
self.assertTrue(len(c) == 2)
|
|
|
|
|
|
class TestValueDict(unittest.TestCase):
|
|
def setUp(self) -> None:
|
|
with paddle.pir_utils.IrGuard():
|
|
main = paddle.static.Program()
|
|
startup = paddle.static.Program()
|
|
with paddle.static.program_guard(main, startup):
|
|
self.x = paddle.static.data('x', [2, 3])
|
|
self.y = paddle.static.data('y', [4, 5])
|
|
|
|
def test_init(self):
|
|
a = ValueDict()
|
|
a[self.x] = 'x'
|
|
a[self.y] = 'y'
|
|
b = ValueDict(a)
|
|
self.assertTrue(len(b) == 2)
|
|
|
|
def test_bool(self):
|
|
a = ValueDict()
|
|
self.assertFalse(bool(a))
|
|
|
|
def test_getitem_and_pop_error(self):
|
|
with paddle.pir_utils.IrGuard():
|
|
main = paddle.static.Program()
|
|
startup = paddle.static.Program()
|
|
with paddle.static.program_guard(main, startup):
|
|
x = paddle.static.data('x', [2, 3])
|
|
y = paddle.static.data('y', [4, 5])
|
|
a = ValueDict()
|
|
a[x] = 'x'
|
|
self.assertRaises(KeyError, a.__getitem__, y)
|
|
self.assertRaises(KeyError, a.pop, y)
|
|
|
|
def test_update(self):
|
|
a = ValueDict()
|
|
a[self.x] = 'x'
|
|
b = ValueDict()
|
|
b[self.y] = 'y'
|
|
a.update(b)
|
|
self.assertTrue(a[self.y] == 'y')
|
|
|
|
|
|
if __name__ == "__main__":
|
|
unittest.main()
|