wehub-resource-sync
712 lines
25 KiB
Python
712 lines
25 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 numbers
|
|
import unittest
|
|
|
|
import numpy as np
|
|
from op_test import get_device_place, is_custom_device
|
|
|
|
import paddle
|
|
from paddle import base
|
|
from paddle.base import core
|
|
|
|
|
|
class TestTensorPtr(unittest.TestCase):
|
|
def test_tensor_ptr(self):
|
|
t = core.DenseTensor()
|
|
np_arr = np.zeros([2, 3])
|
|
t.set(np_arr, core.CPUPlace())
|
|
self.assertGreater(t._ptr(), 0)
|
|
|
|
|
|
class TestTensor(unittest.TestCase):
|
|
def setUp(self):
|
|
self.support_dtypes = [
|
|
'bool',
|
|
'uint8',
|
|
'int8',
|
|
'int16',
|
|
'int32',
|
|
'int64',
|
|
'float16',
|
|
'float32',
|
|
'float64',
|
|
'complex64',
|
|
'complex128',
|
|
]
|
|
|
|
def test_int_tensor(self):
|
|
scope = core.Scope()
|
|
var = scope.var("test_tensor")
|
|
place = core.CPUPlace()
|
|
|
|
tensor = var.get_tensor()
|
|
|
|
tensor._set_dims([1000, 784])
|
|
tensor._alloc_int(place)
|
|
tensor_array = np.array(tensor)
|
|
self.assertEqual((1000, 784), tensor_array.shape)
|
|
tensor_array[3, 9] = 1
|
|
tensor_array[19, 11] = 2
|
|
tensor.set(tensor_array, place)
|
|
|
|
tensor_array_2 = np.array(tensor)
|
|
self.assertEqual(1, tensor_array_2[3, 9])
|
|
self.assertEqual(2, tensor_array_2[19, 11])
|
|
|
|
def test_float_tensor(self):
|
|
scope = core.Scope()
|
|
var = scope.var("test_tensor")
|
|
place = core.CPUPlace()
|
|
|
|
tensor = var.get_tensor()
|
|
|
|
tensor._set_dims([1000, 784])
|
|
tensor._alloc_float(place)
|
|
|
|
tensor_array = np.array(tensor)
|
|
self.assertEqual((1000, 784), tensor_array.shape)
|
|
tensor_array[3, 9] = 1.0
|
|
tensor_array[19, 11] = 2.0
|
|
tensor.set(tensor_array, place)
|
|
|
|
tensor_array_2 = np.array(tensor)
|
|
self.assertAlmostEqual(1.0, tensor_array_2[3, 9])
|
|
self.assertAlmostEqual(2.0, tensor_array_2[19, 11])
|
|
|
|
def test_int8_tensor(self):
|
|
scope = core.Scope()
|
|
var = scope.var("int8_tensor")
|
|
cpu_tensor = var.get_tensor()
|
|
tensor_array = np.random.randint(
|
|
-127, high=128, size=[100, 200], dtype=np.int8
|
|
)
|
|
place = core.CPUPlace()
|
|
cpu_tensor.set(tensor_array, place)
|
|
cpu_tensor_array_2 = np.array(cpu_tensor)
|
|
self.assertAlmostEqual(cpu_tensor_array_2.all(), tensor_array.all())
|
|
|
|
if core.is_compiled_with_cuda() or is_custom_device():
|
|
cuda_tensor = var.get_tensor()
|
|
tensor_array = np.random.randint(
|
|
-127, high=128, size=[100, 200], dtype=np.int8
|
|
)
|
|
place = get_device_place()
|
|
cuda_tensor.set(tensor_array, place)
|
|
cuda_tensor_array_2 = np.array(cuda_tensor)
|
|
self.assertAlmostEqual(
|
|
cuda_tensor_array_2.all(), tensor_array.all()
|
|
)
|
|
|
|
def test_complex64_tensor(self):
|
|
scope = core.Scope()
|
|
var = scope.var("complex64_tensor")
|
|
cpu_tensor = var.get_tensor()
|
|
tensor_array = (
|
|
np.random.uniform(-1, 1, (100, 200))
|
|
+ 1j * np.random.uniform(-1, 1, (100, 200))
|
|
).astype(np.complex64)
|
|
place = core.CPUPlace()
|
|
cpu_tensor.set(tensor_array, place)
|
|
cpu_tensor_array_2 = np.array(cpu_tensor)
|
|
self.assertAlmostEqual(cpu_tensor_array_2.all(), tensor_array.all())
|
|
|
|
if core.is_compiled_with_cuda() or is_custom_device():
|
|
cuda_tensor = var.get_tensor()
|
|
tensor_array = (
|
|
np.random.uniform(-1, 1, (100, 200))
|
|
+ 1j * np.random.uniform(-1, 1, (100, 200))
|
|
).astype(np.complex64)
|
|
place = get_device_place()
|
|
cuda_tensor.set(tensor_array, place)
|
|
cuda_tensor_array_2 = np.array(cuda_tensor)
|
|
self.assertAlmostEqual(
|
|
cuda_tensor_array_2.all(), tensor_array.all()
|
|
)
|
|
|
|
def test_complex128_tensor(self):
|
|
scope = core.Scope()
|
|
var = scope.var("complex128_tensor")
|
|
cpu_tensor = var.get_tensor()
|
|
tensor_array = (
|
|
np.random.uniform(-1, 1, (100, 200))
|
|
+ 1j * np.random.uniform(-1, 1, (100, 200))
|
|
).astype(np.complex128)
|
|
place = core.CPUPlace()
|
|
cpu_tensor.set(tensor_array, place)
|
|
cpu_tensor_array_2 = np.array(cpu_tensor)
|
|
self.assertAlmostEqual(cpu_tensor_array_2.all(), tensor_array.all())
|
|
|
|
if core.is_compiled_with_cuda() or is_custom_device():
|
|
cuda_tensor = var.get_tensor()
|
|
tensor_array = (
|
|
np.random.uniform(-1, 1, (100, 200))
|
|
+ 1j * np.random.uniform(-1, 1, (100, 200))
|
|
).astype(np.complex128)
|
|
place = get_device_place()
|
|
cuda_tensor.set(tensor_array, place)
|
|
cuda_tensor_array_2 = np.array(cuda_tensor)
|
|
self.assertAlmostEqual(
|
|
cuda_tensor_array_2.all(), tensor_array.all()
|
|
)
|
|
|
|
def test_int_lod_tensor(self):
|
|
place = core.CPUPlace()
|
|
scope = core.Scope()
|
|
var_lod = scope.var("test_lod_tensor")
|
|
lod_tensor = var_lod.get_tensor()
|
|
|
|
lod_tensor._set_dims([4, 4, 6])
|
|
lod_tensor._alloc_int(place)
|
|
array = np.array(lod_tensor)
|
|
array[0, 0, 0] = 3
|
|
array[3, 3, 5] = 10
|
|
lod_tensor.set(array, place)
|
|
|
|
lod_v = np.array(lod_tensor)
|
|
self.assertTrue(np.all(array == lod_v))
|
|
|
|
def test_float_lod_tensor(self):
|
|
place = core.CPUPlace()
|
|
scope = core.Scope()
|
|
var_lod = scope.var("test_lod_tensor")
|
|
|
|
lod_tensor = var_lod.get_tensor()
|
|
lod_tensor._set_dims([5, 2, 3, 4])
|
|
lod_tensor._alloc_float(place)
|
|
|
|
tensor_array = np.array(lod_tensor)
|
|
self.assertEqual((5, 2, 3, 4), tensor_array.shape)
|
|
tensor_array[0, 0, 0, 0] = 1.0
|
|
tensor_array[0, 0, 0, 1] = 2.0
|
|
lod_tensor.set(tensor_array, place)
|
|
|
|
lod_v = np.array(lod_tensor)
|
|
self.assertAlmostEqual(1.0, lod_v[0, 0, 0, 0])
|
|
self.assertAlmostEqual(2.0, lod_v[0, 0, 0, 1])
|
|
|
|
def test_empty_tensor(self):
|
|
place = core.CPUPlace()
|
|
scope = core.Scope()
|
|
var = scope.var("test_tensor")
|
|
|
|
tensor = var.get_tensor()
|
|
tensor._set_dims([0, 1])
|
|
tensor._alloc_float(place)
|
|
|
|
tensor_array = np.array(tensor)
|
|
self.assertEqual((0, 1), tensor_array.shape)
|
|
|
|
if core.is_compiled_with_cuda() or is_custom_device():
|
|
gpu_place = get_device_place()
|
|
tensor._alloc_float(gpu_place)
|
|
tensor_array = np.array(tensor)
|
|
self.assertEqual((0, 1), tensor_array.shape)
|
|
|
|
def run_slice_tensor(self, place, dtype):
|
|
tensor = base.Tensor()
|
|
shape = [3, 3, 3]
|
|
tensor._set_dims(shape)
|
|
|
|
tensor_array = np.array(
|
|
[
|
|
[[1, 2, 3], [4, 5, 6], [7, 8, 9]],
|
|
[[10, 11, 12], [13, 14, 15], [16, 17, 18]],
|
|
[[19, 20, 21], [22, 23, 24], [25, 26, 27]],
|
|
]
|
|
).astype(dtype)
|
|
|
|
tensor.set(tensor_array, place)
|
|
n1 = tensor[1]
|
|
t1 = tensor_array[1]
|
|
self.assertTrue((np.array(n1) == np.array(t1)).all())
|
|
|
|
n2 = tensor[1:]
|
|
t2 = tensor_array[1:]
|
|
self.assertTrue((np.array(n2) == np.array(t2)).all())
|
|
|
|
n3 = tensor[0:2:]
|
|
t3 = tensor_array[0:2:]
|
|
self.assertTrue((np.array(n3) == np.array(t3)).all())
|
|
|
|
n4 = tensor[2::-2]
|
|
t4 = tensor_array[2::-2]
|
|
self.assertTrue((np.array(n4) == np.array(t4)).all())
|
|
|
|
n5 = tensor[2::-2][0]
|
|
t5 = tensor_array[2::-2][0]
|
|
self.assertTrue((np.array(n5) == np.array(t5)).all())
|
|
|
|
n6 = tensor[2:-1:-1]
|
|
t6 = tensor_array[2:-1:-1]
|
|
self.assertTrue((np.array(n6) == np.array(t6)).all())
|
|
|
|
n7 = tensor[0:, 0:]
|
|
t7 = tensor_array[0:, 0:]
|
|
self.assertTrue((np.array(n7) == np.array(t7)).all())
|
|
|
|
n8 = tensor[0::1, 0::-1, 2:]
|
|
t8 = tensor_array[0::1, 0::-1, 2:]
|
|
self.assertTrue((np.array(n8) == np.array(t8)).all())
|
|
|
|
def test_slice_tensor(self):
|
|
for dtype in self.support_dtypes:
|
|
# run cpu first
|
|
place = core.CPUPlace()
|
|
self.run_slice_tensor(place, dtype)
|
|
|
|
if core.is_compiled_with_cuda() or is_custom_device():
|
|
place = get_device_place()
|
|
self.run_slice_tensor(place, dtype)
|
|
|
|
def test_print_tensor(self):
|
|
scope = core.Scope()
|
|
var = scope.var("test_tensor")
|
|
place = core.CPUPlace()
|
|
tensor = var.get_tensor()
|
|
tensor._set_dims([10, 10])
|
|
tensor._alloc_int(place)
|
|
tensor_array = np.array(tensor)
|
|
self.assertEqual((10, 10), tensor_array.shape)
|
|
tensor_array[0, 0] = 1
|
|
tensor_array[2, 2] = 2
|
|
tensor.set(tensor_array, place)
|
|
print(tensor)
|
|
self.assertTrue(isinstance(str(tensor), str))
|
|
|
|
if core.is_compiled_with_cuda() or is_custom_device():
|
|
tensor.set(tensor_array, get_device_place())
|
|
print(tensor)
|
|
self.assertTrue(isinstance(str(tensor), str))
|
|
|
|
def test_tensor_pointer(self):
|
|
place = core.CPUPlace()
|
|
scope = core.Scope()
|
|
var = scope.var("test_tensor")
|
|
place = core.CPUPlace()
|
|
tensor = var.get_tensor()
|
|
dtype = paddle.float32
|
|
if paddle.framework.use_pir_api() and isinstance(
|
|
dtype, paddle.base.core.DataType
|
|
):
|
|
dtype = paddle.pir.core.datatype_to_vartype[dtype]
|
|
self.assertTrue(
|
|
isinstance(tensor._mutable_data(place, dtype), numbers.Integral)
|
|
)
|
|
|
|
if core.is_compiled_with_cuda():
|
|
place = get_device_place()
|
|
self.assertTrue(
|
|
isinstance(tensor._mutable_data(place, dtype), numbers.Integral)
|
|
)
|
|
place = core.CUDAPinnedPlace()
|
|
self.assertTrue(
|
|
isinstance(tensor._mutable_data(place, dtype), numbers.Integral)
|
|
)
|
|
places = base.cuda_pinned_places()
|
|
self.assertTrue(
|
|
isinstance(
|
|
tensor._mutable_data(places[0], dtype), numbers.Integral
|
|
)
|
|
)
|
|
elif is_custom_device():
|
|
place = get_device_place()
|
|
self.assertTrue(
|
|
isinstance(tensor._mutable_data(place, dtype), numbers.Integral)
|
|
)
|
|
|
|
def test_tensor_set_fp16(self):
|
|
array = np.random.random((300, 500)).astype("float16")
|
|
tensor = base.Tensor()
|
|
place = core.CPUPlace()
|
|
tensor.set(array, place)
|
|
tensor_dtype = tensor._dtype()
|
|
if paddle.framework.use_pir_api() and isinstance(
|
|
tensor_dtype, paddle.base.libpaddle.VarDesc.VarType
|
|
):
|
|
tensor_dtype = paddle.pir.core.vartype_to_datatype[tensor_dtype]
|
|
self.assertEqual(tensor_dtype, paddle.float16)
|
|
np.testing.assert_array_equal(np.array(tensor), array)
|
|
|
|
if core.is_compiled_with_cuda():
|
|
place = get_device_place()
|
|
tensor.set(array, place)
|
|
self.assertEqual(tensor_dtype, paddle.float16)
|
|
np.testing.assert_array_equal(np.array(tensor), array)
|
|
|
|
place = core.CUDAPinnedPlace()
|
|
tensor.set(array, place)
|
|
self.assertEqual(tensor_dtype, paddle.float16)
|
|
np.testing.assert_array_equal(np.array(tensor), array)
|
|
elif is_custom_device():
|
|
place = get_device_place()
|
|
tensor.set(array, place)
|
|
self.assertEqual(tensor_dtype, paddle.float16)
|
|
np.testing.assert_array_equal(np.array(tensor), array)
|
|
|
|
def test_tensor_set_int16(self):
|
|
array = np.random.randint(100, size=(300, 500)).astype("int16")
|
|
tensor = base.Tensor()
|
|
place = core.CPUPlace()
|
|
tensor.set(array, place)
|
|
tensor_dtype = tensor._dtype()
|
|
if paddle.framework.use_pir_api() and isinstance(
|
|
tensor_dtype, paddle.base.libpaddle.VarDesc.VarType
|
|
):
|
|
tensor_dtype = paddle.pir.core.vartype_to_datatype[tensor_dtype]
|
|
self.assertEqual(tensor_dtype, paddle.int16)
|
|
np.testing.assert_array_equal(np.array(tensor), array)
|
|
|
|
if core.is_compiled_with_cuda():
|
|
place = get_device_place()
|
|
tensor.set(array, place)
|
|
self.assertEqual(tensor_dtype, paddle.int16)
|
|
np.testing.assert_array_equal(np.array(tensor), array)
|
|
|
|
place = core.CUDAPinnedPlace()
|
|
tensor.set(array, place)
|
|
self.assertEqual(tensor_dtype, paddle.int16)
|
|
np.testing.assert_array_equal(np.array(tensor), array)
|
|
elif is_custom_device():
|
|
place = get_device_place()
|
|
tensor.set(array, place)
|
|
self.assertEqual(tensor_dtype, paddle.int16)
|
|
np.testing.assert_array_equal(np.array(tensor), array)
|
|
|
|
def test_tensor_set_from_array_list(self):
|
|
array = np.random.randint(1000, size=(200, 300))
|
|
list_array = [array, array]
|
|
tensor = base.Tensor()
|
|
place = core.CPUPlace()
|
|
tensor.set(list_array, place)
|
|
self.assertEqual([2, 200, 300], tensor.shape())
|
|
np.testing.assert_array_equal(np.array(tensor), list_array)
|
|
|
|
if core.is_compiled_with_cuda():
|
|
place = get_device_place()
|
|
tensor.set(list_array, place)
|
|
self.assertEqual([2, 200, 300], tensor.shape())
|
|
np.testing.assert_array_equal(np.array(tensor), list_array)
|
|
|
|
place = core.CUDAPinnedPlace()
|
|
tensor.set(list_array, place)
|
|
self.assertEqual([2, 200, 300], tensor.shape())
|
|
np.testing.assert_array_equal(np.array(tensor), list_array)
|
|
elif is_custom_device():
|
|
place = get_device_place()
|
|
tensor.set(list_array, place)
|
|
self.assertEqual([2, 200, 300], tensor.shape())
|
|
np.testing.assert_array_equal(np.array(tensor), list_array)
|
|
|
|
def test_tensor_set_error(self):
|
|
scope = core.Scope()
|
|
var = scope.var("test_tensor")
|
|
place = core.CPUPlace()
|
|
|
|
tensor = var.get_tensor()
|
|
|
|
exception = None
|
|
try:
|
|
error_array = ["1", "2"]
|
|
tensor.set(error_array, place)
|
|
except ValueError as ex:
|
|
exception = ex
|
|
|
|
self.assertIsNotNone(exception)
|
|
|
|
def test_tensor_set_item_complex128(self):
|
|
array = (
|
|
np.random.random((100, 100)) + 1j * np.random.random((100, 100))
|
|
).astype(np.complex128)
|
|
tensor = base.Tensor()
|
|
place = core.CPUPlace()
|
|
tensor.set(array, place)
|
|
tensor_dtype = tensor._dtype()
|
|
if paddle.framework.use_pir_api() and isinstance(
|
|
tensor_dtype, paddle.base.libpaddle.VarDesc.VarType
|
|
):
|
|
tensor_dtype = paddle.pir.core.vartype_to_datatype[tensor_dtype]
|
|
self.assertEqual(tensor_dtype, paddle.complex128)
|
|
tensor._set_complex128_element(0, 42.1 + 42.1j)
|
|
np.testing.assert_allclose(
|
|
tensor._get_complex128_element(0), 42.1 + 42.1j
|
|
)
|
|
|
|
if core.is_compiled_with_cuda():
|
|
place = get_device_place()
|
|
tensor.set(array, place)
|
|
self.assertEqual(tensor_dtype, paddle.complex128)
|
|
tensor._set_complex128_element(0, 42.1 + 42.1j)
|
|
np.testing.assert_allclose(
|
|
tensor._get_complex128_element(0), 42.1 + 42.1j
|
|
)
|
|
|
|
place = core.CUDAPinnedPlace()
|
|
tensor.set(array, place)
|
|
self.assertEqual(tensor_dtype, paddle.complex128)
|
|
tensor._set_complex128_element(0, 42.1 + 42.1j)
|
|
np.testing.assert_allclose(
|
|
tensor._get_complex128_element(0), 42.1 + 42.1j
|
|
)
|
|
elif is_custom_device():
|
|
place = get_device_place()
|
|
tensor.set(array, place)
|
|
self.assertEqual(tensor_dtype, paddle.complex128)
|
|
tensor._set_complex128_element(0, 42.1 + 42.1j)
|
|
np.testing.assert_allclose(
|
|
tensor._get_complex128_element(0), 42.1 + 42.1j
|
|
)
|
|
|
|
def test_tensor_set_item_complex64(self):
|
|
array = (
|
|
np.random.random((100, 100)) + 1j * np.random.random((100, 100))
|
|
).astype(np.complex64)
|
|
tensor = base.Tensor()
|
|
place = core.CPUPlace()
|
|
tensor.set(array, place)
|
|
tensor_dtype = tensor._dtype()
|
|
if paddle.framework.use_pir_api() and isinstance(
|
|
tensor_dtype, paddle.base.libpaddle.VarDesc.VarType
|
|
):
|
|
tensor_dtype = paddle.pir.core.vartype_to_datatype[tensor_dtype]
|
|
self.assertEqual(tensor_dtype, paddle.complex64)
|
|
tensor._set_complex64_element(0, 42.1 + 42.1j)
|
|
np.testing.assert_allclose(
|
|
np.complex64(tensor._get_complex64_element(0)),
|
|
np.complex64(42.1 + 42.1j),
|
|
)
|
|
|
|
if core.is_compiled_with_cuda():
|
|
place = get_device_place()
|
|
tensor.set(array, place)
|
|
self.assertEqual(tensor_dtype, paddle.complex64)
|
|
tensor._set_complex64_element(0, 42.1 + 42.1j)
|
|
np.testing.assert_allclose(
|
|
np.complex64(tensor._get_complex64_element(0)),
|
|
np.complex64(42.1 + 42.1j),
|
|
)
|
|
|
|
place = core.CUDAPinnedPlace()
|
|
tensor.set(array, place)
|
|
self.assertEqual(tensor_dtype, paddle.complex64)
|
|
tensor._set_complex64_element(0, 42.1 + 42.1j)
|
|
np.testing.assert_allclose(
|
|
np.complex64(tensor._get_complex64_element(0)),
|
|
np.complex64(42.1 + 42.1j),
|
|
)
|
|
elif is_custom_device():
|
|
place = get_device_place()
|
|
tensor.set(array, place)
|
|
self.assertEqual(tensor_dtype, paddle.complex64)
|
|
tensor._set_complex64_element(0, 42.1 + 42.1j)
|
|
np.testing.assert_allclose(
|
|
np.complex64(tensor._get_complex64_element(0)),
|
|
np.complex64(42.1 + 42.1j),
|
|
)
|
|
|
|
|
|
class TestTensorDataSetter(unittest.TestCase):
|
|
def setUp(self):
|
|
paddle.disable_static()
|
|
|
|
def test_same_shape_same_dtype(self):
|
|
x = paddle.tensor([[1, 2], [3, 4]], dtype="float32")
|
|
y = paddle.rand_like(x)
|
|
x.requires_grad = True
|
|
|
|
loss = x.sum()
|
|
loss.backward(retain_graph=True)
|
|
|
|
x_grad_expected = paddle.ones_like(x)
|
|
|
|
np.testing.assert_equal(x.grad.numpy(), x_grad_expected.numpy())
|
|
|
|
x.data = y
|
|
|
|
self.assertEqual(x.data_ptr(), y.data_ptr())
|
|
np.testing.assert_allclose(x.numpy(), y.numpy())
|
|
self.assertEqual(
|
|
x.requires_grad,
|
|
True,
|
|
"x's requires_grad should be True after data setting.",
|
|
)
|
|
|
|
loss.backward()
|
|
x_grad_expected = paddle.ones_like(x) * 2
|
|
|
|
np.testing.assert_equal(x.grad.numpy(), x_grad_expected.numpy())
|
|
|
|
def test_new_shape_same_dtype_same_place(self):
|
|
x: paddle.Tensor = paddle.tensor([[1, 2], [3, 4]], dtype="float32")
|
|
y = paddle.rand([3, 4, 5], dtype="float32")
|
|
x.requires_grad = True
|
|
|
|
loss = x.sum()
|
|
loss.backward()
|
|
|
|
x_grad_expected = paddle.ones_like(x)
|
|
|
|
np.testing.assert_equal(x.grad.numpy(), x_grad_expected.numpy())
|
|
assert x.grad.dtype == x_grad_expected.dtype
|
|
|
|
x.data = y
|
|
|
|
self.assertEqual(x.data_ptr(), y.data_ptr())
|
|
np.testing.assert_allclose(x.numpy(), y.numpy())
|
|
self.assertEqual(
|
|
x.requires_grad,
|
|
True,
|
|
"x's requires_grad should be True after data setting.",
|
|
)
|
|
with self.assertRaises((ValueError, RuntimeError)):
|
|
loss = x.sum()
|
|
loss.backward()
|
|
|
|
x.clear_gradient()
|
|
z = x.sum()
|
|
z.backward()
|
|
x_grad_expected = paddle.ones_like(x)
|
|
np.testing.assert_equal(x.grad.numpy(), x_grad_expected.numpy())
|
|
assert x.grad.dtype == x_grad_expected.dtype
|
|
|
|
def test_same_shape_new_dtype_same_place(self):
|
|
x = paddle.tensor([[1, 2], [3, 4]], dtype="float32")
|
|
x_grad_expected = paddle.ones_like(x)
|
|
y = x.to(paddle.float16)
|
|
x.requires_grad = True
|
|
|
|
loss = x.sum()
|
|
|
|
x.data = y
|
|
|
|
loss.backward()
|
|
np.testing.assert_equal(x.grad.numpy(), x_grad_expected.numpy())
|
|
assert x.grad.dtype == x_grad_expected.dtype
|
|
|
|
x.clear_gradient(False)
|
|
assert x.grad is None
|
|
z = x.sum()
|
|
z.backward()
|
|
x_grad_expected = paddle.ones_like(x)
|
|
np.testing.assert_equal(x.grad.numpy(), x_grad_expected.numpy())
|
|
assert x.dtype == x.grad.dtype
|
|
|
|
def test_same_shape_same_dtype_new_place(self):
|
|
if not paddle.device.is_compiled_with_cuda():
|
|
return
|
|
x = paddle.tensor([[1, 2], [3, 4]], dtype="float32").cuda()
|
|
y = x.cpu()
|
|
x_grad_expected = paddle.ones_like(x)
|
|
x.requires_grad = True
|
|
|
|
loss = x.sum()
|
|
x.data = y.data
|
|
|
|
self.assertEqual(x.data_ptr(), y.data_ptr())
|
|
np.testing.assert_allclose(x.numpy(), y.numpy())
|
|
self.assertEqual(
|
|
x.requires_grad,
|
|
True,
|
|
"x's requires_grad should be True after data setting.",
|
|
)
|
|
loss.backward()
|
|
np.testing.assert_equal(x.grad.numpy(), x_grad_expected.numpy())
|
|
assert x.grad.dtype == x_grad_expected.dtype
|
|
assert x.grad.place == x_grad_expected.place
|
|
|
|
x.clear_gradient(False)
|
|
assert x.grad is None
|
|
loss = x.sum()
|
|
loss.backward()
|
|
x_grad_expected = paddle.ones_like(x)
|
|
np.testing.assert_equal(x.grad.numpy(), x_grad_expected.numpy())
|
|
assert x.grad.place == x.place
|
|
|
|
|
|
class TestTensorNewSharedTensor(unittest.TestCase):
|
|
def test_shared_data(self):
|
|
x = paddle.to_tensor([1.0, 2.0, 3.0])
|
|
y = x._new_shared_tensor()
|
|
x[0] = 88.88
|
|
# Test that they share the same data
|
|
self.assertEqual(y.shape, x.shape)
|
|
self.assertEqual(y.dtype, x.dtype)
|
|
self.assertEqual(y.place, x.place)
|
|
np.testing.assert_allclose(y.numpy(), x.numpy())
|
|
|
|
def test_new_shared_tensor_retain_holder_true(self):
|
|
"""Test _new_shared_tensor with retain_holder=True (default)"""
|
|
x = paddle.to_tensor([1.0, 2.0, 3.0], stop_gradient=False)
|
|
y = x._new_shared_tensor(retain_holder=True)
|
|
|
|
# Test that they share the same data
|
|
self.assertEqual(y.shape, x.shape)
|
|
self.assertEqual(y.dtype, x.dtype)
|
|
self.assertEqual(y.place, x.place)
|
|
np.testing.assert_allclose(y.numpy(), x.numpy())
|
|
|
|
# Test autograd metadata sharing
|
|
self.assertEqual(y.stop_gradient, x.stop_gradient)
|
|
|
|
# Test gradient sharing after backward
|
|
loss = x.sum() + y.sum()
|
|
loss.backward()
|
|
self.assertIsNotNone(x.grad)
|
|
self.assertIsNotNone(y.grad)
|
|
np.testing.assert_allclose(x.grad.numpy(), y.grad.numpy())
|
|
|
|
self.assertEqual(id(x.grad), id(y.grad))
|
|
# Test data sharing - modification affects both
|
|
|
|
def test_new_shared_tensor_retain_holder_false(self):
|
|
"""Test _new_shared_tensor with retain_holder=False"""
|
|
x = paddle.to_tensor([1.0, 2.0, 3.0], stop_gradient=False)
|
|
z = x._new_shared_tensor(retain_holder=False)
|
|
|
|
# Test metadata is the same
|
|
self.assertEqual(z.shape, x.shape)
|
|
self.assertEqual(z.dtype, x.dtype)
|
|
self.assertEqual(z.place, x.place)
|
|
|
|
# Test that new tensor has empty data allocation
|
|
# It should be uninitialized or have default values
|
|
self.assertEqual(z.stop_gradient, x.stop_gradient)
|
|
|
|
# Test autograd metadata is still shared
|
|
loss = x.sum()
|
|
loss.backward()
|
|
self.assertIsNotNone(z.grad)
|
|
|
|
def test_new_shared_tensor_default_behavior(self):
|
|
"""Test _new_shared_tensor with default parameters"""
|
|
x = paddle.to_tensor([[1.0, 2.0], [3.0, 4.0]], stop_gradient=False)
|
|
y = x._new_shared_tensor() # default retain_holder=True
|
|
|
|
# Test gradient calculation
|
|
loss = (x + y).sum()
|
|
loss.backward()
|
|
|
|
np.testing.assert_allclose(x.grad.numpy(), y.grad.numpy())
|
|
|
|
def test_new_shared_tensor_uninitialized_error(self):
|
|
"""Test error when original tensor is not initialized"""
|
|
x = paddle.Tensor()
|
|
x._clear_dataptr() # Ensure Tensor is uninitialized tensor
|
|
with self.assertRaises(ValueError):
|
|
x._new_shared_tensor()
|
|
|
|
|
|
if __name__ == '__main__':
|
|
unittest.main()
|