未验证 提交 9f5afa62 编写于 作者: H HongyuJia 提交者: GitHub

[Custom Extension] Polish xpu testcase (#49158)

* clean custom_xpu testcase test_static_pe

* use assert_allclose to solve precision error

* adjust precision

* flatten tensor

* fix flatten
上级 72973d5a
...@@ -150,26 +150,29 @@ def custom_relu_double_grad_dynamic(func, device, dtype, np_x, use_func=True): ...@@ -150,26 +150,29 @@ def custom_relu_double_grad_dynamic(func, device, dtype, np_x, use_func=True):
t = paddle.to_tensor(np_x, dtype=dtype, stop_gradient=False) t = paddle.to_tensor(np_x, dtype=dtype, stop_gradient=False)
out = func(t) if use_func else paddle.nn.functional.relu(t) out = func(t) if use_func else paddle.nn.functional.relu(t)
out.stop_gradient = False
dx = paddle.grad( dx = paddle.grad(
outputs=[out], inputs=[t], create_graph=True, retain_graph=True outputs=out,
inputs=t,
grad_outputs=paddle.ones_like(t),
create_graph=True,
retain_graph=True,
) )
dx[0].backward() ddout = paddle.grad(
outputs=dx[0],
assert dx[0].grad is not None inputs=out.grad,
return dx[0].numpy(), dx[0].grad.numpy() grad_outputs=paddle.ones_like(t),
create_graph=False,
)
fluid.set_flags({"FLAGS_retain_grad_for_all_tensor": False}) fluid.set_flags({"FLAGS_retain_grad_for_all_tensor": False})
assert ddout[0].numpy() is not None
return dx[0].numpy(), ddout[0].numpy()
class TestNewCustomOpSetUpInstall(unittest.TestCase): class TestNewCustomOpXpuSetUpInstall(unittest.TestCase):
def setUp(self): def setUp(self):
cur_dir = os.path.dirname(os.path.abspath(__file__)) cur_dir = os.path.dirname(os.path.abspath(__file__))
# compile, install the custom op egg into site-packages under background
# Currently custom XPU op does not support Windows
if os.name == 'nt':
return
cmd = 'cd {} && {} custom_relu_xpu_setup.py install'.format( cmd = 'cd {} && {} custom_relu_xpu_setup.py install'.format(
cur_dir, sys.executable cur_dir, sys.executable
) )
...@@ -192,7 +195,7 @@ class TestNewCustomOpSetUpInstall(unittest.TestCase): ...@@ -192,7 +195,7 @@ class TestNewCustomOpSetUpInstall(unittest.TestCase):
self.custom_op = custom_relu_xpu_module_setup.custom_relu self.custom_op = custom_relu_xpu_module_setup.custom_relu
self.dtypes = ['float32', 'float64'] self.dtypes = ['float32', 'float64']
self.devices = ['xpu'] self.device = 'xpu'
# config seed # config seed
SEED = 2021 SEED = 2021
...@@ -200,12 +203,11 @@ class TestNewCustomOpSetUpInstall(unittest.TestCase): ...@@ -200,12 +203,11 @@ class TestNewCustomOpSetUpInstall(unittest.TestCase):
paddle.framework.random._manual_program_seed(SEED) paddle.framework.random._manual_program_seed(SEED)
def test_static(self): def test_static(self):
for device in self.devices:
for dtype in self.dtypes: for dtype in self.dtypes:
x = np.random.uniform(-1, 1, [4, 8]).astype(dtype) x = np.random.uniform(-1, 1, [4, 8]).astype(dtype)
out = custom_relu_static(self.custom_op, device, dtype, x) out = custom_relu_static(self.custom_op, self.device, dtype, x)
pd_out = custom_relu_static( pd_out = custom_relu_static(
self.custom_op, device, dtype, x, False self.custom_op, self.device, dtype, x, False
) )
np.testing.assert_array_equal( np.testing.assert_array_equal(
out, out,
...@@ -216,30 +218,29 @@ class TestNewCustomOpSetUpInstall(unittest.TestCase): ...@@ -216,30 +218,29 @@ class TestNewCustomOpSetUpInstall(unittest.TestCase):
) )
def test_static_pe(self): def test_static_pe(self):
for device in self.devices:
for dtype in self.dtypes: for dtype in self.dtypes:
x = np.random.uniform(-1, 1, [4, 8]).astype(dtype) x = np.random.uniform(-1, 1, [4, 8]).astype(dtype)
out = custom_relu_static_pe(self.custom_op, device, dtype, x) out = custom_relu_static_pe(self.custom_op, self.device, dtype, x)
pd_out = custom_relu_static_pe( pd_out = custom_relu_static_pe(
self.custom_op, device, dtype, x, False self.custom_op, self.device, dtype, x, False
) )
np.testing.assert_array_equal( np.testing.assert_allclose(
out, out,
pd_out, pd_out,
atol=1e-2,
err_msg='custom op out: {},\n paddle api out: {}'.format( err_msg='custom op out: {},\n paddle api out: {}'.format(
out, pd_out out, pd_out
), ),
) )
def test_dynamic(self): def test_dynamic(self):
for device in self.devices:
for dtype in self.dtypes: for dtype in self.dtypes:
x = np.random.uniform(-1, 1, [4, 8]).astype(dtype) x = np.random.uniform(-1, 1, [4, 8]).astype(dtype)
out, x_grad = custom_relu_dynamic( out, x_grad = custom_relu_dynamic(
self.custom_op, device, dtype, x self.custom_op, self.device, dtype, x
) )
pd_out, pd_x_grad = custom_relu_dynamic( pd_out, pd_x_grad = custom_relu_dynamic(
self.custom_op, device, dtype, x, False self.custom_op, self.device, dtype, x, False
) )
np.testing.assert_array_equal( np.testing.assert_array_equal(
out, out,
...@@ -261,9 +262,9 @@ class TestNewCustomOpSetUpInstall(unittest.TestCase): ...@@ -261,9 +262,9 @@ class TestNewCustomOpSetUpInstall(unittest.TestCase):
np_data = np.random.random((1, 1, 28, 28)).astype("float32") np_data = np.random.random((1, 1, 28, 28)).astype("float32")
np_label = np.random.random((1, 1)).astype("int64") np_label = np.random.random((1, 1)).astype("int64")
path_prefix = "self.custom_op_inference/custom_relu" path_prefix = "self.custom_op_inference/custom_relu"
for device in self.devices:
predict = custom_relu_static_inference( predict = custom_relu_static_inference(
self.custom_op, device, np_data, np_label, path_prefix self.custom_op, self.device, np_data, np_label, path_prefix
) )
# load inference model # load inference model
with static.scope_guard(static.Scope()): with static.scope_guard(static.Scope()):
...@@ -278,9 +279,10 @@ class TestNewCustomOpSetUpInstall(unittest.TestCase): ...@@ -278,9 +279,10 @@ class TestNewCustomOpSetUpInstall(unittest.TestCase):
feed={feed_target_names[0]: np_data}, feed={feed_target_names[0]: np_data},
fetch_list=fetch_targets, fetch_list=fetch_targets,
) )
np.testing.assert_array_equal( np.testing.assert_allclose(
predict, predict,
predict_infer, predict_infer,
atol=1e-2,
err_msg='custom op predict: {},\n custom op infer predict: {}'.format( err_msg='custom op predict: {},\n custom op infer predict: {}'.format(
predict, predict_infer predict, predict_infer
), ),
...@@ -294,14 +296,11 @@ class TestNewCustomOpSetUpInstall(unittest.TestCase): ...@@ -294,14 +296,11 @@ class TestNewCustomOpSetUpInstall(unittest.TestCase):
path_prefix = "self.custom_op_inference/custom_relu" path_prefix = "self.custom_op_inference/custom_relu"
from paddle.inference import Config, create_predictor from paddle.inference import Config, create_predictor
for device in self.devices:
predict = custom_relu_static_inference( predict = custom_relu_static_inference(
self.custom_op, device, np_data, np_label, path_prefix self.custom_op, self.device, np_data, np_label, path_prefix
) )
# load inference model # load inference model
config = Config( config = Config(path_prefix + ".pdmodel", path_prefix + ".pdiparams")
path_prefix + ".pdmodel", path_prefix + ".pdiparams"
)
predictor = create_predictor(config) predictor = create_predictor(config)
input_tensor = predictor.get_input_handle( input_tensor = predictor.get_input_handle(
predictor.get_input_names()[0] predictor.get_input_names()[0]
...@@ -313,23 +312,27 @@ class TestNewCustomOpSetUpInstall(unittest.TestCase): ...@@ -313,23 +312,27 @@ class TestNewCustomOpSetUpInstall(unittest.TestCase):
predictor.get_output_names()[0] predictor.get_output_names()[0]
) )
predict_infer = output_tensor.copy_to_cpu() predict_infer = output_tensor.copy_to_cpu()
self.assertTrue( predict = np.array(predict).flatten()
np.isclose(predict, predict_infer, rtol=5e-5).any(), predict_infer = np.array(predict_infer).flatten()
"custom op predict: {},\n custom op infer predict: {}".format( np.testing.assert_allclose(
predict,
predict_infer,
rtol=5e-5,
atol=1e-2,
err_msg="custom op predict: {},\n custom op infer predict: {}".format(
predict, predict_infer predict, predict_infer
), ),
) )
paddle.disable_static() paddle.disable_static()
def test_func_double_grad_dynamic(self): def test_func_double_grad_dynamic(self):
for device in self.devices:
for dtype in self.dtypes: for dtype in self.dtypes:
x = np.random.uniform(-1, 1, [4, 8]).astype(dtype) x = np.random.uniform(-1, 1, [4, 8]).astype(dtype)
out, dx_grad = custom_relu_double_grad_dynamic( out, dx_grad = custom_relu_double_grad_dynamic(
self.custom_op, device, dtype, x self.custom_op, self.device, dtype, x
) )
pd_out, pd_dx_grad = custom_relu_double_grad_dynamic( pd_out, pd_dx_grad = custom_relu_double_grad_dynamic(
self.custom_op, device, dtype, x, False self.custom_op, self.device, dtype, x, False
) )
np.testing.assert_array_equal( np.testing.assert_array_equal(
out, out,
...@@ -348,8 +351,7 @@ class TestNewCustomOpSetUpInstall(unittest.TestCase): ...@@ -348,8 +351,7 @@ class TestNewCustomOpSetUpInstall(unittest.TestCase):
def test_with_dataloader(self): def test_with_dataloader(self):
paddle.disable_static() paddle.disable_static()
for device in self.devices: paddle.set_device(self.device)
paddle.set_device(device)
# data loader # data loader
transform = Compose( transform = Compose(
[Normalize(mean=[127.5], std=[127.5], data_format='CHW')] [Normalize(mean=[127.5], std=[127.5], data_format='CHW')]
...@@ -368,9 +370,10 @@ class TestNewCustomOpSetUpInstall(unittest.TestCase): ...@@ -368,9 +370,10 @@ class TestNewCustomOpSetUpInstall(unittest.TestCase):
for batch_id, (image, _) in enumerate(train_loader()): for batch_id, (image, _) in enumerate(train_loader()):
out = self.custom_op(image) out = self.custom_op(image)
pd_out = paddle.nn.functional.relu(image) pd_out = paddle.nn.functional.relu(image)
np.testing.assert_array_equal( np.testing.assert_allclose(
out, out,
pd_out, pd_out,
atol=1e-2,
err_msg='custom op out: {},\n paddle api out: {}'.format( err_msg='custom op out: {},\n paddle api out: {}'.format(
out, pd_out out, pd_out
), ),
...@@ -382,4 +385,8 @@ class TestNewCustomOpSetUpInstall(unittest.TestCase): ...@@ -382,4 +385,8 @@ class TestNewCustomOpSetUpInstall(unittest.TestCase):
if __name__ == '__main__': if __name__ == '__main__':
# compile, install the custom op egg into site-packages under background
# Currently custom XPU op does not support Windows
if os.name == 'nt':
exit()
unittest.main() unittest.main()
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册