提交 88c192c8 编写于 作者: M Megvii Engine Team

feat(lite): add get_data_by_share in megenginelite python interface

GitOrigin-RevId: 0ddbb75e823106a61d5802d8e395db99a3e9f1d6
上级 8624ec22
...@@ -8,6 +8,7 @@ ...@@ -8,6 +8,7 @@
# "AS IS" BASIS, WITHOUT ARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # "AS IS" BASIS, WITHOUT ARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
from .base import * from .base import *
from .base import version as __version__
from .global_setting import * from .global_setting import *
from .network import * from .network import *
from .struct import * from .struct import *
......
...@@ -69,7 +69,9 @@ class LiteOptions(Structure): ...@@ -69,7 +69,9 @@ class LiteOptions(Structure):
"const_shape": bool(self.const_shape), "const_shape": bool(self.const_shape),
"force_dynamic_alloc": bool(self.force_dynamic_alloc), "force_dynamic_alloc": bool(self.force_dynamic_alloc),
"force_output_dynamic_alloc": bool(self.force_output_dynamic_alloc), "force_output_dynamic_alloc": bool(self.force_output_dynamic_alloc),
"force_output_nocopy": bool(self.force_output_nocopy), "force_output_use_user_specified_memory": bool(
self.force_output_use_user_specified_memory
),
"no_profiling_on_shape_change": bool(self.no_profiling_on_shape_change), "no_profiling_on_shape_change": bool(self.no_profiling_on_shape_change),
"jit_level": self.jit_level, "jit_level": self.jit_level,
"comp_node_seq_record_level": self.comp_node_seq_record_level, "comp_node_seq_record_level": self.comp_node_seq_record_level,
...@@ -99,7 +101,7 @@ class LiteConfig(Structure): ...@@ -99,7 +101,7 @@ class LiteConfig(Structure):
("device_id", c_int), ("device_id", c_int),
("device_type", c_int), ("device_type", c_int),
("backend", c_int), ("backend", c_int),
("bare_model_cryption_name", c_char_p), ("_bare_model_cryption_name", c_char_p),
("options", LiteOptions), ("options", LiteOptions),
] ]
...@@ -110,18 +112,30 @@ class LiteConfig(Structure): ...@@ -110,18 +112,30 @@ class LiteConfig(Structure):
else: else:
self.options = LiteOptions() self.options = LiteOptions()
self.bare_model_cryption_name = c_char_p(b"") self._bare_model_cryption_name = c_char_p(b"")
self.use_loader_dynamic_param = 0 self.use_loader_dynamic_param = 0
self.has_compression = 0 self.has_compression = 0
self.backend = LiteBackend.LITE_DEFAULT self.backend = LiteBackend.LITE_DEFAULT
@property
def bare_model_cryption_name(self):
return self._bare_model_cryption_name.decode("utf-8")
@bare_model_cryption_name.setter
def bare_model_cryption_name(self, name):
if isinstance(name, str):
self._bare_model_cryption_name = name.encode("utf-8")
else:
assert isinstance(name, bytes), "name should be str or bytes type."
self._bare_model_cryption_name = name
def __repr__(self): def __repr__(self):
data = { data = {
"has_compression": bool(self.has_compression), "has_compression": bool(self.has_compression),
"device_id": LiteDeviceType(self.device_id), "device_id": LiteDeviceType(self.device_id),
"device_type": LiteDeviceType(self.device_type), "device_type": LiteDeviceType(self.device_type),
"backend": LiteBackend(self.backend), "backend": LiteBackend(self.backend),
"bare_model_cryption_name": self.bare_model_cryption_name.decode("utf-8"), "bare_model_cryption_name": self.bare_model_cryption_name,
"options": self.options, "options": self.options,
} }
return data.__repr__() return data.__repr__()
...@@ -149,7 +163,7 @@ class LiteIO(Structure): ...@@ -149,7 +163,7 @@ class LiteIO(Structure):
""" """
_fields_ = [ _fields_ = [
("name", c_char_p), ("_name", c_char_p),
("is_host", c_int), ("is_host", c_int),
("io_type", c_int), ("io_type", c_int),
("config_layout", LiteLayout), ("config_layout", LiteLayout),
...@@ -159,9 +173,9 @@ class LiteIO(Structure): ...@@ -159,9 +173,9 @@ class LiteIO(Structure):
self, name, is_host=True, io_type=LiteIOType.LITE_IO_VALUE, layout=None self, name, is_host=True, io_type=LiteIOType.LITE_IO_VALUE, layout=None
): ):
if type(name) == str: if type(name) == str:
self.name = c_char_p(name.encode("utf-8")) self._name = c_char_p(name.encode("utf-8"))
else: else:
self.name = c_char_p(name) self._name = c_char_p(name)
if layout: if layout:
self.config_layout = layout self.config_layout = layout
...@@ -171,6 +185,18 @@ class LiteIO(Structure): ...@@ -171,6 +185,18 @@ class LiteIO(Structure):
self.is_host = is_host self.is_host = is_host
self.io_type = io_type self.io_type = io_type
@property
def name(self):
return self._name.decode("utf-8")
@name.setter
def name(self, name):
if isinstance(name, str):
self._name = name.encode("utf-8")
else:
assert isinstance(name, bytes), "name should be str or bytes type."
self._name = name
def __repr__(self): def __repr__(self):
data = { data = {
"name": self.name, "name": self.name,
...@@ -208,17 +234,45 @@ class LiteNetworkIO(object): ...@@ -208,17 +234,45 @@ class LiteNetworkIO(object):
the input and output information for user to construct _LiteNetWorkIO the input and output information for user to construct _LiteNetWorkIO
""" """
def __init__(self): def __init__(self, inputs=None, outputs=None):
self.inputs = [] self.inputs = []
self.outputs = [] self.outputs = []
if inputs:
for i in inputs:
if isinstance(i, list):
self.inputs.append(LiteIO(*i))
else:
assert isinstance(
i, LiteIO
), "the param to construct LiteNetworkIO must be list of the LiteIO member or the LiteIO."
self.inputs.append(i)
if outputs:
for i in outputs:
if isinstance(i, list):
self.outputs.append(LiteIO(*i))
else:
assert isinstance(
i, LiteIO
), "the param to construct LiteNetworkIO must be list of the LiteIO member or the LiteIO."
self.outputs.append(i)
def add_input(self, input_io): def add_input(
assert isinstance(input_io, LiteIO) self, obj, is_host=True, io_type=LiteIOType.LITE_IO_VALUE, layout=None
self.inputs.append(input_io) ):
if isinstance(obj, LiteIO):
self.inputs.append(obj)
else:
name = obj
self.add_input(LiteIO(name, is_host, io_type, layout))
def add_output(self, output_io): def add_output(
assert isinstance(output_io, LiteIO) self, obj, is_host=True, io_type=LiteIOType.LITE_IO_VALUE, layout=None
self.outputs.append(output_io) ):
if isinstance(obj, LiteIO):
self.outputs.append(obj)
else:
name = obj
self.add_output(LiteIO(name, is_host, io_type, layout))
def _create_network_io(self): def _create_network_io(self):
network_io = _LiteNetworkIO() network_io = _LiteNetworkIO()
......
...@@ -48,6 +48,15 @@ ctype_to_lite_dtypes = { ...@@ -48,6 +48,15 @@ ctype_to_lite_dtypes = {
c_ushort: LiteDataType.LITE_UINT16, c_ushort: LiteDataType.LITE_UINT16,
} }
_lite_dtypes_to_ctype = {
LiteDataType.LITE_INT: c_int,
LiteDataType.LITE_FLOAT: c_float,
LiteDataType.LITE_UINT8: c_ubyte,
LiteDataType.LITE_INT8: c_byte,
LiteDataType.LITE_INT16: c_short,
LiteDataType.LITE_UINT16: c_ushort,
}
class LiteLayout(Structure): class LiteLayout(Structure):
""" """
...@@ -55,7 +64,7 @@ class LiteLayout(Structure): ...@@ -55,7 +64,7 @@ class LiteLayout(Structure):
""" """
_fields_ = [ _fields_ = [
("shapes", c_size_t * MAX_DIM), ("_shapes", c_size_t * MAX_DIM),
("ndim", c_size_t), ("ndim", c_size_t),
("data_type", c_int), ("data_type", c_int),
] ]
...@@ -64,10 +73,10 @@ class LiteLayout(Structure): ...@@ -64,10 +73,10 @@ class LiteLayout(Structure):
if shape: if shape:
shape = list(shape) shape = list(shape)
assert len(shape) <= MAX_DIM, "Layout max dim is 7." assert len(shape) <= MAX_DIM, "Layout max dim is 7."
self.shapes = (c_size_t * MAX_DIM)(*shape) self._shapes = (c_size_t * MAX_DIM)(*shape)
self.ndim = len(shape) self.ndim = len(shape)
else: else:
self.shapes = (c_size_t * MAX_DIM)() self._shapes = (c_size_t * MAX_DIM)()
self.ndim = 0 self.ndim = 0
if not dtype: if not dtype:
self.data_type = LiteDataType.LITE_FLOAT self.data_type = LiteDataType.LITE_FLOAT
...@@ -83,9 +92,24 @@ class LiteLayout(Structure): ...@@ -83,9 +92,24 @@ class LiteLayout(Structure):
else: else:
raise RuntimeError("unkonw data type") raise RuntimeError("unkonw data type")
@property
def dtype(self):
return _lite_type_to_nptypes[LiteDataType(self.data_type)]
@property
def shapes(self):
return list(self._shapes)[0 : self.ndim]
@shapes.setter
def shapes(self, shape):
shape = list(shape)
assert len(shape) <= MAX_DIM, "Layout max dim is 7."
self._shapes = (c_size_t * MAX_DIM)(*shape)
self.ndim = len(shape)
def __repr__(self): def __repr__(self):
data = { data = {
"shapes": list(self.shapes)[0 : self.ndim], "shapes": self.shapes,
"ndim": self.ndim, "ndim": self.ndim,
"data_type": _lite_type_to_nptypes[LiteDataType(self.data_type)], "data_type": _lite_type_to_nptypes[LiteDataType(self.data_type)],
} }
...@@ -177,15 +201,20 @@ class LiteTensor(object): ...@@ -177,15 +201,20 @@ class LiteTensor(object):
device_type=LiteDeviceType.LITE_CPU, device_type=LiteDeviceType.LITE_CPU,
device_id=0, device_id=0,
is_pinned_host=False, is_pinned_host=False,
shapes=None,
dtype=None,
): ):
""" """
create a Tensor with layout, device, is_pinned_host param create a Tensor with layout, device, is_pinned_host or shapes, dtype,
device_type, device_id, is_pinned_host param
""" """
self._tensor = _Ctensor() self._tensor = _Ctensor()
if layout:
self._layout = layout
else:
self._layout = LiteLayout() self._layout = LiteLayout()
if layout is not None:
self._layout = layout
elif shapes is not None:
shapes = list(shapes)
self._layout = LiteLayout(shapes, dtype)
self._device_type = device_type self._device_type = device_type
self._device_id = device_id self._device_id = device_id
self._is_pinned_host = is_pinned_host self._is_pinned_host = is_pinned_host
...@@ -222,9 +251,12 @@ class LiteTensor(object): ...@@ -222,9 +251,12 @@ class LiteTensor(object):
@layout.setter @layout.setter
def layout(self, layout): def layout(self, layout):
assert isinstance(layout, LiteLayout) if isinstance(layout, LiteLayout):
self._layout = layout self._layout = layout
self._api.LITE_set_tensor_layout(self._tensor, layout) elif isinstance(layout, list):
self._layout.shapes = layout
self._api.LITE_set_tensor_layout(self._tensor, self._layout)
@property @property
def is_pinned_host(self): def is_pinned_host(self):
...@@ -270,7 +302,6 @@ class LiteTensor(object): ...@@ -270,7 +302,6 @@ class LiteTensor(object):
""" """
get the length of the meomry in byte get the length of the meomry in byte
""" """
self.update()
length = c_size_t() length = c_size_t()
self._api.LITE_get_tensor_total_size_in_byte(self._tensor, byref(length)) self._api.LITE_get_tensor_total_size_in_byte(self._tensor, byref(length))
return length.value return length.value
...@@ -336,7 +367,6 @@ class LiteTensor(object): ...@@ -336,7 +367,6 @@ class LiteTensor(object):
""" """
get the memory of the tensor, return c_void_p of the tensor memory get the memory of the tensor, return c_void_p of the tensor memory
""" """
self.update()
mem = c_void_p() mem = c_void_p()
self._api.LITE_get_tensor_memory(self._tensor, byref(mem)) self._api.LITE_get_tensor_memory(self._tensor, byref(mem))
return mem return mem
...@@ -347,7 +377,6 @@ class LiteTensor(object): ...@@ -347,7 +377,6 @@ class LiteTensor(object):
param data: the data will shared to the tensor, it should be a param data: the data will shared to the tensor, it should be a
numpy.ndarray or ctypes data numpy.ndarray or ctypes data
""" """
self.update()
if isinstance(data, np.ndarray): if isinstance(data, np.ndarray):
assert ( assert (
self.is_continue self.is_continue
...@@ -356,8 +385,7 @@ class LiteTensor(object): ...@@ -356,8 +385,7 @@ class LiteTensor(object):
self.is_pinned_host or self.device_type == LiteDeviceType.LITE_CPU self.is_pinned_host or self.device_type == LiteDeviceType.LITE_CPU
), "set_data_by_share can only apply in cpu tensor or pinned tensor." ), "set_data_by_share can only apply in cpu tensor or pinned tensor."
np_type = _lite_type_to_nptypes[LiteDataType(self._layout.data_type)] c_type = _lite_dtypes_to_ctype[LiteDataType(self._layout.data_type)]
c_type = np.ctypeslib.as_ctypes_type(np_type)
if self.nbytes != data.nbytes: if self.nbytes != data.nbytes:
self.layout = LiteLayout(data.shape, ctype_to_lite_dtypes[c_type]) self.layout = LiteLayout(data.shape, ctype_to_lite_dtypes[c_type])
...@@ -377,7 +405,6 @@ class LiteTensor(object): ...@@ -377,7 +405,6 @@ class LiteTensor(object):
param data: the data to copy to tensor, it should be list, param data: the data to copy to tensor, it should be list,
numpy.ndarraya or ctypes with length numpy.ndarraya or ctypes with length
""" """
self.update()
if layout is not None: if layout is not None:
self.layout = layout self.layout = layout
...@@ -386,8 +413,7 @@ class LiteTensor(object): ...@@ -386,8 +413,7 @@ class LiteTensor(object):
self.is_pinned_host or self.device_type == LiteDeviceType.LITE_CPU self.is_pinned_host or self.device_type == LiteDeviceType.LITE_CPU
), "set_data_by_copy can only apply in cpu tensor or pinned tensor." ), "set_data_by_copy can only apply in cpu tensor or pinned tensor."
np_type = _lite_type_to_nptypes[LiteDataType(self._layout.data_type)] c_type = _lite_dtypes_to_ctype[LiteDataType(self._layout.data_type)]
c_type = np.ctypeslib.as_ctypes_type(np_type)
tensor_memory = c_void_p() tensor_memory = c_void_p()
...@@ -415,6 +441,22 @@ class LiteTensor(object): ...@@ -415,6 +441,22 @@ class LiteTensor(object):
self._api.LITE_get_tensor_memory(self._tensor, byref(tensor_memory)) self._api.LITE_get_tensor_memory(self._tensor, byref(tensor_memory))
memmove(tensor_memory, data, data_length) memmove(tensor_memory, data, data_length)
def get_data_by_share(self):
"""
get the data in the tensor, add share the data with a new numpy, and
return the numpy arrray, be careful, the data in numpy is valid before
the tensor memory is write again, such as LiteNetwok forward next time.
"""
assert self.is_continue, "get_data_by_share can only apply in continue tensor."
assert (
self.is_pinned_host or self.device_type == LiteDeviceType.LITE_CPU
), "get_data_by_share can only apply in CPU tensor or cpu pinned tensor."
memory = self.get_ctypes_memory()
c_type = _lite_dtypes_to_ctype[LiteDataType(self._layout.data_type)]
pnt = cast(memory, POINTER(c_type))
return np.ctypeslib.as_array(pnt, self._layout.shapes)
def to_numpy(self): def to_numpy(self):
""" """
get the buffer of the tensor get the buffer of the tensor
...@@ -475,3 +517,13 @@ def LiteTensorConcat( ...@@ -475,3 +517,13 @@ def LiteTensorConcat(
) )
result_tensor.update() result_tensor.update()
return result_tensor return result_tensor
def lite_dtype_2_numpy(dtype):
"""
convert lite dtype to corresponding numpy dtype
"""
assert isinstance(
dtype, LiteDataType
), "input must be LiteDataType when using lite_dtype_2_numpy."
return _lite_type_to_nptypes[dtype]
...@@ -21,6 +21,12 @@ def test_version(): ...@@ -21,6 +21,12 @@ def test_version():
print("Lite verson: {}".format(version)) print("Lite verson: {}".format(version))
def test_config():
config = LiteConfig()
config.bare_model_cryption_name = "nothing"
print(config)
def test_network_io(): def test_network_io():
input_io1 = LiteIO("data1", is_host=False, io_type=LiteIOType.LITE_IO_VALUE) input_io1 = LiteIO("data1", is_host=False, io_type=LiteIOType.LITE_IO_VALUE)
input_io2 = LiteIO( input_io2 = LiteIO(
...@@ -32,6 +38,7 @@ def test_network_io(): ...@@ -32,6 +38,7 @@ def test_network_io():
io = LiteNetworkIO() io = LiteNetworkIO()
io.add_input(input_io1) io.add_input(input_io1)
io.add_input(input_io2) io.add_input(input_io2)
io.add_input("data3", False)
output_io1 = LiteIO("out1", is_host=False) output_io1 = LiteIO("out1", is_host=False)
output_io2 = LiteIO("out2", is_host=True, layout=LiteLayout([1, 1000])) output_io2 = LiteIO("out2", is_host=True, layout=LiteLayout([1, 1000]))
...@@ -39,7 +46,7 @@ def test_network_io(): ...@@ -39,7 +46,7 @@ def test_network_io():
io.add_output(output_io1) io.add_output(output_io1)
io.add_output(output_io2) io.add_output(output_io2)
assert len(io.inputs) == 2 assert len(io.inputs) == 3
assert len(io.outputs) == 2 assert len(io.outputs) == 2
assert io.inputs[0] == input_io1 assert io.inputs[0] == input_io1
...@@ -47,9 +54,25 @@ def test_network_io(): ...@@ -47,9 +54,25 @@ def test_network_io():
c_io = io._create_network_io() c_io = io._create_network_io()
assert c_io.input_size == 2 assert c_io.input_size == 3
assert c_io.output_size == 2 assert c_io.output_size == 2
ins = [["data1", True], ["data2", False, LiteIOType.LITE_IO_SHAPE]]
outs = [["out1", True], ["out2", False, LiteIOType.LITE_IO_VALUE]]
io2 = LiteNetworkIO(ins, outs)
assert len(io2.inputs) == 2
assert len(io2.outputs) == 2
io3 = LiteNetworkIO([input_io1, input_io2], [output_io1, output_io2])
assert len(io3.inputs) == 2
assert len(io3.outputs) == 2
test_io = LiteIO("test")
assert test_io.name == "test"
test_io.name = "test2"
assert test_io.name == "test2"
class TestShuffleNet(unittest.TestCase): class TestShuffleNet(unittest.TestCase):
source_dir = os.getenv("LITE_TEST_RESOURCE") source_dir = os.getenv("LITE_TEST_RESOURCE")
...@@ -319,9 +342,9 @@ class TestNetwork(TestShuffleNet): ...@@ -319,9 +342,9 @@ class TestNetwork(TestShuffleNet):
data = ios[key].to_numpy().flatten() data = ios[key].to_numpy().flatten()
input_data = self.input_data.flatten() input_data = self.input_data.flatten()
assert data.size == input_data.size assert data.size == input_data.size
assert io.name.decode("utf-8") == "data" assert io.name == "data"
for i in range(data.size): for i in range(data.size):
assert data[i] == input_data[i] assert abs(data[i] - input_data[i]) < 1e-5
return 0 return 0
network.set_start_callback(start_callback) network.set_start_callback(start_callback)
...@@ -343,7 +366,7 @@ class TestNetwork(TestShuffleNet): ...@@ -343,7 +366,7 @@ class TestNetwork(TestShuffleNet):
output_data = self.correct_data.flatten() output_data = self.correct_data.flatten()
assert data.size == output_data.size assert data.size == output_data.size
for i in range(data.size): for i in range(data.size):
assert data[i] == output_data[i] assert abs(data[i] - output_data[i]) < 1e-5
return 0 return 0
network.set_finish_callback(finish_callback) network.set_finish_callback(finish_callback)
...@@ -404,3 +427,27 @@ class TestNetwork(TestShuffleNet): ...@@ -404,3 +427,27 @@ class TestNetwork(TestShuffleNet):
binary_equal_between_batch=True, binary_equal_between_batch=True,
) )
self.do_forward(network) self.do_forward(network)
def test_device_tensor_no_copy(self):
# construct LiteOption
net_config = LiteConfig()
net_config.options.force_output_use_user_specified_memory = True
network = LiteNetwork(config=net_config)
network.load(self.model_path)
input_tensor = network.get_io_tensor("data")
# fill input_data with device data
input_tensor.set_data_by_share(self.input_data)
output_tensor = network.get_io_tensor(network.get_output_name(0))
out_array = np.zeros(output_tensor.layout.shapes, output_tensor.layout.dtype)
output_tensor.set_data_by_share(out_array)
# inference
for i in range(2):
network.forward()
network.wait()
self.check_correct(out_array)
...@@ -54,6 +54,16 @@ def test_tensor_make(): ...@@ -54,6 +54,16 @@ def test_tensor_make():
tensor = LiteTensor(layout, device_id=1) tensor = LiteTensor(layout, device_id=1)
assert tensor.device_id == 1 assert tensor.device_id == 1
tensor.layout = [8, 14]
assert tensor.layout.shapes[0] == 8
assert tensor.layout.shapes[1] == 14
assert tensor.layout.data_type == LiteDataType.LITE_FLOAT
tensor_new = LiteTensor(shapes=[1, 3, 224], dtype=np.int8)
assert tensor_new.layout.shapes[1] == 3
assert tensor_new.layout.shapes[2] == 224
assert tensor_new.layout.data_type == LiteDataType.LITE_INT8
def test_tensor_set_data(): def test_tensor_set_data():
layout = LiteLayout([2, 16], "int8") layout = LiteLayout([2, 16], "int8")
...@@ -292,3 +302,24 @@ def test_tensor_concat(): ...@@ -292,3 +302,24 @@ def test_tensor_concat():
for i in range(128): for i in range(128):
index = j * 128 + i index = j * 128 + i
assert real_data[index // 32][index % 32] == j assert real_data[index // 32][index % 32] == j
def test_tensor_get_memory_by_share():
layout = LiteLayout([4, 32], "int16")
tensor = LiteTensor(layout)
assert tensor.nbytes == 4 * 32 * 2
arr = np.ones([4, 32], "int16")
for i in range(128):
arr[i // 32][i % 32] = i
tensor.set_data_by_copy(arr)
test_data = tensor.get_data_by_share()
real_data = tensor.to_numpy()
for i in range(128):
assert real_data[i // 32][i % 32] == test_data[i // 32][i % 32]
arr[1][18] = 5
arr[3][7] = 345
tensor.set_data_by_copy(arr)
assert test_data[1][18] == 5
assert test_data[3][7] == 345
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