-
Notifications
You must be signed in to change notification settings - Fork 664
/
ofblob.py
209 lines (181 loc) · 8.16 KB
/
ofblob.py
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
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
"""
Copyright 2020 The OneFlow 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.
"""
from __future__ import absolute_import
import collections
from functools import reduce
import numpy as np
import oneflow as flow
import oneflow.oneflow_internal as oneflow_api
from google.protobuf import text_format
from oneflow.python.framework.dtype import convert_proto_dtype_to_oneflow_dtype
from oneflow.python.lib.core.box import Box
class OfBlob(object):
def __init__(self, of_blob_ptr):
self.of_blob_ptr_ = of_blob_ptr
@property
def dtype(self):
return convert_proto_dtype_to_oneflow_dtype(
oneflow_api.Ofblob_GetDataType(self.of_blob_ptr_)
)
@property
def static_shape(self):
num_axes = oneflow_api.OfBlob_NumAxes(self.of_blob_ptr_)
dst_ndarray = np.ndarray(num_axes, dtype=np.int64)
oneflow_api.OfBlob_CopyStaticShapeTo(self.of_blob_ptr_, dst_ndarray)
return tuple(dst_ndarray.tolist())
@property
def shape(self):
num_axes = oneflow_api.OfBlob_NumAxes(self.of_blob_ptr_)
dst_ndarray = np.zeros(num_axes, dtype=np.int64)
oneflow_api.OfBlob_CopyShapeToNumpy(self.of_blob_ptr_, dst_ndarray)
return tuple(dst_ndarray.tolist())
@property
def shape_list(self):
tensor_shape_list = []
num_axes = oneflow_api.OfBlob_NumAxes(self.of_blob_ptr_)
oneflow_api.OfBlob_ResetTensorIterator(self.of_blob_ptr_)
while not oneflow_api.OfBlob_CurTensorIteratorEqEnd(self.of_blob_ptr_):
shape_tensor = np.zeros(self.num_axes, dtype=np.int64)
oneflow_api.OfBlob_CurTensorCopyShapeTo(self.of_blob_ptr_, shape_tensor)
assert len(shape_tensor.shape) == 1
assert shape_tensor.size == num_axes
tensor_shape_list.append(tuple(shape_tensor.tolist()))
oneflow_api.OfBlob_IncTensorIterator(self.of_blob_ptr_)
return tensor_shape_list
def set_shape(self, shape):
assert isinstance(shape, (list, tuple))
assert len(shape) == oneflow_api.OfBlob_NumAxes(self.of_blob_ptr_)
oneflow_api.OfBlob_CopyShapeFromNumpy(
self.of_blob_ptr_, np.array(shape, dtype=np.int64)
)
@property
def num_axes(self):
return oneflow_api.OfBlob_NumAxes(self.of_blob_ptr_)
@property
def is_dynamic(self):
return oneflow_api.OfBlob_IsDynamic(self.of_blob_ptr_)
@property
def is_tensor_list(self):
return oneflow_api.OfBlob_IsTensorList(self.of_blob_ptr_)
def CopyToNdarray(self):
ndarray_lists = self._CopyToNdarrayLists()
assert len(ndarray_lists) == 1
assert len(ndarray_lists[0]) == 1
return ndarray_lists[0][0]
def CopyToNdarrayLists(self):
return self._CopyToNdarrayLists()
def CopyToFlatNdarrayList(self):
ndarray_lists = self._CopyToNdarrayLists()
ret_ndarray_list = []
for ndarray_list in ndarray_lists:
for ndarray in ndarray_list:
ret_ndarray_list.append(ndarray)
return ret_ndarray_list
def CopyFromNdarray(self, src_ndarray):
if self.is_dynamic:
return self._CopyFromNdarrayLists([[src_ndarray]])
else:
return self._CopyBodyFromNdarray(src_ndarray)
def _CopyBodyFromNdarray(self, src_ndarray):
assert not self.is_dynamic
method_name = oneflow_api.Dtype_GetOfBlobStaticTensorCopyFromBufferFuncName(
self.dtype.oneflow_proto_dtype
)
copy_method = getattr(oneflow_api, method_name)
copy_method(self.of_blob_ptr_, src_ndarray)
def CopyFromNdarrayList(self, src_ndarray_list):
return self._CopyFromNdarrayLists([src_ndarray_list])
def CopyFromNdarrayLists(self, ndarray_lists):
assert self.is_dynamic
return self._CopyFromNdarrayLists(ndarray_lists)
def _CopyToNdarrayLists(self):
(
tensor_list,
is_new_slice_start_mask,
) = self._CopyToNdarrayListAndIsNewSliceStartMask()
tensor_lists = []
for tensor, is_new_slice_start in zip(tensor_list, is_new_slice_start_mask):
if is_new_slice_start:
tensor_lists.append([])
tensor_lists[-1].append(tensor)
return tensor_lists
def _CopyToNdarrayListAndIsNewSliceStartMask(self):
# get tensor list
method_name = oneflow_api.Dtype_GetOfBlobCurTensorCopyToBufferFuncName(
self.dtype.oneflow_proto_dtype
)
copy_method = getattr(oneflow_api, method_name)
tensor_list = []
oneflow_api.OfBlob_ResetTensorIterator(self.of_blob_ptr_)
while oneflow_api.OfBlob_CurTensorIteratorEqEnd(self.of_blob_ptr_) == False:
shape_tensor = np.zeros(self.num_axes, dtype=np.int64)
oneflow_api.OfBlob_CurTensorCopyShapeTo(self.of_blob_ptr_, shape_tensor)
shape = tuple(shape_tensor.tolist())
tensor = np.zeros(
shape, dtype=flow.convert_oneflow_dtype_to_numpy_dtype(self.dtype)
)
copy_method(self.of_blob_ptr_, tensor)
tensor_list.append(tensor)
oneflow_api.OfBlob_IncTensorIterator(self.of_blob_ptr_)
assert len(tensor_list) == oneflow_api.OfBlob_TotalNumOfTensors(
self.of_blob_ptr_
)
# generate is_new_slice_start_mask
is_new_slice_start_mask = [False] * len(tensor_list)
num_slices = oneflow_api.OfBlob_NumOfTensorListSlices(self.of_blob_ptr_)
for x in range(num_slices):
tensor_list_start = oneflow_api.OfBlob_TensorIndex4SliceId(
self.of_blob_ptr_, x
)
assert tensor_list_start >= 0
assert tensor_list_start < len(is_new_slice_start_mask)
is_new_slice_start_mask[tensor_list_start] = True
return tensor_list, is_new_slice_start_mask
def _CopyFromNdarrayLists(self, ndarray_lists):
assert isinstance(ndarray_lists, (list, tuple))
flat_ndarray_list = []
is_new_slice_start_mask = []
for ndarray_list in ndarray_lists:
assert isinstance(ndarray_list, (list, tuple))
for i, ndarray in enumerate(ndarray_list):
assert isinstance(ndarray, np.ndarray)
flat_ndarray_list.append(ndarray)
is_new_slice_start_mask.append(i == 0)
self._CopyFromNdarrayListAndIsNewSliceStartMask(
flat_ndarray_list, is_new_slice_start_mask
)
def _CopyFromNdarrayListAndIsNewSliceStartMask(
self, tensor_list, is_new_slice_start_mask
):
assert len(tensor_list) == len(is_new_slice_start_mask)
method_name = oneflow_api.Dtype_GetOfBlobCurMutTensorCopyFromBufferFuncName(
self.dtype.oneflow_proto_dtype
)
copy_method = getattr(oneflow_api, method_name)
oneflow_api.OfBlob_ClearTensorLists(self.of_blob_ptr_)
for i, tensor in enumerate(tensor_list):
assert tensor.data.contiguous
if is_new_slice_start_mask[i]:
oneflow_api.OfBlob_AddTensorListSlice(self.of_blob_ptr_)
oneflow_api.OfBlob_AddTensor(self.of_blob_ptr_)
assert oneflow_api.OfBlob_CurMutTensorAvailable(self.of_blob_ptr_)
shape_tensor = np.array(tensor.shape, dtype=np.int64)
oneflow_api.OfBlob_CurMutTensorCopyShapeFrom(
self.of_blob_ptr_, shape_tensor
)
copy_method(self.of_blob_ptr_, tensor)
assert len(tensor_list) == oneflow_api.OfBlob_TotalNumOfTensors(
self.of_blob_ptr_
)
num_slices = reduce(lambda a, b: a + b, is_new_slice_start_mask, 0)
assert num_slices == oneflow_api.OfBlob_NumOfTensorListSlices(self.of_blob_ptr_)