-
Notifications
You must be signed in to change notification settings - Fork 243
/
util.py
372 lines (341 loc) · 14.4 KB
/
util.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
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
#
# -*- coding: utf-8 -*-
#
# Copyright (c) 2021 Intel Corporation
#
# 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 os
import numpy as np
from onnx import helper, numpy_helper
from onnx import onnx_pb as onnx_proto
from enum import Enum
from pathlib import Path
import abc
__producer__ = "onnx.quantize"
__version__ = "0.1.0"
onnx_domain = "ai.onnx"
ms_domain = "com.microsoft"
support_pair = {
'uint8 uint8': True,
'2 2': True,
'float16 float16': True,
'10 10': True,
'float32 float16': True,
'1 10': True,
'float16 float32': True,
'10 1': True
}
dtype_mapping = {
'fp32': 1,
'uint8': 2,
'int8': 3,
'uint16': 4,
'int16': 5,
'int32': 6,
'int64': 7,
'string': 8,
'bool': 9,
'fp16': 10,
'double': 11,
'uint32': 12,
'uint64': 13,
'complex64': 14,
'complex128': 15,
}
def dtype_to_name(dtype_mapping, dtype):
return list(dtype_mapping.keys())[list(dtype_mapping.values()).index(dtype)]
class QuantType(Enum): # pragma: no cover
QInt8 = 0
QUInt8 = 1
def make_quant_node(name, inputs, outputs):
return helper.make_node("QuantizeLinear", inputs, outputs, name)
def make_dquant_node(name, inputs, outputs):
return helper.make_node("DequantizeLinear", inputs, outputs, name)
def _get_qrange_for_qType(qType, reduce_range=False):
'''
Helper function to get the quantization range for a type.
parameter qType: quantization type.
return: quantization range.
'''
if qType == onnx_proto.TensorProto.UINT8:
return 127 if reduce_range else 255
elif qType == onnx_proto.TensorProto.INT8:
# [-64, 64] for reduce_range, and [-127, 127] full_range.
return 128 if reduce_range else 254
else:
raise ValueError('unsupported quantization data type')
def split_shared_bias(model):
for input_name, node_list in model.input_name_to_nodes.items():
if len(node_list) > 1 and input_name in [i.name for i in model.model.graph.initializer]:
for node in node_list[1:]:
if node.op_type not in ['Conv', 'FusedConv']:
continue
if len(node.input) > 2 and node.input[2] == input_name:
new_input_name = node.input[2] + '_nc_split_' + node.name
new_input = helper.make_tensor(
new_input_name,
model.get_initializer(input_name).data_type,
model.get_initializer(input_name).dims,
model.get_initializer(input_name).raw_data,
True)
model.add_initializer(new_input)
node.input[2] = new_input_name
return model
def cast_tensor(tensor, dtype): # pragma: no cover
'''
Convert tensor float to target dtype.
param tensor: TensorProto object
return tensor_target_dtype: converted TensorProto object
'''
if not isinstance(tensor, onnx_proto.TensorProto):
raise ValueError('Expected input type is an ONNX TensorProto but got %s' % type(tensor))
if tensor.data_type == onnx_proto.TensorProto.FLOAT:
new_tensor = helper.make_tensor(
name=tensor.name,
data_type=dtype_mapping[dtype],
dims=numpy_helper.to_array(tensor).shape,
vals=numpy_helper.to_array(tensor)
)
return new_tensor
return None
def remove_init_from_model_input(model):
inputs = model.model.graph.input
name_to_input = {}
for inp in inputs:
name_to_input[inp.name] = inp
for initializer in model.model.graph.initializer:
if initializer.name in name_to_input:
inputs.remove(name_to_input[initializer.name])
def collate_preds(results):
batch = results[0]
if isinstance(batch, list):
results = zip(*results)
collate_results = []
for output in results:
collate_results.append(np.concatenate(output))
elif isinstance(batch, np.ndarray):
collate_results = np.concatenate(results)
return collate_results
def quantize_data_with_scale_zero(data, qType, scheme, scale, zero_point):
'''
:parameter data: data to quantize
:parameter qType: data type to quantize to. Supported types UINT8 and INT8
:parameter scheme: sym or asym quantization.
:parameter scale: computed scale of quantized data
:parameter zero_point: computed zero point of quantized data
:return: quantized weights
To pack weights, we compute a linear transformation
- when data type == uint8 mode, from [rmin, rmax] -> [0, 2^{b-1}] and
- when data type == int8, from [-m , m] -> [-(2^{b-1}-1), 2^{b-1}-1] where
m = max(abs(rmin), abs(rmax))
'''
data = np.asarray(data)
if qType == onnx_proto.TensorProto.INT8 and scheme == 'sym':
# signed byte type
quantized_data = (data.astype(np.float32) / scale).round().astype('b')
elif qType == onnx_proto.TensorProto.UINT8 and scheme == 'asym':
quantized_data = ((data.astype(np.float32) / scale).round() + zero_point).astype('B')
else:
raise ValueError("Unexpected combination of data type {} and scheme {}.".format(
qType, scheme))
return quantized_data
def quantize_data(data, quantize_range, qType, scheme):
'''
:parameter data: data to quantize
:parameter quantize_range: list of data to weight pack.
:parameter qType: data type to quantize to. Supported types UINT8 and INT8
:param scheme: sym or asym quantization.
:return: minimum, maximum, zero point, scale, and quantized weights
To pack weights, we compute a linear transformation
- when data type == uint8 mode, from [rmin, rmax] -> [0, 2^{b-1}] and
- when data type == int8, from [-m , m] -> [-(2^{b-1}-1), 2^{b-1}-1] where
m = max(abs(rmin), abs(rmax))
and add necessary intermediate nodes to trasnform quantized weight to full weight
using the equation r = S(q-z), where
r: real original value
q: quantized value
S: scale
z: zero point
'''
rmin = min(min(data), 0)
rmax = max(max(data), 0)
if scheme == 'sym' and qType == onnx_proto.TensorProto.INT8:
max_range = max(abs(rmin), abs(rmax))
scale = (float(max_range) * 2) / quantize_range if max_range > 0 else 1
zero_point = 0
elif scheme == 'asym' and qType == onnx_proto.TensorProto.UINT8:
scale = (float(rmax) - rmin) / quantize_range if rmin != rmax else 1
zero_point = round((0 - rmin) / scale)
else:
raise ValueError("Unexpected combination of data type {} and scheme {}.".format(
qType, scheme))
quantized_data = quantize_data_with_scale_zero(data, qType, scheme, scale, zero_point)
return rmin, rmax, zero_point, scale, quantized_data
def quantize_data_per_channel(tensor_value, qType, scheme, scale_value, zo_value):
channel_count = tensor_value.shape[0] # TBD, default from axis 0
new_per_channel_tensor_values = []
for i in range(channel_count):
per_channel_tensor_value = tensor_value.take(i, 0)
per_channel_scale_value = scale_value.take(i)
per_channel_zero_value = zo_value.take(i)
new_per_channel_tensor_values.append(quantize_data_with_scale_zero(\
per_channel_tensor_value,
qType,
scheme,
per_channel_scale_value,
per_channel_zero_value))
# combine per_channel_data into one
reshape_dims = list(tensor_value.shape) # deep copy
reshape_dims[0] = 1 # only one per channel for reshape
new_tensor_value = new_per_channel_tensor_values[0].reshape(reshape_dims)
for i in range(1, channel_count):
new_per_channel_tensor_value = new_per_channel_tensor_values[i].\
reshape(reshape_dims)
new_tensor_value = np.concatenate((new_tensor_value, \
new_per_channel_tensor_value), 0)
return new_tensor_value
def dequantize_data_with_scale_zero(tensor_value, scale_value, zo_value): # pragma: no cover
return (tensor_value.astype(np.float32) - zo_value.astype(np.float32)) * scale_value
def dequantize_data(tensor_value, scale_value, zo_value, axis=0): # pragma: no cover
if scale_value.size == 1:
return dequantize_data_with_scale_zero(tensor_value, scale_value, zo_value)
else:
channel_count = tensor_value.shape[axis] # TBD, default from axis 0
new_per_channel_tensor_values = []
for i in range(channel_count):
per_channel_tensor_value = tensor_value.take(i, 0)
per_channel_scale_value = scale_value.take(i)
per_channel_zero_value = zo_value.take(i)
new_per_channel_tensor_values.append(dequantize_data_with_scale_zero(\
per_channel_tensor_value,
per_channel_scale_value,
per_channel_zero_value))
# combine per_channel_data into one
reshape_dims = list(tensor_value.shape) # deep copy
reshape_dims[0] = 1 # only one per channel for reshape
new_tensor_value = new_per_channel_tensor_values[0].reshape(reshape_dims)
for i in range(1, channel_count):
new_per_channel_tensor_value = new_per_channel_tensor_values[i].\
reshape(reshape_dims)
new_tensor_value = np.concatenate((new_tensor_value, \
new_per_channel_tensor_value), 0)
return new_tensor_value
class ValueInfo: # pragma: no cover
def __init__(self,
tensor_name,
dtype,
new_dtype):
self.tensor_name = tensor_name
self.dtype = dtype
self.new_dtype = new_dtype
class QuantizedValue:
'''
Represents a linearly quantized value (input\output\intializer)
'''
def __init__(self,
name,
new_quantized_name,
scale_name,
zero_point_name,
quantized_value_type,
axis=None,
qType=QuantType.QUInt8):
self.name = name
self.q_name = new_quantized_name
self.scale_name = scale_name
self.zp_name = zero_point_name
self.value_type = quantized_value_type
self.axis = axis
self.qType = qType
class QuantizedInitializer:
'''
Represents a linearly quantized weight input from ONNX operators
'''
def __init__(self,
name,
initializer,
rmins,
rmaxs,
zero_points,
scales,
data=[],
quantized_data=[],
axis=None,
qType=QuantType.QUInt8):
self.name = name
self.initializer = initializer # TensorProto initializer in ONNX graph
self.rmins = rmins # List of minimum range for each axis
self.rmaxs = rmaxs # List of maximum range for each axis
# 1D tensor of zero points computed for each axis. scalar if axis is empty
self.zero_points = zero_points
self.scales = scales # 1D tensor of scales computed for each axis. scalar if axis is empty
self.data = data # original data from initializer TensorProto
self.quantized_data = quantized_data # weight-packed data from data
# Scalar to specify which dimension in the initializer to weight pack.
self.axis = axis
# If empty, single zero point and scales computed from a single rmin and rmax
self.qType = qType
class QuantizationMode(Enum): # pragma: no cover
IntegerOps = 0
QLinearOps = 1
class QuantizedValueType(Enum): # pragma: no cover
Input = 0
Initializer = 1
class QuantFormat(Enum): # pragma: no cover
QOperator = 0
QDQ = 1
def quantize_nparray(qtype, arr, scale, zero_point, low=None, high=None):
dtype = np.uint8 if qtype == "uint8" else np.int8
cliplow = max(0 if dtype == np.uint8 else -127, -127 if low is None else low)
cliphigh = min(255 if dtype == np.uint8 else 127, 255 if high is None else high)
arr_fp32 = np.asarray((arr.astype(np.float32) / scale).round() + zero_point)
np.clip(arr_fp32, cliplow, cliphigh, out=arr_fp32)
return arr_fp32.astype(dtype)
def attribute_to_kwarg(attribute):
'''
Convert attribute to kwarg format for use with onnx.helper.make_node.
'''
attribute_mapping = {
1: attribute.f,
2: attribute.i,
3: attribute.s,
4: attribute.t,
5: attribute.g,
6: attribute.floats,
7: attribute.ints,
8: attribute.strings,
9: attribute.tensors,
10: attribute.graphs
}
if attribute.type in attribute_mapping:
value = attribute_mapping[attribute.type]
else: # pragma: no cover
raise ValueError(
'attribute {} has no type specified '
'or unsupported type {}.'.format(attribute.name, attribute.type))
return {attribute.name: value}
def find_by_name(name, item_list):
'''
Helper function to find item by name in a list.
'''
items = []
for item in item_list:
assert hasattr(item, "name"), \
"{} should have a 'name' atrribute defined".format(item) # pragma: no cover
if item.name == name:
items.append(item)
if len(items) > 0:
return items[0]
else:
return None