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
|
# @package sparse_to_dense
# Module caffe2.python.layers.sparse_to_dense
from collections import defaultdict
import numpy as np
from caffe2.python import schema
from caffe2.python.layers.layers import AccessedFeatures, ModelLayer
class FeatureSparseToDense(ModelLayer):
def __init__(
self,
model,
input_record,
input_specs,
name="feature_sparse_to_dense",
default_dense_value=None,
**kwargs
):
"""
`input_specs` follows the format of FeatureSpec from schema. To be more
precise it's a namedtuple that should have:
'feature_type', 'feature_names', 'feature_ids'
Default_dense_value can only be 0.0 or float("NaN"). Any input that isn't
None will be NaN.
"""
super(FeatureSparseToDense, self).__init__(model, name, input_record, **kwargs)
if default_dense_value is None:
default_dense_value = 0.0
default_dense_value = float(default_dense_value)
assert (
np.isnan(default_dense_value) or default_dense_value == 0.0
), "default_dense_value can only be 0.0 or NaN"
self.input_specs = input_specs
self.default_float_value = (
model.global_constants["NAN"]
if np.isnan(default_dense_value)
else model.global_constants["ZERO"]
)
self.zero_range = model.global_constants["ZERO_RANGE"]
outputs = []
for field, feature_specs in self.input_specs:
assert len(feature_specs.feature_names) == len(feature_specs.feature_ids)
if feature_specs.feature_type == "FLOAT":
outputs.append(
(
field,
schema.Scalar(
(np.float32, (len(feature_specs.feature_ids),)),
self.get_next_blob_reference(field + "_output"),
),
)
)
elif feature_specs.feature_type == "ID_LIST":
outputs.append(
(
field,
schema.Struct(
(
"ranges",
schema.Scalar(
(np.int32, (len(feature_specs.feature_ids), 2)),
self.get_next_blob_reference(field + "_ranges"),
),
),
(
"values",
schema.Scalar(
np.int64,
self.get_next_blob_reference(field + "_values"),
),
),
),
)
)
elif feature_specs.feature_type == "ID_SCORE_LIST":
outputs.append(
(
field,
schema.Struct(
(
"ranges",
schema.Scalar(
(np.int32, (len(feature_specs.feature_ids), 2)),
self.get_next_blob_reference(field + "_ranges"),
),
),
(
"ids",
schema.Scalar(
np.int64,
self.get_next_blob_reference(field + "_ids"),
),
),
(
"scores",
schema.Scalar(
np.float32,
self.get_next_blob_reference(field + "_scores"),
),
),
),
)
)
elif feature_specs.feature_type == "EMBEDDING":
# We don't know dimensions of embeddings in input data.
# Even though they should match dimensions from feature config,
# we keep ranges blob to check input data later.
outputs.append(
(
field,
schema.Struct(
(
"ranges",
schema.Scalar(
(np.int32, (len(feature_specs.feature_ids), 2)),
self.get_next_blob_reference(field + "_ranges"),
),
),
(
"values",
schema.Scalar(
np.float32,
self.get_next_blob_reference(field + "_values"),
),
),
),
)
)
elif feature_specs.feature_type == "GENERIC_FEATURE":
# We don't know dimensions of embeddings in input data.
# Even though they should match dimensions from feature config,
# we keep ranges blob to check input data later.
# Currently this schema with ranges and values is only for
# generic type enum 1. If new types are implemented, we need to
# modify the ParseGeneric operator, and this part accordingly
outputs.append(
(
field,
schema.Struct(
(
"ranges",
schema.Scalar(
(np.int32, (len(feature_specs.feature_ids), 2)),
self.get_next_blob_reference(field + "_ranges"),
),
),
(
"values",
schema.Scalar(
np.float32,
self.get_next_blob_reference(field + "_values"),
),
),
),
)
)
else:
raise TypeError(
"Unsupported input type: {0}".format(feature_specs.feature_type)
)
# TODO(amalevich): This schema is producing ranges. And thus if there is
# something using it it should support ranges as well. It might be
# confusing, if we don't add better support for ranges/have it as a
# first layer
self.output_schema = schema.Struct(*outputs)
# TODO(amalevich): Consider moving this data to schema, instead
# Structs doesn't support attaching metadata to them and clonning
# will break things badly, but this is the most elegant way to pass
# this info around. Should we change it or it'll be too much work and
# not worse it?
for field, feature_specs in input_specs:
schema.attach_metadata_to_scalars(
self.output_schema[field], schema.Metadata(feature_specs=feature_specs)
)
# Add operators to all types that need to be densified
def add_ops(self, net):
record = self.input_record
for field, feature_specs in self.input_specs:
if feature_specs.feature_type == "FLOAT":
net.SparseToDenseMask(
[
record[field].keys(),
record[field].values(),
self.default_float_value,
record[field].lengths(),
],
[self.output_schema[field]()],
mask=feature_specs.feature_ids,
)
elif feature_specs.feature_type == "ID_LIST":
id_list_ranges = net.LengthsToRanges(
record[field].values.lengths(), net.NextScopedBlob("id_list_ranges")
)
net.SparseToDenseMask(
[
record[field].keys(),
id_list_ranges,
self.zero_range,
record[field].lengths(),
],
self.output_schema[field].ranges(),
mask=feature_specs.feature_ids,
)
# Alias helps to enforce the fact that all SparseToDense calls
# produce new blobs.
# Reusing blob names might result in some weird consequences
# during the delivery time, when content of the blobs is
# generated based on the inputSpecs.
net.Alias(
record[field].values.items(), self.output_schema[field].values()
)
elif feature_specs.feature_type == "ID_SCORE_LIST":
# TODO: merge this to the case above?
id_list_ranges = net.LengthsToRanges(
record[field].values.lengths(),
net.NextScopedBlob("id_score_list_ranges"),
)
net.SparseToDenseMask(
[
record[field].keys(),
id_list_ranges,
self.zero_range,
record[field].lengths(),
],
self.output_schema[field].ranges(),
mask=feature_specs.feature_ids,
)
# Alias helps to enforce the fact that all SparseToDense calls
# produce new blobs.
# Reusing blob names might result in some weird consequences
# during the delivery time, when content of the blobs is
# generated based on the inputSpecs.
net.Alias(record[field].values.keys(), self.output_schema[field].ids())
net.Alias(
record[field].values.values(), self.output_schema[field].scores()
)
elif feature_specs.feature_type == "EMBEDDING":
ranges = net.LengthsToRanges(
record[field].values.lengths(),
net.NextScopedBlob("embeddings_ranges"),
)
net.SparseToDenseMask(
[
record[field].keys(),
ranges,
self.zero_range,
record[field].lengths(),
],
self.output_schema[field].ranges(),
mask=feature_specs.feature_ids,
)
# Alias helps to enforce the fact that all SparseToDense calls
# produce new blobs.
# Reusing blob names might result in some weird consequences
# during the delivery time, when content of the blobs is
# generated based on the inputSpecs.
net.Alias(
record[field].values.items(), self.output_schema[field].values()
)
elif feature_specs.feature_type == "GENERIC_FEATURE":
(
feature_lengths_blob,
feature_ids_blob,
value_lengths_blob,
value_values_blob,
) = net.ParseGeneric(
[record[field]()],
["feature_lengths", "feature_ids", "value_lengths", "value_values"],
feature_type_enum=1,
)
# Currently our implementation only supports
# generic type enum 1. If new types are implemented, we need to
# modify the ParseGeneric operator, the schema above,
# and this part accordingly to parse the generic feature strings
# into input_record
ranges = net.LengthsToRanges(
value_lengths_blob, net.NextScopedBlob("generics_ranges")
)
net.SparseToDenseMask(
[feature_ids_blob, ranges, self.zero_range, feature_lengths_blob],
self.output_schema[field].ranges(),
mask=feature_specs.feature_ids,
)
# Alias helps to enforce the fact that all SparseToDense calls
# produce new blobs.
# Reusing blob names might result in some weird consequences
# during the delivery time, when content of the blobs is
# generated based on the inputSpecs.
net.Alias(value_values_blob, self.output_schema[field].values())
def get_metadata(self):
metadata = []
for field, feature_specs in self.input_specs:
metadata.append(
(
{
"type": feature_specs.feature_type,
"names": feature_specs.feature_names,
"ids": feature_specs.feature_ids,
},
self.output_schema[field].field_blobs(),
self.output_schema[field].field_types(),
)
)
if feature_specs.feature_type == "FLOAT":
metadata[-1][0]["cardinality"] = 1
return metadata
def get_accessed_features(self):
accessed_features = defaultdict(list)
# The features that are accessed are just those features that appear in
# the input specs
for field, feature_specs in self.input_specs:
accessed_features[field].append(
AccessedFeatures(
feature_specs.feature_type, set(feature_specs.feature_ids)
)
)
return accessed_features
|
