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
|
/* Copyright 2016 The TensorFlow 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.
==============================================================================*/
#ifndef TENSORFLOW_CORE_LIB_MONITORING_SAMPLER_H_
#define TENSORFLOW_CORE_LIB_MONITORING_SAMPLER_H_
// We replace this implementation with a null implementation for mobile
// platforms.
#include "tensorflow/core/platform/platform.h"
#ifdef IS_MOBILE_PLATFORM
#include "tensorflow/core/lib/monitoring/mobile_sampler.h"
#else
#include <float.h>
#include <map>
#include "tensorflow/core/framework/summary.pb.h"
#include "tensorflow/core/lib/histogram/histogram.h"
#include "tensorflow/core/lib/monitoring/collection_registry.h"
#include "tensorflow/core/lib/monitoring/metric_def.h"
#include "tensorflow/core/platform/macros.h"
#include "tensorflow/core/platform/mutex.h"
#include "tensorflow/core/platform/thread_annotations.h"
namespace tensorflow {
namespace monitoring {
// SamplerCell stores each value of an Sampler.
//
// A cell can be passed off to a module which may repeatedly update it without
// needing further map-indexing computations. This improves both encapsulation
// (separate modules can own a cell each, without needing to know about the map
// to which both cells belong) and performance (since map indexing and
// associated locking are both avoided).
//
// This class is thread-safe.
class SamplerCell {
public:
SamplerCell(const std::vector<double>& bucket_limits)
: histogram_(bucket_limits) {}
~SamplerCell() {}
// Atomically adds a sample.
void Add(double sample);
// Returns the current histogram value as a proto.
HistogramProto value() const;
private:
histogram::ThreadSafeHistogram histogram_;
TF_DISALLOW_COPY_AND_ASSIGN(SamplerCell);
};
// Bucketing strategies for the samplers.
//
// We automatically add -DBL_MAX and DBL_MAX to the ranges, so that no sample
// goes out of bounds.
//
// WARNING: If you are changing the interface here, please do change the same in
// mobile_sampler.h.
class Buckets {
public:
virtual ~Buckets() = default;
// Sets up buckets of the form:
// [-DBL_MAX, ..., scale * growth^i,
// scale * growth_factor^(i + 1), ..., DBL_MAX].
//
// So for powers of 2 with a bucket count of 10, you would say (1, 2, 10)
static std::unique_ptr<Buckets> Exponential(double scale,
double growth_factor,
int bucket_count);
// Sets up buckets of the form:
// [-DBL_MAX, ..., bucket_limits[i], bucket_limits[i + 1], ..., DBL_MAX].
static std::unique_ptr<Buckets> Explicit(
std::initializer_list<double> bucket_limits);
virtual const std::vector<double>& explicit_bounds() const = 0;
};
// A stateful class for updating a cumulative histogram metric.
//
// This class encapsulates a set of histograms (or a single histogram for a
// label-less metric) configured with a list of increasing bucket boundaries.
// Each histogram is identified by a tuple of labels. The class allows the
// user to add a sample to each histogram value.
//
// Sampler allocates storage and maintains a cell for each value. You can
// retrieve an individual cell using a label-tuple and update it separately.
// This improves performance since operations related to retrieval, like
// map-indexing and locking, are avoided.
//
// This class is thread-safe.
template <int NumLabels>
class Sampler {
public:
~Sampler() {
// Deleted here, before the metric_def is destroyed.
registration_handle_.reset();
}
// Creates the metric based on the metric-definition arguments and buckets.
//
// Example;
// auto* sampler_with_label = Sampler<1>::New({"/tensorflow/sampler",
// "Tensorflow sampler", "MyLabelName"}, {10.0, 20.0, 30.0});
static Sampler* New(const MetricDef<MetricKind::kCumulative, HistogramProto,
NumLabels>& metric_def,
std::unique_ptr<Buckets> buckets);
// Retrieves the cell for the specified labels, creating it on demand if
// not already present.
template <typename... Labels>
SamplerCell* GetCell(const Labels&... labels) LOCKS_EXCLUDED(mu_);
private:
friend class SamplerCell;
Sampler(const MetricDef<MetricKind::kCumulative, HistogramProto, NumLabels>&
metric_def,
std::unique_ptr<Buckets> buckets)
: metric_def_(metric_def),
buckets_(std::move(buckets)),
registration_handle_(CollectionRegistry::Default()->Register(
&metric_def_, [&](MetricCollectorGetter getter) {
auto metric_collector = getter.Get(&metric_def_);
mutex_lock l(mu_);
for (const auto& cell : cells_) {
metric_collector.CollectValue(cell.first, cell.second.value());
}
})) {}
mutable mutex mu_;
// The metric definition. This will be used to identify the metric when we
// register it for collection.
const MetricDef<MetricKind::kCumulative, HistogramProto, NumLabels>
metric_def_;
// Bucket limits for the histograms in the cells.
std::unique_ptr<Buckets> buckets_;
// Registration handle with the CollectionRegistry.
std::unique_ptr<CollectionRegistry::RegistrationHandle> registration_handle_;
using LabelArray = std::array<string, NumLabels>;
// we need a container here that guarantees pointer stability of the value,
// namely, the pointer of the value should remain valid even after more cells
// are inserted.
std::map<LabelArray, SamplerCell> cells_ GUARDED_BY(mu_);
TF_DISALLOW_COPY_AND_ASSIGN(Sampler);
};
////
// Implementation details follow. API readers may skip.
////
inline void SamplerCell::Add(const double sample) { histogram_.Add(sample); }
inline HistogramProto SamplerCell::value() const {
HistogramProto pb;
histogram_.EncodeToProto(&pb, true /* preserve_zero_buckets */);
return pb;
}
template <int NumLabels>
Sampler<NumLabels>* Sampler<NumLabels>::New(
const MetricDef<MetricKind::kCumulative, HistogramProto, NumLabels>&
metric_def,
std::unique_ptr<Buckets> buckets) {
return new Sampler<NumLabels>(metric_def, std::move(buckets));
}
template <int NumLabels>
template <typename... Labels>
SamplerCell* Sampler<NumLabels>::GetCell(const Labels&... labels)
LOCKS_EXCLUDED(mu_) {
// Provides a more informative error message than the one during array
// construction below.
static_assert(sizeof...(Labels) == NumLabels,
"Mismatch between Sampler<NumLabels> and number of labels "
"provided in GetCell(...).");
const LabelArray& label_array = {{labels...}};
mutex_lock l(mu_);
const auto found_it = cells_.find(label_array);
if (found_it != cells_.end()) {
return &(found_it->second);
}
return &(cells_
.emplace(std::piecewise_construct,
std::forward_as_tuple(label_array),
std::forward_as_tuple(buckets_->explicit_bounds()))
.first->second);
}
} // namespace monitoring
} // namespace tensorflow
#endif // IS_MOBILE_PLATFORM
#endif // TENSORFLOW_CORE_LIB_MONITORING_SAMPLER_H_
|
