sampling-lm-test.cc
5.31 KB
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
// rnnlm/sampling-lm-test.cc
// Copyright 2017 Ke Li
// 2017 Johns Hopkins University (author: Daniel Povey)
// See ../../COPYING for clarification regarding multiple authors
//
// 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
//
// THIS CODE IS PROVIDED *AS IS* BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
// KIND, EITHER EXPRESS OR IMPLIED, INCLUDING WITHOUT LIMITATION ANY IMPLIED
// WARRANTIES OR CONDITIONS OF TITLE, FITNESS FOR A PARTICULAR PURPOSE,
// MERCHANTABLITY OR NON-INFRINGEMENT.
// See the Apache 2 License for the specific language governing permissions and
// limitations under the License.
#include "rnnlm/sampling-lm.h"
namespace kaldi {
namespace rnnlm {
class SamplingLmTest {
public:
typedef SamplingLm::HistType HistType;
typedef SamplingLm::WeightedHistType WeightedHistType;
explicit SamplingLmTest(SamplingLm *arpa) {
arpa_ = arpa;
}
// This function reads in a list of histories and their weights from a file
// only text form is supported
void ReadHistories(std::istream &is, bool binary,
WeightedHistType *histories);
void TestUnigramDistribution();
void TestGetDistribution(const WeightedHistType &histories);
private:
// This SamplingLm object is used to get accesses to private and protected
// members in SamplingLm class
SamplingLm *arpa_;
};
void SamplingLmTest::ReadHistories(std::istream &is, bool binary,
WeightedHistType *histories) {
if (binary) {
KALDI_ERR << "binary-mode reading is not implemented for ArpaFileParser";
}
const fst::SymbolTable* sym = arpa_->Symbols();
std::string line;
KALDI_LOG << "Start reading histories from file...";
// int32 ngram_order = arpa_->ngram_order_;
int32 ngram_order = arpa_->Order();
while (getline(is, line)) {
std::istringstream is(line);
std::istream_iterator<std::string> begin(is), end;
std::vector<std::string> tokens(begin, end);
HistType history;
int32 word;
BaseFloat hist_weight = 0;
for (int32 i = 0; i < tokens.size() - 1; ++i) {
word = sym->Find(tokens[i]);
if (word == -1) { // fst::kNoSymbol
KALDI_ERR << "Found history contains word that is not in Arpa LM";
}
history.push_back(word);
}
HistType h1;
if (history.size() >= ngram_order) {
HistType h(history.end() - ngram_order + 1, history.end());
h1 = h;
}
if (!ConvertStringToReal(tokens.back(), &hist_weight)) {
KALDI_ERR << arpa_->LineReference() << ": invalid history weight '"
<< tokens.back() << "'";
}
KALDI_ASSERT(hist_weight >= 0);
std::pair<HistType, BaseFloat> hist_pair;
if (history.size() >= ngram_order) {
hist_pair = std::make_pair(h1, hist_weight);
} else {
hist_pair = std::make_pair(history, hist_weight);
}
(*histories).push_back(hist_pair);
}
KALDI_LOG << "Successfully reading histories from file.";
}
void SamplingLmTest::TestUnigramDistribution() {
std::vector<BaseFloat> unigram_probs;
unigram_probs = arpa_->GetUnigramDistribution();
// Check 0 (epsilon) has probability 0.0
KALDI_ASSERT(unigram_probs[0] == 0.0);
// Assert the sum of unigram probs of all words is 1.0
BaseFloat probsum = 0.0;
for (int32 i = 0; i < unigram_probs.size(); ++i) {
probsum += unigram_probs[i];
}
KALDI_ASSERT(ApproxEqual(probsum, 1.0));
}
void SamplingLmTest::TestGetDistribution(const WeightedHistType &histories) {
// get total input weights of histories
BaseFloat total_weights = 0.0;
WeightedHistType::const_iterator it = histories.begin();
for (; it != histories.end(); ++it) {
total_weights += it->second;
}
BaseFloat unigram_weight = 0.0;
BaseFloat non_unigram_probsum = 0.0;
std::vector<std::pair<int32, BaseFloat> > pdf;
unigram_weight = arpa_->GetDistribution(histories, &pdf);
for (int32 i = 0; i < pdf.size(); ++i) {
non_unigram_probsum += pdf[i].second;
}
// assert unigram weight plus total non_unigram probs equals
// the total input histories' weights
KALDI_ASSERT(ApproxEqual(unigram_weight + non_unigram_probsum, total_weights));
}
} // namespace rnnlm
} // namespace kaldi
int main(int argc, char **argv) {
using namespace kaldi;
using namespace kaldi::rnnlm;
const char *usage = "";
ParseOptions po(usage);
po.Read(argc, argv);
std::string arpa_file = "test/0.1k_3gram_unpruned.arpa",
history_file = "test/hists";
ArpaParseOptions options;
fst::SymbolTable symbols;
enum {
kEps = 0,
kBos, kEos, kUnk
};
// Use spaces on special symbols, so we rather fail than read them by mistake.
symbols.AddSymbol(" <eps>", kEps);
options.bos_symbol = symbols.AddSymbol("<s>", kBos);
options.eos_symbol = symbols.AddSymbol("</s>", kEos);
options.unk_symbol = symbols.AddSymbol("<unk>", kUnk);
options.oov_handling = ArpaParseOptions::kAddToSymbols;
SamplingLm arpa(options, &symbols);
bool binary;
Input k1(arpa_file);
arpa.Read(k1.Stream());
SamplingLmTest mdl(&arpa);
mdl.TestUnigramDistribution();
Input k2(history_file, &binary);
SamplingLmTest::WeightedHistType histories;
mdl.ReadHistories(k2.Stream(), binary, &histories);
mdl.TestGetDistribution(histories);
KALDI_LOG << "Tests for SamplingLm class succeed.";
return 0;
}