MCPcopy Create free account
hub / github.com/catboost/catboost / BuildBpeWord

Function BuildBpeWord

library/cpp/text_processing/app_helpers/app_helpers.cpp:108–147  ·  view source on GitHub ↗

Source from the content-addressed store, hash-verified

106}
107
108static TIntrusivePtr<TBpeDictionary> BuildBpeWord(
109 const TDictionaryBuilderOptions& dictionaryBuilderOptions,
110 const TDictionaryOptions& dictionaryOptions,
111 const TBpeDictionaryOptions& bpeOptions,
112 const TString& inputFilePath,
113 const TTokenizerOptions& tokenizerOptions,
114 bool useTokenizer,
115 bool verbose
116) {
117 if (verbose) {
118 Cerr << "Stage [1/2]: Dictionary building\n";
119 }
120
121 auto dictionary = BuildDictionary(
122 inputFilePath,
123 dictionaryBuilderOptions,
124 dictionaryOptions,
125 tokenizerOptions,
126 useTokenizer,
127 verbose
128 );
129
130 TBpeDictionaryBuilder bpeBuilder(bpeOptions.NumUnits, bpeOptions.SkipUnknown, dictionary);
131
132 if (verbose) {
133 Cerr << "Stage [2/2]: Bpe building\n";
134 }
135
136 ApplyFuncTotokenizedText(
137 inputFilePath,
138 tokenizerOptions,
139 useTokenizer,
140 verbose,
141 [&](const TVector<TString>& tokens) {
142 bpeBuilder.Add(tokens, /*weight*/1);
143 }
144 );
145
146 return bpeBuilder.FinishBuilding();
147}
148
149static TIntrusivePtr<TBpeDictionary> BuildBpeLetter(
150 const TDictionaryBuilderOptions& dictionaryBuilderOptions,

Callers 1

BuildBpeMethod · 0.85

Calls 3

ApplyFuncTotokenizedTextFunction · 0.85
AddMethod · 0.45
FinishBuildingMethod · 0.45

Tested by

no test coverage detected