Class BPETrainer
java.lang.Object
io.github.kirstenali.deepj.tokenizers.bpe.BPETrainer
-
Field Summary
Fields -
Constructor Summary
Constructors -
Method Summary
Modifier and TypeMethodDescriptiontrainFromFile(Path path, int targetVocabSize, List<String> specialTokens) trainFromFile(Path path, int targetVocabSize, List<String> specialTokens, int sampleChars) trainTokenizer(String text, int targetVocabSize) trainTokenizer(String text, int targetVocabSize, List<String> specialTokens) trainTokenizerFromFile(Path path, int targetVocabSize) trainTokenizerFromFile(Path path, int targetVocabSize, List<String> specialTokens) trainTokenizerWithDefaults(String text, int targetVocabSize) trainTokenizerWithDefaultsFromFile(Path path, int targetVocabSize)
-
Field Details
-
DEFAULT_SPECIAL_TOKENS
-
-
Constructor Details
-
BPETrainer
public BPETrainer()
-
-
Method Details
-
train
-
train
-
trainFromFile
public BPEModel trainFromFile(Path path, int targetVocabSize, List<String> specialTokens) throws IOException - Throws:
IOException
-
trainFromFile
public BPEModel trainFromFile(Path path, int targetVocabSize, List<String> specialTokens, int sampleChars) throws IOException - Throws:
IOException
-
trainTokenizer
-
trainTokenizer
-
trainTokenizerWithDefaults
-
trainTokenizerWithDefaultsFromFile
public BPETokenizer trainTokenizerWithDefaultsFromFile(Path path, int targetVocabSize) throws IOException - Throws:
IOException
-
trainTokenizerFromFile
- Throws:
IOException
-
trainTokenizerFromFile
public BPETokenizer trainTokenizerFromFile(Path path, int targetVocabSize, List<String> specialTokens) throws IOException - Throws:
IOException
-