Mila
Deep Neural Network Library
Loading...
Searching...
No Matches
Mila::Data::Tokenizer Class Referenceabstractexport
Inheritance diagram for Mila::Data::Tokenizer:
Mila::Data::BpeTokenizer Mila::Data::CharTokenizer

Public Member Functions

virtual std::string decode (std::span< const TokenId > tokens)=0
virtual std::vector< TokenId > encode (const std::string &text)=0
virtual std::optional< TokenId > getBosTokenId () const =0
virtual std::optional< TokenId > getEosTokenId () const =0
virtual std::optional< TokenId > getPadTokenId () const =0
virtual size_t getVocabSize () const =0
virtual bool isValidToken (TokenId tokenId) const =0
virtual std::string tokenToString (TokenId tokenId) const =0

Member Function Documentation

◆ decode()

virtual std::string Mila::Data::Tokenizer::decode ( std::span< const TokenId > tokens)
pure virtual

◆ encode()

virtual std::vector< TokenId > Mila::Data::Tokenizer::encode ( const std::string & text)
pure virtual

◆ getBosTokenId()

virtual std::optional< TokenId > Mila::Data::Tokenizer::getBosTokenId ( ) const
pure virtual

Implemented in Mila::Data::CharTokenizer.

◆ getEosTokenId()

virtual std::optional< TokenId > Mila::Data::Tokenizer::getEosTokenId ( ) const
pure virtual

Implemented in Mila::Data::CharTokenizer.

◆ getPadTokenId()

virtual std::optional< TokenId > Mila::Data::Tokenizer::getPadTokenId ( ) const
pure virtual

Implemented in Mila::Data::CharTokenizer.

◆ getVocabSize()

virtual size_t Mila::Data::Tokenizer::getVocabSize ( ) const
pure virtual

Implemented in Mila::Data::CharTokenizer.

◆ isValidToken()

virtual bool Mila::Data::Tokenizer::isValidToken ( TokenId tokenId) const
pure virtual

Implemented in Mila::Data::CharTokenizer.

◆ tokenToString()

virtual std::string Mila::Data::Tokenizer::tokenToString ( TokenId tokenId) const
pure virtual

Implemented in Mila::Data::CharTokenizer.


The documentation for this class was generated from the following file:
  • Mila/Src/Data/Tokenizers/Tokenizer.ixx