nndeploy C++ API  0.2.0
nndeploy C++ API
Classes | Enumerations | Functions
nndeploy::tokenizer Namespace Reference

Classes

class  TokenizerDecode
 TokenizerDecode. More...
 
class  TokenizerEncode
 TokenizerEncode. More...
 
class  TokenizerIds
 
class  TokenizerPraram
 
class  TokenizerText
 

Enumerations

enum  TokenizerType : int {
  kTokenizerTypeHF = 0x0000 , kTokenizerTypeBPE , kTokenizerTypeSentencePiece , kTokenizerTypeRWKVWorld ,
  kTokenizerTypeNotSupport
}
 

Functions

TokenizerType stringToTokenizerType (const std::string &src)
 
std::string tokenizerTypeToString (TokenizerType type)
 

Enumeration Type Documentation

◆ TokenizerType

Enumerator
kTokenizerTypeHF 

Create HF tokenizer from a single in-memory json blob.

Parameters
json_blobThe json blob.
Returns
The created tokenzier.
kTokenizerTypeBPE 

Create BPE tokenizer.

Parameters
vocab_blobThe blob that contains vocabs.
merges_blobThe blob that contains the merges.
added_tokensThe added tokens.
Returns
The created tokenizer.
kTokenizerTypeSentencePiece 

Create SentencePiece.

Parameters
model_blobThe blob that contains vocabs.
Returns
The created tokenizer.
kTokenizerTypeRWKVWorld 

Create RWKVWorldTokenizer.

Parameters
model_blobThe blob that contains vocabs.
Returns
The created tokenizer.
kTokenizerTypeNotSupport 

Definition at line 31 of file tokenizer.h.

Function Documentation

◆ stringToTokenizerType()

TokenizerType nndeploy::tokenizer::stringToTokenizerType ( const std::string &  src)

◆ tokenizerTypeToString()

std::string nndeploy::tokenizer::tokenizerTypeToString ( TokenizerType  type)