public class GPT3Tokenizer
extends java.lang.Object
Modifications:
| Constructor and Description |
|---|
GPT3Tokenizer(Encoding encoding) |
| Modifier and Type | Method and Description |
|---|---|
protected int |
bytePairMerge(ByteSequence piece,
java.util.Collection<java.lang.Integer> result) |
protected java.util.regex.Pattern |
createSpecialRegex(java.util.Map<java.lang.String,?> specialTokensEncoder) |
java.lang.String |
decode(java.util.List<java.lang.Integer> tokens) |
protected java.lang.String |
decodeImpl(java.util.List<java.lang.Integer> tokens) |
java.util.List<java.lang.Integer> |
encode(java.lang.String text) |
java.util.List<java.lang.Integer> |
encode(java.lang.String text,
boolean allowedSpecial) |
java.util.List<java.lang.Integer> |
encode(java.lang.String text,
java.util.Set<java.lang.String> allowedSpecial) |
protected java.util.List<java.lang.Integer> |
encodeImpl(java.lang.String text,
java.util.Set<java.lang.String> allowedSpecial) |
protected int |
getRank(ByteSequence piece,
java.util.List<xyz.felh.openai.gpt3.tokenizer.GPT3Tokenizer.IntPair> partsList,
int startIdx) |
protected java.util.regex.Pattern |
getTlRegex()
Returns the regular expression for tokenizing text
|
protected java.util.regex.Pattern |
getTlSpecialRegex()
Returns the regular expression for detecting special tokens
|
public GPT3Tokenizer(Encoding encoding)
protected java.util.regex.Pattern createSpecialRegex(java.util.Map<java.lang.String,?> specialTokensEncoder)
public java.lang.String decode(java.util.List<java.lang.Integer> tokens)
protected java.lang.String decodeImpl(java.util.List<java.lang.Integer> tokens)
protected java.util.regex.Pattern getTlSpecialRegex()
protected java.util.regex.Pattern getTlRegex()
public java.util.List<java.lang.Integer> encode(java.lang.String text)
public java.util.List<java.lang.Integer> encode(java.lang.String text,
boolean allowedSpecial)
public java.util.List<java.lang.Integer> encode(java.lang.String text,
java.util.Set<java.lang.String> allowedSpecial)
protected java.util.List<java.lang.Integer> encodeImpl(java.lang.String text,
java.util.Set<java.lang.String> allowedSpecial)
protected int getRank(ByteSequence piece, java.util.List<xyz.felh.openai.gpt3.tokenizer.GPT3Tokenizer.IntPair> partsList, int startIdx)
protected int bytePairMerge(ByteSequence piece, java.util.Collection<java.lang.Integer> result)
Copyright © 2023. All Rights Reserved.