LIMA
Libre Multilingual Analyzer — C++ API
Loading...
Searching...
No Matches
deeplima::segmentation::impl Namespace Reference

Classes

class  CharNgramEncoder
 
class  CharReader
 
struct  ngram_descr_t
 
class  SegmentationDecoder
 
class  SegmentationImpl
 The implementation of the segmenter, a SegmentationClassifier, itself a RnnSequenceClassifier. More...
 
class  SegmentationInferenceWrapper
 
class  Utf8Reader
 

Typedefs

using CharNgramEncoderFromUtf8 = CharNgramEncoder< Utf8Reader<> >
 
using SegmentationClassifier = RnnSequenceClassifier< eigen_impl::Model, eigen_impl::EmbdVectorizer, uint8_t >
 
using InputEncoder = CharNgramEncoderFromUtf8
 
using OutputDecoder = SegmentationDecoder
 

Enumerations

enum  uc_channel_t : uint8_t {
  channel_char = 0 , channel_type = 1 , channel_len = 2 , channel_script_change = 3 ,
  max_channel_value
}
 

Variables

const uint8_t bits_per_position_table []
 

Typedef Documentation

◆ CharNgramEncoderFromUtf8

◆ InputEncoder

◆ OutputDecoder

◆ SegmentationClassifier

Enumeration Type Documentation

◆ uc_channel_t

Enumerator
channel_char 
channel_type 
channel_len 
channel_script_change 
max_channel_value 

Definition at line 27 of file utf8_reader.h.

Variable Documentation

◆ bits_per_position_table

const uint8_t deeplima::segmentation::impl::bits_per_position_table[]
Initial value:
= {
21,
8,
3,
1
}

Definition at line 35 of file utf8_reader.h.