Maswali yanayounganishwa na mwongozo · Kati Kiwango

Maswali ya Tokeni ya WordPiece

Kagua msamiati wa WordPiece na sheria za uchoyo za ugawaji, ikijumuisha jukumu lake katika BERT ya lugha nyingi na utofautishaji wake na BPE.

Njia za mwongozo zinazohusianaTokenization ya Neno
Swali 1 ya 8

Je, kiambishi awali cha '##' kinaonyesha nini katika pato la WordPiece?

Hakiki kila swali8 maswali

Majibu na maelezo huonekana unapojibu maswali.

  1. Je, kiambishi awali cha '##' kinaonyesha nini katika pato la WordPiece?
    • Ishara ni kipande cha mwendelezo ndani ya neno, sio neno-awali
    • Ishara ni kitenganishi maalum kama [SEP]
    • Ishara ina masafa ya juu zaidi katika msamiati
    • Ishara ni alama isiyojulikana ya msamiati
  2. Takriban msamiati wa WordPiece unaotumiwa na BERT asili ni ukubwa gani?
    • Takriban tokeni 256
    • Takriban tokeni 30,000
    • Takriban tokeni milioni 1
    • Takriban tokeni 500
  3. Je, kigezo cha kuunganisha cha WordPiece kinatofautiana vipi na Usimbaji wa Kawaida wa Byte-Jozi?
    • Inaunganisha jozi adimu zaidi ili kupunguza msamiati
    • Haiunganishi kamwe na hutumia wahusika pekee
    • Inaunganishwa kwa kuongeza uwezekano wa data ya mafunzo badala ya masafa ghafi
    • Inaunganisha jozi bila mpangilio wakati wa mafunzo
  4. What does a BERT-style WordPiece tokenizer return if it cannot segment a word into vocabulary pieces?
    • It always falls back to individual Unicode characters
    • A random subword selected from the vocabulary
    • The original word encoded as one new token
    • The unknown-token marker, such as [UNK]
  5. Kwa wakati wa uelekezaji, WordPiece huchagua vipi jinsi ya kugawanya neno?
    • Inalingana na neno ndogo refu zaidi katika msamiati kwa pupa, kushoto kwenda kulia
    • Inachukua vipande vifupi iwezekanavyo kwanza
    • Inajaribu kila mgawanyiko unaowezekana na kuwaweka wastani
    • Inatumia mtandao wa neva kutabiri migawanyiko
  6. Je, WordPiece inahusishwa na familia gani ya kielelezo inayojulikana zaidi?
    • Mifano ya GPT
    • BERT na derivatives yake
    • Mifano ya LLaMA
    • Stable Diffusion
  7. What kind of vocabulary did multilingual BERT use across its training languages?
    • A completely separate vocabulary for every language
    • A vocabulary made only of complete words
    • One shared WordPiece vocabulary
    • A list of byte sequences with no subwords
  8. Which tokenizer class in Hugging Face Transformers constructs a BERT tokenizer that uses WordPiece?
    • GPT2Tokenizer
    • RobertaTokenizer
    • CLIPTokenizer
    • BertTokenizer