Ikibazo gihuza ibibazo · Hagati Urwego

Ikibazo cya WordPiece Ikibazo

Ongera usubiremo amagambo ya WordPiece hamwe numururumba wo gutandukanya umururumba, harimo uruhare rwayo mu ndimi nyinshi BERT no gutandukana na BPE.

Inzira zijyanye no kuyoboraIjambo ryamagambo
Ikibazo 1 Bya 8

Niki prefix '##' yerekana mubisohoka muri WordPiece?

Reba buri kibazo8 ibibazo

Ibisubizo nibisobanuro bigaragara mugihe ufata ikibazo.

  1. Niki prefix '##' yerekana mubisohoka muri WordPiece?
    • Ikimenyetso nigice gikomeza imbere yijambo, ntabwo ijambo-intangiriro
    • Ikimenyetso nikintu kidasanzwe gitandukanya nka [SEP]
    • Ikimenyetso gifite inshuro nyinshi mumagambo
    • Ikimenyetso nikimenyetso kitazwi-cyamagambo
  2. Hafi yamagambo angana gute WordPiece ikoreshwa na BERT yumwimerere?
    • Ibimenyetso bigera kuri 256
    • Ibimenyetso bigera ku 30.000
    • Ibimenyetso bigera kuri miliyoni
    • Ibimenyetso bigera kuri 500
  3. Nigute amahame ya WordPiece yo guhuza atandukanye na Encoding ya Byte-Pair?
    • Ihuza couple idakunze kugabanya amagambo
    • Ntabwo ihuza kandi ikoresha inyuguti gusa
    • Ihuza mukugwiza amahugurwa-amakuru bishoboka aho kuba inshuro nke
    • Ihuza ibice byombi mugihe cyamahugurwa
  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. Mugihe cyo gufata umwanzuro, ni gute WordPiece ihitamo uburyo bwo gutandukanya ijambo?
    • Ihuye nijambo rirerire mumagambo umururumba, ibumoso ugana iburyo
    • Ihitamo ibice bigufi bishoboka mbere
    • Iragerageza ibishoboka byose gutandukana no kubigereranya
    • Ikoresha umuyoboro udasanzwe kugirango uhanure gutandukana
  6. Nuwuhe muryango w'icyitegererezo uzwi cyane WordPiece ifitanye isano cyane?
    • Moderi ya GPT
    • BERT n'ibiyikomokaho
    • Moderi 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