Àdánwò Àmi WordPiece

Awọn idahun ati awọn alaye han bi o ṣe n mu ibeere naa.

1. Kini ami-iṣaaju '##' tọkasi ninu iṣẹjade WordPiece?
   • Aami naa jẹ nkan itesiwaju ninu ọrọ kan, kii ṣe ọrọ-akọkọ
   • Aami naa jẹ oluyapa pataki bi [SEP]
   • Aami naa ni igbohunsafẹfẹ ti o ga julọ ninu ọrọ-ọrọ
   • Àmi jẹ ami aimọ jade-ti-fokabulari
2. Ni aijọju bawo ni ọrọ ti WordPiece ṣe tobi to nipasẹ BERT atilẹba?
   • About 256 àmi
   • About 30.000 àmi
   • About 1 million àmi
   • About 500 àmi
3. Báwo ni àyẹ̀wò àkópọ̀ WordPiece ṣe yàtọ̀ sí Bíte-Pair Ìṣirò palẹ̀?
   • O dapọ mọ awọn toje bata lati din fokabulari
   • Ko dapọ ati lo awọn ohun kikọ nikan
   • O dapọ nipasẹ mimuuṣiṣẹpọ ikẹkọ-data o ṣeeṣe kuku ju igbohunsafẹfẹ aise lọ
   • O dapọ awọn orisii ni ID nigba ikẹkọ
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. Ni akoko itọkasi, bawo ni WordPiece ṣe yan bi o ṣe le pin ọrọ kan?
   • Ó bá ọ̀rọ̀ abẹ́lẹ̀ tó gùn jù lọ nínú ọ̀rọ̀ ìfojúkòkòrò, òsì sí ọ̀tún
   • O mu awọn ege ti o kuru ju ti ṣee ṣe akọkọ
   • O gbiyanju gbogbo pipin ti o ṣeeṣe ati awọn aropin wọn
   • O nlo netiwọki nkankikan lati ṣe asọtẹlẹ awọn ipin
6. Ewo ni idile awoṣe ti a mọ daradara ni WordPiece julọ ni nkan ṣe pẹlu?
   • Awọn awoṣe GPT
   • BERT ati awọn itọsẹ rẹ
   • Awọn awoṣe LLMA
   • 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