Quiz model yu toppalante ci toppalante

Sooy def quiz bi, tontu yi ak leeral yi dañuy feeñ.

1. Lan mooy ñaari mbir yi gëna am solo ci xeetu toppalante ci toppalante?
   • Encodeur ak decodeur
   • Tokenisatër ak xaajkat
   • Ñaari encoder yu nuróo
   • Seggal ak rang
2. Ban jafe-jafe bu mag la jumtukaayu bàyyi xel bi saafara ci xeetu seq2seq yi?
   • Tokenisation bu yeex
   • Bottleneck vecteur bu am guddaay bu takku ngir dugal yu gudd
   • Ñàkum kàddu
   • Done yu bari ci wàllu tàggat yaram
3. Ban architecture la xeetu tekki seq2seq 2014 bi njëkk jëfandikoo?
   • Reseau konvolusioŋ
   • Garab yiy jël dogal
   • LSTMs yuñ dajale
   • feed-forward bu amul dara
4. naka la ko bàyyi xel di tabaxee ci jéego bu nekk ci dekodaas bi?
   • Ni ñuy jëlee token yiñ dugal yépp ñu tolloo
   • Soo bañee bàyyi xel ci encoder bi
   • Sooy kopie token bi mujjee dugal
   • Ci lim buñ pondéree ci etaa encodeur yi jëfandikoo poñ yiñ méngale
5. Lan moo waral génnug seq2seq mëna wuute ci guddaay ak duggal?
   • Decodeur bi dafay defar ay token ba keroog muy genne siñaal stop
   • Duggal ak genn dañuy tolloo guddaay saa yu nekk
   • Encoder bi dafay pad lépp
   • Guddaay bi ñu ngi koy tëral bu njëkk
6. Ban xeetu arsitektir seq2seq bu bees moo gëna siiw ci Transformer?
   • ResNet ak VGG
   • T5 ak BART
   • K-means ak PCA
   • Word2Vec ak Gant
7. In encoder-decoder attention, which representations are commonly used as keys and values?
   • Only the decoder’s future hidden states
   • The vocabulary file’s byte offsets
   • The encoder’s output representations
   • The optimizer’s learning-rate history
8. What does a bidirectional encoder provide to an attention-based decoder?
   • A fixed one-token output regardless of the input
   • A guarantee that output length equals input length
   • A causal mask over future target tokens
   • Contextual representations of source tokens that can use information from both directions