Maswali ya Kuzingatia Maswali ya Vikundi

Majibu na maelezo huonekana unapojibu maswali.

1. Katika Umakini wa Hoja-ya Kikundi, ni nini kinachoshirikiwa kati ya kikundi cha vichwa vya hoja?
   • Vichwa vya ufunguo na thamani
   • Makadirio ya pato pekee
   • Upachikaji wa ingizo
   • Kiwango cha kujifunza
2. GQA inaelezewa vyema kama msingi wa kati kati ya misimamo miwili iliyokithiri?
   • Convolution na tahadhari
   • Uangalifu kamili wa vichwa vingi na umakini wa maswali mengi
   • FP32 na FP8 usahihi
   • Miundo ya encoder pekee na avkodare pekee
3. Je, ni sehemu gani ya makisio ambayo GQA inaifanya iwe nafuu?
   • Tokenization ya haraka
   • Uhesabuji wa upotezaji wa mafunzo
   • Kumbukumbu ya kashe ya KV na kipimo data chake wakati wa uzalishaji
   • Laini ya mwisho juu ya msamiati
4. Ikiwa mtindo una vichwa 32 vya maswali na vikundi 8 vya KV, kashe ya KV inapunguzwa kwa takriban sababu gani?
   • Karibu 32x
   • Karibu 2x
   • Hakuna kupunguzwa
   • Kuhusu 4x
5. What quality/speed tradeoff did the GQA paper report in its experiments?
   • Quality close to MHA with inference speed closer to MQA
   • Lower quality and slower inference than both
   • Exactly the same latency and quality for every model
   • A result only on image classification
6. Ni vipi miundo iliyopo ya vichwa vingi mara nyingi hubadilishwa kuwa GQA?
   • Kwa kufuta tabaka zote za tahadhari
   • Kwa hatua fupi ya 'kuinua' kutoka kwa kituo cha ukaguzi kilichopo
   • Kwa kubadili tokenizer tofauti
   • Hawawezi kugeuzwa hata kidogo
7. Which task families did the GQA paper evaluate?
   • Image segmentation and object detection
   • Speech synthesis and speaker verification
   • Language modeling and machine translation
   • Molecular generation and toxicity prediction
8. How much of the original pretraining compute did the GQA paper report using for its uptraining recipe?
   • Takriban 50%
   • About 100%
   • No additional training compute
   • Takriban 5%