Maswali ya Miundo ya Mixtral na Sparse

Majibu na maelezo huonekana unapojibu maswali.

1. Katika Mixtral 8x7B, ni wataalam wangapi wanaochakata kila tokeni ya mtu binafsi?
   • Wataalamu wawili
   • Wataalam wote wanane
   • Mtaalamu mmoja kabisa
   • Nambari ya nasibu kila wakati
2. Ni sehemu gani inayoamua ni wataalam gani ishara inatumwa?
   • Kiashiria
   • Kipanga njia (mtandao wa lango)
   • Safu ya tahadhari
   • Jedwali la kupachika
3. Ingawa Mixtral 8x7B ina jumla ya vigezo 47B, takribani ni ngapi zinazotumika kwa kila tokeni?
   • Takriban bilioni 47
   • Takriban bilioni 7
   • Takriban bilioni 13
   • Takriban bilioni 70
4. What must a Mixtral serving setup account for even when only two experts run per token?
   • Only the active parameters ever need to be stored anywhere
   • All experts must always fit in GPU memory with no offloading option
   • Unused experts can be deleted without changing the checkpoint
   • The full expert weights must be available to the system, with residency or offloading depending on deployment
5. Je, Mistral AI ilitoa Mixtral chini ya leseni gani, na hivyo kuchochea kupitishwa kwa wazi?
   • Apache 2.0
   • Umiliki, chanzo funge pekee
   • GPL v3
   • Creative Commons Isiyo ya Kibiashara
6. What training steps does Mistral report for the Mixtral 8x7B Instruct variant?
   • Huondoa hitaji la data yoyote ya mafunzo
   • Supervised fine-tuning and direct preference optimization (DPO)
   • Inaondoa hitaji la GPU kabisa
   • Inahakikisha mifano haifanyi makosa kamwe
7. How many expert networks are available in each Mixtral 8x7B sparse mixture-of-experts layer?
   • Wataalamu wawili
   • Seven experts
   • Eight experts
   • Twelve experts
8. Which languages does Mistral list for Mixtral 8x7B?
   • Yes; all unused expert weights are deleted
   • Yes; it means the model is dense with 12.9B total parameters
   • No; it means every parameter is active for every token
   • English, French, Italian, German, and Spanish