Guide-yakabatana mibvunzo · Zvakaoma Nhanho

Mixtral uye Sparse Models Quiz

Review Mixtral routing, expert activation, total versus active parameters, and the practical memory and deployment trade-offs of sparse models.

Nzira dzekutungamira dzinoenderanaMixtral Uye Sparse Models
Mubvunzo 1 ye 8

MuMixtral 8x7B, vangani nyanzvi dzinogadzira chiratidzo chega chega?

Tarisa mubvunzo wese8 mibvunzo

Mhinduro uye tsananguro zvinoonekwa paunenge uchitora bvunzo.

  1. MuMixtral 8x7B, vangani nyanzvi dzinogadzira chiratidzo chega chega?
    • Nyanzvi mbiri
    • Nyanzvi dzese sere
    • Nyanzvi imwe chete
    • Nhamba isina kurongeka nguva imwe neimwe
  2. Ndechipi chikamu chinosarudza kuti ndedzipi nyanzvi chiratidzo chinotumirwa?
    • The tokenizer
    • Iyo router (geting network)
    • The attention layer
    • Tafura yekumisikidza
  3. Kunyangwe Mixtral 8x7B ine inenge 47B yakazara paramita, ingangoita mangani ari kushanda pachiratidzo?
    • Anenge mabhiriyoni 47
    • Anenge mabhiriyoni manomwe
    • Anenge mabhiriyoni gumi nematatu
    • Anenge mabhiriyoni makumi manomwe
  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. Iri pasi ripi rezinesi iro Mistral AI yakaburitsa Mixtral, ichikurudzira kutorwa kwakavhurika?
    • Apache 2.0
    • Proprietary, yakavharwa-source chete
    • GPL v3
    • Creative Commons NonCommerce
  6. What training steps does Mistral report for the Mixtral 8x7B Instruct variant?
    • Inobvisa kudiwa kwechero data yekudzidziswa
    • Supervised fine-tuning and direct preference optimization (DPO)
    • Inobvisa kudiwa kweGPUs zvachose
    • Inovimbisa mamodheru kusambokanganisa
  7. How many expert networks are available in each Mixtral 8x7B sparse mixture-of-experts layer?
    • Nyanzvi mbiri
    • 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