quiz bu lëkkale ak tegtal · Diggante Niveau

Modèlu Quiz

Jëfandikool dencukaay buñ kantite, fakteer yuy eskale, ak kompromis yu jaar yoon, boole ci lu tax wàññi poid bu ñor bi mënul gaaraati fit wala gaawaay bu mat ci model bi.

Yooni njiit yu jëm ciModel Kantifikaasioŋ
Laaj 1 de 8

Quantizing benn model dafay wàññi lan?

Xoolaatal laaj bu nekk8 laaj

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

  1. Quantizing benn model dafay wàññi lan?
    • Nimero bu jaar yoon bi ñuy jëfandikoo ngir denc poid yi ak aktivasioŋ yi
    • Limu couche yi ci reso bi
    • Done yiñ tabax ci modelu tàggat yaram
    • Làkk wi model bi mëna dégg
  2. Ignoring scale metadata and other runtime tensors, what is the weight-payload ratio from sixteen bits to four bits per weight?
    • The payload is unchanged
    • It is one hundredth of the original payload
    • The four-bit payload is one quarter of the sixteen-bit payload
    • The payload becomes four times larger
  3. Lan mooy njariñ li gëna mag ci kwantisation bu am bit bu woyof?
    • Royuwaay bi fàtte na làkku tàggat yaram
    • Mën na ñàkka dëggu ndax dañuy tëye valeur yi ci ay seau yu néew
    • Mën na dox ludul ci super ordinatër
    • Dafay efaase ay diisaay ba fàww
  4. Lan moo wuutale tàggat xam-xam kantite ak kantite ginaaw tàggat?
    • Soxlawul benn done
    • Ci nataal kese lay dox
    • Dafay yokk limu bit yi ci diir bi
    • Dafay nuru wërsëg ci diiru tàggat yaram suko defee model bi jàng muñ njubte gu néew
  5. Ban pexe mooy jëfandikoo 4-bit quantization ngir defar ay model yu yaatu yu yomb ci benn GPU?
    • Dàddu
    • Normalisasioŋ ci lots
    • QLoRA
    • Tokenisasioŋ
  6. Lan moo waral scaling ci chaine bu nekk wala ci groupe bu nekk di jàppale model yu bari yi?
    • Dafay jox factëri eskaal yu wuute ay daggitu matriks diisaay ngir mëna am njub
    • Dina dindi bépp poids
    • Dafay gëna rëy model bi
    • Dafay soppi mbind mu nekk audio
  7. A four-bit model’s raw weights fit in available VRAM, but serving fails out of memory. What could explain this?
    • Every four-bit weight expands permanently to a gigabyte
    • Quantization guarantees no runtime allocations
    • Only the model’s filename consumes the remaining memory
    • Caches, scales, unquantized tensors, or workspace need additional memory
  8. A team quantizes to INT8 but sees no inference speedup. What should it check?
    • Whether the training data are alphabetized
    • Whether the workload uses efficient supported kernels on its hardware
    • Whether lower bit width always guarantees faster execution
    • Whether the model must have changed languages