Guide-yakabatana mibvunzo · Pakati Nhanho

Muenzaniso Pruning Quiz

Checks structured and unstructured pruning, sparse-kernel limits, iterative recovery, lottery-ticket claims, and deployment measurements.

Nzira dzekutungamira dzinoenderanaKuchekwa kweModel
Mubvunzo 1 ye 8

What does model pruning remove from a trained network?

Tarisa mubvunzo wese8 mibvunzo

Mhinduro uye tsananguro zvinoonekwa paunenge uchitora bvunzo.

  1. What does model pruning remove from a trained network?
    • Kubvisa uremu kana zvimiro zvinopa zvishoma kune zvakabuda
    • Kuwedzera mamwe ma parameter kuti uvandudze huchokwadi
    • Kudzidzisa modhi kune mamwe maepoch
    • Kuwedzera kugadzirisa kwekuisa
  2. Chii chinosiyanisa kuchekerera kwakarongwa kubva kune zvisina kurongwa?
    • Structured pruning inoshanda chete pamifananidzo
    • Structured pruning inobvisa mayuniti akazara senge neurons kana misoro; isina kuumbwa inobvisa huremu hwega
    • Kuchekerera kusina kurongwa kunoda kudzidziswa kubva pakatanga
    • Hapana musiyano unoshanda pakati pavo
  3. Nei kuchekerera kusina kurongeka kuchiwanzotadza kumhanyisa modhi pane zvakajairika hardware?
    • Inobvisa ma layer akawandisa
    • Nguva dzose inoderedza kurongeka kusvika ku zero
    • Sparse uremu matrices anoda yakakosha hardware kana maraibhurari kushandisa mazero
    • Inowedzera parameter kuverenga
  4. Chii chinowanzoitwa iterative pruning resipi?
    • Chenesa zvese kamwechete, wozoisa
    • Chengetedza kutanga, uye usazodzidzirazve
    • Dhonza bvisa zvidimbu kusvika chokwadi chadonha
    • Rovedza, chekerera zvidiki-zvakakosha ma paramita, nyatso-tune kupora, uye dzokorora
  5. Chii chinorehwa neLottery Tikiti Hypothesis?
    • Dense network ine sparse subnetworks inogona kuenderana neiyo yakazara modhi kana yakadzidziswa kubva pakutanga kwayo
    • Kuchekerera nguva dzose kunovandudza kururama
    • Kuchekerera zvisina tsarukano kunogara kwakanakisa
    • Mamodheru makuru haagone kudzvanywa
  6. Ndeipi mucherechedzo wakapfava unowanzoshandiswa kusarudza kuti ndeupi huremu hwekuchekerera?
    • Kurongeka kwearufabheti yemazita akaturikidzana
    • Hukuru hwakakwana hwezviyereso
    • Ruvara rwemamepu ekuita
    • Nhamba yemuenzaniso wekudzidzisa
  7. When is pruning most likely to reduce inference latency on ordinary dense hardware?
    • When individual zeros are added but the dense kernel still computes them
    • When the model is renamed without changing its graph
    • When whole channels or other structures are removed to produce a smaller dense model
    • When a sparse checkpoint is stored in a larger file
  8. What should a team measure after pruning for a deployment target?
    • Only the number of zeros in a checkpoint
    • Only training accuracy before pruning
    • Whether the file extension changed
    • Task quality and latency or memory on the intended hardware