Stochastic Weight Average Quiz

Mhinduro uye tsananguro zvinoonekwa paunenge uchitora bvunzo.

1. Chii chinonzi Stochastic Weight Averaging chaizvo?
   • Huremu hwemuenzaniso wakatorwa pazvibodzwa zvakawanda kunonoka kudzidziswa
   • Mafungiro emamodheru akati wandei akadzidziswa
   • Iwo ma gradients akaverengera kune ese mini-batches
   • Mari yekudzidza inoshandiswa panguva yega yega
2. Sei SWA ichida kuvandudza generalization?
   • Inobata nemusoro kudzidziswa kwakaiswa zvakanyanya
   • Inowana yakafara, yakapfava minima iyo yakanyanya kusimba pane isingaonekwe data
   • Inowedzera huwandu hwema parameters mumuenzaniso
   • Inobvisa kudiwa kweseti yekusimbisa
3. Nderipi rimwe danho rinodiwa neSWA kune network inoshandisa batch normalization?
   • Kudzidzira zvakare network kubva pakutanga
   • Kupeta kaviri saizi yebatch
   • Imwe yekuwedzera yekupfuura kudzoreredza BN inomhanya manhamba
   • Kubvisa ese dropout layers
4. Ndeapi maitiro ekudzidza-chiyero anowanzo shandiswa panguva yeavhareji yeSWA chikamu?
   • Mwero wekudzidza waderera kusvika paziro chaiyo
   • Chiyero chekudzidza chechimwe chekugadzikana
   • Hapana mwero wekudzidza zvachose sezvo kudzidziswa kwakaoma nechando
   • Chiyero chenguva dzose kana chinotenderera, chine mwero wekudzidza wakanyanya
5. Mari yekufungidzira yeSWA inofananidzwa sei neyechinyakare ensemble yeN modhi?
   • Inodhura inenge imwe modhi, sezvo mhedzisiro iri seti yehuremu
   • Inodhura kakapetwa N
   • Zvakafanana nekumhanyisa N modhi
   • Haikwanise kushandiswa kufungidzira
6. Why does the batch-normalization adjustment belong after SWA weight averaging?
   • The averaging step deletes every convolution
   • Its running statistics must be recomputed for the averaged weights
   • It converts batch normalization into layer normalization
   • It restores the original optimizer learning rate
7. An SWA running mean combines n previous samples with a new weight vector w_i. Which update gives equal weight to all n+1 samples?
   • w_SWA + n · w_i
   • (w_SWA + w_i) divided by n
   • (n · w_SWA + w_i) divided by (n + 1)
   • n · (w_SWA − w_i)
8. Why is SWA commonly started during the later part of training?
   • It requires the model to remain at random initialization
   • It can only average models with different architectures
   • It replaces the training data with validation examples
   • It averages nearby useful checkpoints from a trajectory after the model has made substantial progress