Guide-yakabatana mibvunzo · Zvakaoma Nhanho

InfoNCE uye SimCLR Zvinangwa Quiz

Inotarisa iyo InfoNCE yakanaka-inopesana-yakashata, maonero eSimCLR uye musoro wekufungidzira, uye mabasa ekushisa uye batch size.

Nzira dzekutungamira dzinoenderanaInfonce Uye Simclr Zvinangwa
Mubvunzo 1 ye 8

Chii chinonzi InfoNCE chinangwa chinokurudzira modhi kuita?

Tarisa mubvunzo wese8 mibvunzo

Mhinduro uye tsananguro zvinoonekwa paunenge uchitora bvunzo.

  1. Chii chinonzi InfoNCE chinangwa chinokurudzira modhi kuita?
    • Ita kuti mubvunzo ufanane nechawo chechokwadi chezvakanaka pane zvakawanda zvisina kunaka
    • Bata nemusoro chaiwo pixel kukosha kwemufananidzo wega wega
    • Kuderedza nhamba yemuenzaniso parameters
    • Fanofanotaura izwi rinotevera mumutsara
  2. MuSimCLR, vaviri vakanaka vanogadzirwa sei?
    • Nokubatanidza mufananidzo nemavara awo omunhu
    • Nekushandisa maviri akasiyana akasarudzika ekuwedzera kune iwo mufananidzo wakafanana
    • Nokusarudza mifananidzo miviri yakasiyana zvachose
    • Nekutenderedza dataset rese
  3. Ndeipi basa rinoitwa tembiricha τ muNT-Xent / InfoNCE?
    • Inoisa mwero wekudzidza
    • Iyo inodzora saizi yebatch otomatiki
    • Inoyera zvibodzwa zvakafanana; pasi τ inorodza kugovera uye inosimbisa zvakashata zvakashata
    • Inotara nhamba yezvikamu mu encoder
  4. Chii chinoitika kune SimCLR's fungidziro musoro mushure mekudzidzira?
    • Inova iyo yekupedzisira classifier
    • Iyo yakaomeswa nechando uye inoshandiswa se encoder
    • Inodzokororwa kuitira kufungidzira
    • Inoraswa; chinomiririra musoro usati washandiswa pamabasa ezasi
  5. Sei SimCLR yaivimba nehukuru hukuru?
    • Kupa mienzaniso yakawanda yakaipa mukati mechikamu chega chekudzidzira
    • Kuderedza kushandiswa kwendangariro
    • Kuti udzivise kushandisa augmentations
    • Kuti abvise kudiwa kweparameter yekushisa
  6. In SimCLR, which examples usually act as negatives for one anchor view?
    • The matching view of the same image
    • The other augmented views in the batch that come from different images
    • Only examples with the same class label
    • Views produced by a second encoder trained on labels
  7. Why does SimCLR use a nonlinear projection head before computing its contrastive loss?
    • It replaces the image encoder
    • It supplies class labels during pretraining
    • The learned projection helps the contrastive objective while the representation before the head can transfer downstream
    • It removes the need for data augmentation
  8. For an InfoNCE anchor, what should happen to the positive similarity relative to sampled negatives?
    • Every negative should have a higher score than the positive
    • All scores must be identical
    • The positive is excluded from the objective
    • The positive should receive a higher score than the negatives