Guide-yakabatana mibvunzo · Pakati Nhanho

Classifier-Yemahara Yekutungamira Mibvunzo

Review classifier-free guidance, conditional dropout, quality-diversity tradeoff, inference costs, and guidance scale. Check scope and assumptions.

Nzira dzekutungamira dzinoenderanaClassifier Yemahara Guidance
Mubvunzo 1 ye 8

Sei mamodheru ekuparadzira achisevenzesa classifier-yemahara nhungamiro?

Tarisa mubvunzo wese8 mibvunzo

Mhinduro uye tsananguro zvinoonekwa paunenge uchitora bvunzo.

  1. Sei mamodheru ekuparadzira achisevenzesa classifier-yemahara nhungamiro?
    • Kusimbisa kuti zvakadii kuburitsa kunotevera kurudziro yekumisikidza
    • Kumanikidza modhi kuita diki saizi
    • Kubvisa encoder yemavara zvachose
    • Kushandura mifananidzo kuita odhiyo
  2. Mutevedzeri unodzidza sei kuita fungidziro dzisina zvisungo zvekutungamira-yemahara?
    • Nokudzidzisa wechipiri akaparadzana classifier
    • Nekungodonhedza mamiriro panguva yekudzidziswa
    • Nekushandisa mifananidzo mitema chete
    • Nekudzima iyo GPU
  3. Kuwedzera chiyero chekutungamira kunowanzoita chii?
    • Nguva dzose inovandudza kusiyana-siyana
    • Inobvisa kufurira kwechimbidzo
    • Inosimbisa kutevedzera nekukurumidza asi inogona kukonzera kuwandisa
    • Inoita kuti chifananidzo chive greyscale
  4. Zvingangoita kuti vangani vanopasa mberi padanho rekuita denoising inodiwa yakajairwa classifier-yemahara nhungamiro?
    • Zero
    • Gumi
    • Zana
    • Vaviri (imwe ine zvirevo, imwe isina magumo)
  5. Ndeupi mukana une classifier-yemahara nhungamiro ine pamusoro pekutungamira kwe classifier?
    • Hazvidi mugadziri akadzidziswa zvakasiyana
    • Inoda data rakawanda rekudzidziswa
    • Inoshanda chete pamienzaniso yemavara
    • Inobvisa nzira yekupararira
  6. In the common classifier-free guidance formula, what does guidance scale s=0 produce?
    • The conditional prediction with extra amplification
    • The unconditional model prediction without conditioning
    • A prediction from a separately trained classifier
    • A sample with all denoising removed
  7. How does classifier-free guidance combine model outputs at inference?
    • It combines two image classifiers
    • It removes the conditional prompt
    • It combines conditional and unconditional score estimates
    • It adds a learned reward model after decoding
  8. What does a guidance scale of 1 generally correspond to in the common formulation?
    • The unconditional prediction only
    • A classifier-based gradient
    • A zero-noise image
    • The conditional prediction without extra guidance amplification