Quiz ci wàcci degrade

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

1. In plain gradient descent, which direction is used relative to the objective's gradient?
   • The negative-gradient direction
   • The positive-gradient direction
   • An arbitrary perpendicular direction by definition
   • A direction chosen without evaluating any gradient
2. Lan mooy saytu tolluwaayu jàng bi?
   • Limu diisaay yi ci model bi
   • Ñaata klaas la model bi mëna wax
   • Dayo jéego bu nekk buñu jël ci diiru optimisation
   • Bariwaayu done yiñ jëfandikoo ci tàggat yaram
3. How does a mini-batch gradient relate to the full selected training-data objective?
   • It uses no examples
   • It is computed on the batch and used as an estimate of the full-data gradient
   • It must equal the full-data gradient on every step
   • It only exists for images
4. Lan mooy jafe-jafe bi nit ñi di jàng bu bari lool mëna jur?
   • Taggat yaram dafay yeex lool waaye dafay dëgër
   • Modèle bi du bàyyi xel ci degrade bi
   • Done yi dañu yàqu
   • Modèle bi dafa weesu li gëna ndaw te mën na wuute
5. What information does momentum carry into later gradient-descent updates?
   • Only the model's vocabulary size
   • A permanent deletion of all earlier gradients
   • A running weighted combination of earlier gradients
   • An unrelated random reset of every weight
6. Lan mooy màndarga bi gëna am solo ci optimizer Adam buñu ko méngale ak wàccinu gradient bu leer?
   • Dafay méngale tolluwaayu jàng bu baax ci parametre bu nekk
   • Dafay jëfandikoo benn jéego bu takku ci parametre bu nekk
   • Soxlawul ay degrade
   • Dafay tàggat rek couche de sortie bi
7. For plain gradient descent with old weight two, gradient three, and learning rate one-tenth, what is the new weight?
   • Two-point-three
   • One-point-nine-seven
   • Ñatt
   • One-point-seven
8. What does a zero gradient fail to prove for a nonconvex training objective?
   • That the gradient value is zero at the evaluated point
   • That the current point is a global minimum
   • That the objective has been evaluated locally
   • That a stationary-point condition holds