ml-recall

#adam

15 questions from Optimization & Calculus, Training Deep Networks, Deep Learning Frameworks, Fine-tuning & PEFT

Timed test

15 questions · 15 min · score and explanations at the end

Take a test

Q 1 / 15Score 0 / 0

Shuffle questions

Question 1: Short answer

mediumAI-checkedOptimization & Calculus

Adam uses $\beta_1 = 0.9$ and $\beta_2 = 0.999$ with $\mathbf{v}_0 = \mathbf{s}_0 = 0$ . Ignore $\epsilon$ . On the first step, by what factor would the update be larger if you skipped bias correction, i.e. used $\mathbf{v}_1/\sqrt{\mathbf{s}_1}$ instead of $\hat{\mathbf{v}}_1/\sqrt{\hat{\mathbf{s}}_1}$ ?

Enter: check / next