Update-To-Data Ratio

Problem

Check whether the update-to-data ratio matches a target: |gradient_steps / new_samples - target_utd| <= tol. A non-positive sample count returns False.

Hint

UTD above 1 means the critic is updated more than once per new transition.

Test cases

case 1

in: [20,10,2,1e-9]

expected: true

case 2

in: [20,10,3,1e-9]

expected: false

case 3

in: [5,0,1,1e-9]

expected: false

Starter codePython

def utd_ratio_check(gradient_steps, new_samples, target_utd, tol):
    # Your code here
    pass

3 test cases · run your code against them in the editor