MediumNLPnlp-199

Language Model Rescoring Gain

Problem

Return the average log-probability gain from rescoring: mean(rescored_logprobs) - mean(base_logprobs), zipped to the shorter list.

Return 0.0 when either list is empty.

Hint

A positive gain means rescoring improved the average log probability.

case 1

in: [[-1,-2],[-0.5,-1]]

expected: 0.75

case 2

in: [[0],[1]]

expected: 1

case 3

in: [[],[]]

expected: 0

case 4

in: [[-1,-1,-1],[-2,-2]]

expected: -1

case 5

in: [[5],[5]]

expected: 0

Starter codePython

def lm_rescoring_gain(base_logprobs, rescored_logprobs):
    # Your code here
    pass

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