Perplexity from Mean NLL

Problem

Return perplexity = exp(mean_nll) for a language model's mean negative log-likelihood measured in nats. Return 0.0 when mean_nll is negative.

case 1

in: [2.302585092994046]

expected: 10.000000000000002

case 2

in: [0]

expected: 1

case 3

in: [2]

expected: 7.38905609893065

case 4

in: [0.6931471805599453]

expected: 2

Starter codePython

import math
def perplexity_from_nll(mean_nll):
    # Your code here
    pass

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