MediumNLPnlp-392

Token Level Precision Generation

Problem

Generation precision is the multiset overlap divided by the number of predicted tokens; return 0.0 when empty.

Given predicted and reference tokens, return the precision.

Hint

Count matched multiplicities before dividing by the prediction length.

case 1

in: [["a","b"],["a","b","c"]]

expected: 1

case 2

in: [["x"],["y"]]

expected: 0

case 3

in: [[],["a"]]

expected: 0

Starter codePython

def token_level_precision_generation(predicted, reference):
    # Your code here
    pass

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