EasyNLPnlp-084

Token Type IDs

Problem

Build BERT token type ids for the packed pair [CLS] A [SEP] B [SEP].

Return len_a + 2 zeros followed by len_b + 1 ones.

Hint

The first [CLS] and [SEP] belong to segment 0; the second [SEP] closes segment 1.

case 1

in: [3,2]

expected: [0,0,0,0,0,1,1,1]

case 2

in: [1,1]

expected: [0,0,0,1,1]

case 3

in: [0,0]

expected: [0,0,1]

case 4

in: [2,0]

expected: [0,0,0,0,1]

Starter codePython

def token_type_ids(len_a, len_b):
    # Your code here
    pass

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