EasyNLPnlp-240

Special Token Budget

Problem

Compute how many content tokens fit after special tokens and a generation reserve are subtracted from the context limit.

Return max(0, context_limit - n_special - reserve).

Hint

Clamp the result at zero so an overfull prompt reports no budget.

case 1

in: [512,3,64]

expected: 445

case 2

in: [128,200,0]

expected: 0

case 3

in: [0,0,0]

expected: 0

case 4

in: [2048,2,512]

expected: 1534

Starter codePython

def special_token_budget(context_limit, n_special, reserve):
    # Your code here
    pass

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