[FIXED] how the code '-input[range(target.shape[0]),target]' works?

Issue

I'm learing pytorch.Reading the official tutorial,I met the preplexing code. input is a tensor, so is target.

def nll(input,target):
    return -input[range(target.shape[0]),target].mean()

And the pred is:

target is:

the '-input[range(target.shape[0]),target]' is:

Output shows this is not substracting target from input or merging two tensors

Solution

The code input[range(target.shape[0]), target] simply picks, from each row i of input the element at column indicated by the corresponding element of target, that is target[i].
In other words, if out = input[range(target.shape[0]), target] then out[i] = input[i, target[i]].

This is very similar to torch.gather.

Answered By - Shai