import torch
​
sentence_int = torch.tensor(
    [dc[s] for s in sentence.replace(',', '').split()]
)
print(sentence_int)