22 duplicated lines in:
  - fairseq/criterions/cross_entropy.py (60:90, 40%)
  - fairseq/model_parallel/criterions/vocab_parallel_cross_entropy.py (58:87, 40%)

20 duplicated lines in:
  - fairseq/criterions/hubert_criterion.py (143:165, 13%)
  - fairseq/model_parallel/criterions/vocab_parallel_cross_entropy.py (56:78, 37%)

17 duplicated lines in:
  - fairseq/criterions/adaptive_loss.py (101:123, 22%)
  - fairseq/model_parallel/criterions/vocab_parallel_cross_entropy.py (65:87, 31%)

8 duplicated lines in:
  - fairseq/criterions/sentence_ranking.py (97:105, 10%)
  - fairseq/model_parallel/criterions/vocab_parallel_cross_entropy.py (63:71, 14%)

6 duplicated lines in:
  - fairseq/criterions/masked_lm.py (74:81, 11%)
  - fairseq/model_parallel/criterions/vocab_parallel_cross_entropy.py (54:61, 11%)

6 duplicated lines in:
  - fairseq/criterions/masked_lm.py (87:98, 11%)
  - fairseq/model_parallel/criterions/vocab_parallel_cross_entropy.py (76:87, 11%)