SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

cleaning - change ' to " (black requirements)

Browse Source
This commit is contained in:
VictorSanh
2020-01-10 19:34:25 -05:00
parent ebba9e929d
commit e83d9f1c1d
1 changed files with 4 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
examples/distillation/lm_seqs_dataset.py
View File

@@ -114,17 +114,17 @@ class LmSeqsDataset(Dataset):
""" """
Remove sequences with a (too) high level of unknown tokens. Remove sequences with a (too) high level of unknown tokens.
""" """
if 'unk_token' not in self.params.special_tok_ids: if "unk_token" not in self.params.special_tok_ids:
return return
else: else:
unk_token_id = self.params.special_tok_ids['unk_token'] unk_token_id = self.params.special_tok_ids["unk_token"]
init_size = len(self) init_size = len(self)
unk_occs = np.array([np.count_nonzero(a == unk_token_id) for a in self.token_ids]) unk_occs = np.array([np.count_nonzero(a == unk_token_id) for a in self.token_ids])
indices = (unk_occs / self.lengths) < 0.5 indices = (unk_occs / self.lengths) < 0.5
self.token_ids = self.token_ids[indices] self.token_ids = self.token_ids[indices]
self.lengths = self.lengths[indices] self.lengths = self.lengths[indices]
new_size = len(self) new_size = len(self)
logger.info(f'Remove {init_size - new_size} sequences with a high level of unknown tokens (50%).') logger.info(f"Remove {init_size - new_size} sequences with a high level of unknown tokens (50%).")
def print_statistics(self): def print_statistics(self):
""" """