Fix TiktokenTokenizer missing num_special_tokens_to_add (#30643)
Co-authored-by: Sushil Dubey <sushildubey171@users.noreply.github.com> Co-authored-by: Mohammad Miadh Angkad <176301910+mmangkad@users.noreply.github.com>
This commit is contained in:
co-authored by
Sushil Dubey
Mohammad Miadh Angkad
parent
7045e0fdff
commit
94de28764c
@@ -113,6 +113,10 @@ class TiktokenTokenizer:
|
|||||||
def decode(self, x, *args, **kwargs):
|
def decode(self, x, *args, **kwargs):
|
||||||
return self.tokenizer.decode(x)
|
return self.tokenizer.decode(x)
|
||||||
|
|
||||||
|
def num_special_tokens_to_add(self, *args, **kwargs) -> int:
|
||||||
|
# tiktoken's encode adds no special tokens (add_special_tokens is ignored).
|
||||||
|
return 0
|
||||||
|
|
||||||
def batch_decode(
|
def batch_decode(
|
||||||
self, batch, skip_special_tokens=True, spaces_between_special_tokens=False
|
self, batch, skip_special_tokens=True, spaces_between_special_tokens=False
|
||||||
):
|
):
|
||||||
|
|||||||
Reference in New Issue
Block a user