fix AttributeError: 'YayiTokenizer' object has no attribute 'sp_model'
#5
by
mzbac
- opened
- tokenization_yayi.py +3 -2
tokenization_yayi.py
CHANGED
@@ -71,6 +71,8 @@ class YayiTokenizer(PreTrainedTokenizer):
|
|
71 |
eos_token = AddedToken(eos_token, lstrip=False, rstrip=False) if isinstance(eos_token, str) else eos_token
|
72 |
unk_token = AddedToken(unk_token, lstrip=False, rstrip=False) if isinstance(unk_token, str) else unk_token
|
73 |
pad_token = AddedToken(pad_token, lstrip=False, rstrip=False) if isinstance(pad_token, str) else pad_token
|
|
|
|
|
74 |
super().__init__(
|
75 |
bos_token=bos_token,
|
76 |
eos_token=eos_token,
|
@@ -85,8 +87,7 @@ class YayiTokenizer(PreTrainedTokenizer):
|
|
85 |
self.vocab_file = vocab_file
|
86 |
self.add_bos_token = add_bos_token
|
87 |
self.add_eos_token = add_eos_token
|
88 |
-
|
89 |
-
self.sp_model.Load(vocab_file)
|
90 |
|
91 |
def __getstate__(self):
|
92 |
state = self.__dict__.copy()
|
|
|
71 |
eos_token = AddedToken(eos_token, lstrip=False, rstrip=False) if isinstance(eos_token, str) else eos_token
|
72 |
unk_token = AddedToken(unk_token, lstrip=False, rstrip=False) if isinstance(unk_token, str) else unk_token
|
73 |
pad_token = AddedToken(pad_token, lstrip=False, rstrip=False) if isinstance(pad_token, str) else pad_token
|
74 |
+
self.sp_model = spm.SentencePieceProcessor(**self.sp_model_kwargs)
|
75 |
+
self.sp_model.Load(vocab_file)
|
76 |
super().__init__(
|
77 |
bos_token=bos_token,
|
78 |
eos_token=eos_token,
|
|
|
87 |
self.vocab_file = vocab_file
|
88 |
self.add_bos_token = add_bos_token
|
89 |
self.add_eos_token = add_eos_token
|
90 |
+
|
|
|
91 |
|
92 |
def __getstate__(self):
|
93 |
state = self.__dict__.copy()
|