Chang-1202
commited on
Commit
•
b5532db
1
Parent(s):
56f1e27
Update tokenizer_config.json
Browse files- tokenizer_config.json +7 -8
tokenizer_config.json
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
{
|
2 |
-
"tokenizer_class": "
|
3 |
"do_lower_case": false,
|
4 |
"do_word_tokenize": true,
|
5 |
"do_subword_tokenize": true,
|
@@ -7,12 +7,11 @@
|
|
7 |
"word_form_type": "dictionary_and_surface",
|
8 |
"subword_tokenizer_type": "wordpiece",
|
9 |
"model_max_length": 512,
|
10 |
-
"sudachipy_kwargs": {"split_mode":"A","dict_type":"core"}
|
11 |
"auto_map": {
|
12 |
-
|
13 |
-
|
14 |
-
|
15 |
-
|
16 |
}
|
17 |
-
}
|
18 |
-
|
|
|
1 |
{
|
2 |
+
"tokenizer_class": "ElectraSudachipyTokenizer",
|
3 |
"do_lower_case": false,
|
4 |
"do_word_tokenize": true,
|
5 |
"do_subword_tokenize": true,
|
|
|
7 |
"word_form_type": "dictionary_and_surface",
|
8 |
"subword_tokenizer_type": "wordpiece",
|
9 |
"model_max_length": 512,
|
10 |
+
"sudachipy_kwargs": {"split_mode":"A","dict_type":"core"},
|
11 |
"auto_map": {
|
12 |
+
"AutoTokenizer": [
|
13 |
+
"modeling.ElectraSudachipyTokenizer",
|
14 |
+
null
|
15 |
+
]
|
16 |
}
|
17 |
+
}
|
|