shainaraza
commited on
Commit
•
f069dac
1
Parent(s):
0f3a69a
Upload tokenizer
Browse files- added_tokens.json +4 -0
- special_tokens_map.json +4 -0
- tokenizer.json +18 -0
added_tokens.json
ADDED
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"<BIASED>": 30522,
|
3 |
+
"<NON_BIASED>": 30523
|
4 |
+
}
|
special_tokens_map.json
CHANGED
@@ -1,4 +1,8 @@
|
|
1 |
{
|
|
|
|
|
|
|
|
|
2 |
"cls_token": "[CLS]",
|
3 |
"mask_token": "[MASK]",
|
4 |
"pad_token": "[PAD]",
|
|
|
1 |
{
|
2 |
+
"additional_special_tokens": [
|
3 |
+
"<BIASED>",
|
4 |
+
"<NON_BIASED>"
|
5 |
+
],
|
6 |
"cls_token": "[CLS]",
|
7 |
"mask_token": "[MASK]",
|
8 |
"pad_token": "[PAD]",
|
tokenizer.json
CHANGED
@@ -59,6 +59,24 @@
|
|
59 |
"rstrip": false,
|
60 |
"normalized": false,
|
61 |
"special": true
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
62 |
}
|
63 |
],
|
64 |
"normalizer": {
|
|
|
59 |
"rstrip": false,
|
60 |
"normalized": false,
|
61 |
"special": true
|
62 |
+
},
|
63 |
+
{
|
64 |
+
"id": 30522,
|
65 |
+
"content": "<BIASED>",
|
66 |
+
"single_word": false,
|
67 |
+
"lstrip": false,
|
68 |
+
"rstrip": false,
|
69 |
+
"normalized": true,
|
70 |
+
"special": false
|
71 |
+
},
|
72 |
+
{
|
73 |
+
"id": 30523,
|
74 |
+
"content": "<NON_BIASED>",
|
75 |
+
"single_word": false,
|
76 |
+
"lstrip": false,
|
77 |
+
"rstrip": false,
|
78 |
+
"normalized": true,
|
79 |
+
"special": false
|
80 |
}
|
81 |
],
|
82 |
"normalizer": {
|