Commit
•
f4b6143
1
Parent(s):
8108334
Upload processor
Browse files- added_tokens.json +8 -1
- tokenizer.json +63 -0
added_tokens.json
CHANGED
|
@@ -1,5 +1,12 @@
|
|
| 1 |
{
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
"<s_iitcdip>": 57523,
|
|
|
|
| 3 |
"<s_synthdog>": 57524,
|
| 4 |
-
"<sep/>": 57522
|
|
|
|
| 5 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"</s_answer>": 57530,
|
| 3 |
+
"</s_question>": 57528,
|
| 4 |
+
"<no/>": 57526,
|
| 5 |
+
"<s_answer>": 57529,
|
| 6 |
+
"<s_docvqa>": 57531,
|
| 7 |
"<s_iitcdip>": 57523,
|
| 8 |
+
"<s_question>": 57527,
|
| 9 |
"<s_synthdog>": 57524,
|
| 10 |
+
"<sep/>": 57522,
|
| 11 |
+
"<yes/>": 57525
|
| 12 |
}
|
tokenizer.json
CHANGED
|
@@ -88,6 +88,69 @@
|
|
| 88 |
"rstrip": false,
|
| 89 |
"normalized": false,
|
| 90 |
"special": true
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 91 |
}
|
| 92 |
],
|
| 93 |
"normalizer": {
|
|
|
|
| 88 |
"rstrip": false,
|
| 89 |
"normalized": false,
|
| 90 |
"special": true
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"id": 57525,
|
| 94 |
+
"content": "<yes/>",
|
| 95 |
+
"single_word": false,
|
| 96 |
+
"lstrip": false,
|
| 97 |
+
"rstrip": false,
|
| 98 |
+
"normalized": true,
|
| 99 |
+
"special": false
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"id": 57526,
|
| 103 |
+
"content": "<no/>",
|
| 104 |
+
"single_word": false,
|
| 105 |
+
"lstrip": false,
|
| 106 |
+
"rstrip": false,
|
| 107 |
+
"normalized": true,
|
| 108 |
+
"special": false
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"id": 57527,
|
| 112 |
+
"content": "<s_question>",
|
| 113 |
+
"single_word": false,
|
| 114 |
+
"lstrip": false,
|
| 115 |
+
"rstrip": false,
|
| 116 |
+
"normalized": true,
|
| 117 |
+
"special": false
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"id": 57528,
|
| 121 |
+
"content": "</s_question>",
|
| 122 |
+
"single_word": false,
|
| 123 |
+
"lstrip": false,
|
| 124 |
+
"rstrip": false,
|
| 125 |
+
"normalized": true,
|
| 126 |
+
"special": false
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"id": 57529,
|
| 130 |
+
"content": "<s_answer>",
|
| 131 |
+
"single_word": false,
|
| 132 |
+
"lstrip": false,
|
| 133 |
+
"rstrip": false,
|
| 134 |
+
"normalized": true,
|
| 135 |
+
"special": false
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"id": 57530,
|
| 139 |
+
"content": "</s_answer>",
|
| 140 |
+
"single_word": false,
|
| 141 |
+
"lstrip": false,
|
| 142 |
+
"rstrip": false,
|
| 143 |
+
"normalized": true,
|
| 144 |
+
"special": false
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"id": 57531,
|
| 148 |
+
"content": "<s_docvqa>",
|
| 149 |
+
"single_word": false,
|
| 150 |
+
"lstrip": false,
|
| 151 |
+
"rstrip": false,
|
| 152 |
+
"normalized": true,
|
| 153 |
+
"special": false
|
| 154 |
}
|
| 155 |
],
|
| 156 |
"normalizer": {
|