Training done
Browse files- added_tokens.json +14 -14
- tokenizer.json +16 -30
- tokenizer_config.json +14 -14
added_tokens.json
CHANGED
|
@@ -1,20 +1,20 @@
|
|
| 1 |
{
|
| 2 |
-
"</s_dob>":
|
| 3 |
-
"</s_expiry>":
|
| 4 |
-
"</s_id>":
|
| 5 |
-
"</s_issue>":
|
| 6 |
-
"</s_name>":
|
| 7 |
-
"</s_nationality>":
|
| 8 |
-
"</s_sex>":
|
| 9 |
"<s_cord-v2>": 57539,
|
| 10 |
-
"<s_dob>":
|
| 11 |
-
"<s_expiry>":
|
| 12 |
-
"<s_id>":
|
| 13 |
"<s_iitcdip>": 57523,
|
| 14 |
-
"<s_issue>":
|
| 15 |
-
"<s_name>":
|
| 16 |
-
"<s_nationality>":
|
| 17 |
-
"<s_sex>":
|
| 18 |
"<s_synthdog>": 57524,
|
| 19 |
"<sep/>": 57522
|
| 20 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"</s_dob>": 57538,
|
| 3 |
+
"</s_expiry>": 57530,
|
| 4 |
+
"</s_id>": 57532,
|
| 5 |
+
"</s_issue>": 57528,
|
| 6 |
+
"</s_name>": 57534,
|
| 7 |
+
"</s_nationality>": 57526,
|
| 8 |
+
"</s_sex>": 57536,
|
| 9 |
"<s_cord-v2>": 57539,
|
| 10 |
+
"<s_dob>": 57537,
|
| 11 |
+
"<s_expiry>": 57529,
|
| 12 |
+
"<s_id>": 57531,
|
| 13 |
"<s_iitcdip>": 57523,
|
| 14 |
+
"<s_issue>": 57527,
|
| 15 |
+
"<s_name>": 57533,
|
| 16 |
+
"<s_nationality>": 57525,
|
| 17 |
+
"<s_sex>": 57535,
|
| 18 |
"<s_synthdog>": 57524,
|
| 19 |
"<sep/>": 57522
|
| 20 |
}
|
tokenizer.json
CHANGED
|
@@ -1,21 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"version": "1.0",
|
| 3 |
-
"truncation":
|
| 4 |
-
|
| 5 |
-
"max_length": 768,
|
| 6 |
-
"strategy": "LongestFirst",
|
| 7 |
-
"stride": 0
|
| 8 |
-
},
|
| 9 |
-
"padding": {
|
| 10 |
-
"strategy": {
|
| 11 |
-
"Fixed": 768
|
| 12 |
-
},
|
| 13 |
-
"direction": "Right",
|
| 14 |
-
"pad_to_multiple_of": null,
|
| 15 |
-
"pad_id": 1,
|
| 16 |
-
"pad_type_id": 0,
|
| 17 |
-
"pad_token": "<pad>"
|
| 18 |
-
},
|
| 19 |
"added_tokens": [
|
| 20 |
{
|
| 21 |
"id": 0,
|
|
@@ -91,7 +77,7 @@
|
|
| 91 |
},
|
| 92 |
{
|
| 93 |
"id": 57525,
|
| 94 |
-
"content": "<
|
| 95 |
"single_word": false,
|
| 96 |
"lstrip": false,
|
| 97 |
"rstrip": false,
|
|
@@ -100,7 +86,7 @@
|
|
| 100 |
},
|
| 101 |
{
|
| 102 |
"id": 57526,
|
| 103 |
-
"content": "</
|
| 104 |
"single_word": false,
|
| 105 |
"lstrip": false,
|
| 106 |
"rstrip": false,
|
|
@@ -109,7 +95,7 @@
|
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"id": 57527,
|
| 112 |
-
"content": "<
|
| 113 |
"single_word": false,
|
| 114 |
"lstrip": false,
|
| 115 |
"rstrip": false,
|
|
@@ -118,7 +104,7 @@
|
|
| 118 |
},
|
| 119 |
{
|
| 120 |
"id": 57528,
|
| 121 |
-
"content": "</
|
| 122 |
"single_word": false,
|
| 123 |
"lstrip": false,
|
| 124 |
"rstrip": false,
|
|
@@ -127,7 +113,7 @@
|
|
| 127 |
},
|
| 128 |
{
|
| 129 |
"id": 57529,
|
| 130 |
-
"content": "<
|
| 131 |
"single_word": false,
|
| 132 |
"lstrip": false,
|
| 133 |
"rstrip": false,
|
|
@@ -136,7 +122,7 @@
|
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"id": 57530,
|
| 139 |
-
"content": "</
|
| 140 |
"single_word": false,
|
| 141 |
"lstrip": false,
|
| 142 |
"rstrip": false,
|
|
@@ -145,7 +131,7 @@
|
|
| 145 |
},
|
| 146 |
{
|
| 147 |
"id": 57531,
|
| 148 |
-
"content": "<
|
| 149 |
"single_word": false,
|
| 150 |
"lstrip": false,
|
| 151 |
"rstrip": false,
|
|
@@ -154,7 +140,7 @@
|
|
| 154 |
},
|
| 155 |
{
|
| 156 |
"id": 57532,
|
| 157 |
-
"content": "</
|
| 158 |
"single_word": false,
|
| 159 |
"lstrip": false,
|
| 160 |
"rstrip": false,
|
|
@@ -163,7 +149,7 @@
|
|
| 163 |
},
|
| 164 |
{
|
| 165 |
"id": 57533,
|
| 166 |
-
"content": "<
|
| 167 |
"single_word": false,
|
| 168 |
"lstrip": false,
|
| 169 |
"rstrip": false,
|
|
@@ -172,7 +158,7 @@
|
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"id": 57534,
|
| 175 |
-
"content": "</
|
| 176 |
"single_word": false,
|
| 177 |
"lstrip": false,
|
| 178 |
"rstrip": false,
|
|
@@ -181,7 +167,7 @@
|
|
| 181 |
},
|
| 182 |
{
|
| 183 |
"id": 57535,
|
| 184 |
-
"content": "<
|
| 185 |
"single_word": false,
|
| 186 |
"lstrip": false,
|
| 187 |
"rstrip": false,
|
|
@@ -190,7 +176,7 @@
|
|
| 190 |
},
|
| 191 |
{
|
| 192 |
"id": 57536,
|
| 193 |
-
"content": "</
|
| 194 |
"single_word": false,
|
| 195 |
"lstrip": false,
|
| 196 |
"rstrip": false,
|
|
@@ -199,7 +185,7 @@
|
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"id": 57537,
|
| 202 |
-
"content": "<
|
| 203 |
"single_word": false,
|
| 204 |
"lstrip": false,
|
| 205 |
"rstrip": false,
|
|
@@ -208,7 +194,7 @@
|
|
| 208 |
},
|
| 209 |
{
|
| 210 |
"id": 57538,
|
| 211 |
-
"content": "</
|
| 212 |
"single_word": false,
|
| 213 |
"lstrip": false,
|
| 214 |
"rstrip": false,
|
|
|
|
| 1 |
{
|
| 2 |
"version": "1.0",
|
| 3 |
+
"truncation": null,
|
| 4 |
+
"padding": null,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
"added_tokens": [
|
| 6 |
{
|
| 7 |
"id": 0,
|
|
|
|
| 77 |
},
|
| 78 |
{
|
| 79 |
"id": 57525,
|
| 80 |
+
"content": "<s_nationality>",
|
| 81 |
"single_word": false,
|
| 82 |
"lstrip": false,
|
| 83 |
"rstrip": false,
|
|
|
|
| 86 |
},
|
| 87 |
{
|
| 88 |
"id": 57526,
|
| 89 |
+
"content": "</s_nationality>",
|
| 90 |
"single_word": false,
|
| 91 |
"lstrip": false,
|
| 92 |
"rstrip": false,
|
|
|
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"id": 57527,
|
| 98 |
+
"content": "<s_issue>",
|
| 99 |
"single_word": false,
|
| 100 |
"lstrip": false,
|
| 101 |
"rstrip": false,
|
|
|
|
| 104 |
},
|
| 105 |
{
|
| 106 |
"id": 57528,
|
| 107 |
+
"content": "</s_issue>",
|
| 108 |
"single_word": false,
|
| 109 |
"lstrip": false,
|
| 110 |
"rstrip": false,
|
|
|
|
| 113 |
},
|
| 114 |
{
|
| 115 |
"id": 57529,
|
| 116 |
+
"content": "<s_expiry>",
|
| 117 |
"single_word": false,
|
| 118 |
"lstrip": false,
|
| 119 |
"rstrip": false,
|
|
|
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"id": 57530,
|
| 125 |
+
"content": "</s_expiry>",
|
| 126 |
"single_word": false,
|
| 127 |
"lstrip": false,
|
| 128 |
"rstrip": false,
|
|
|
|
| 131 |
},
|
| 132 |
{
|
| 133 |
"id": 57531,
|
| 134 |
+
"content": "<s_id>",
|
| 135 |
"single_word": false,
|
| 136 |
"lstrip": false,
|
| 137 |
"rstrip": false,
|
|
|
|
| 140 |
},
|
| 141 |
{
|
| 142 |
"id": 57532,
|
| 143 |
+
"content": "</s_id>",
|
| 144 |
"single_word": false,
|
| 145 |
"lstrip": false,
|
| 146 |
"rstrip": false,
|
|
|
|
| 149 |
},
|
| 150 |
{
|
| 151 |
"id": 57533,
|
| 152 |
+
"content": "<s_name>",
|
| 153 |
"single_word": false,
|
| 154 |
"lstrip": false,
|
| 155 |
"rstrip": false,
|
|
|
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"id": 57534,
|
| 161 |
+
"content": "</s_name>",
|
| 162 |
"single_word": false,
|
| 163 |
"lstrip": false,
|
| 164 |
"rstrip": false,
|
|
|
|
| 167 |
},
|
| 168 |
{
|
| 169 |
"id": 57535,
|
| 170 |
+
"content": "<s_sex>",
|
| 171 |
"single_word": false,
|
| 172 |
"lstrip": false,
|
| 173 |
"rstrip": false,
|
|
|
|
| 176 |
},
|
| 177 |
{
|
| 178 |
"id": 57536,
|
| 179 |
+
"content": "</s_sex>",
|
| 180 |
"single_word": false,
|
| 181 |
"lstrip": false,
|
| 182 |
"rstrip": false,
|
|
|
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"id": 57537,
|
| 188 |
+
"content": "<s_dob>",
|
| 189 |
"single_word": false,
|
| 190 |
"lstrip": false,
|
| 191 |
"rstrip": false,
|
|
|
|
| 194 |
},
|
| 195 |
{
|
| 196 |
"id": 57538,
|
| 197 |
+
"content": "</s_dob>",
|
| 198 |
"single_word": false,
|
| 199 |
"lstrip": false,
|
| 200 |
"rstrip": false,
|
tokenizer_config.json
CHANGED
|
@@ -65,7 +65,7 @@
|
|
| 65 |
"special": true
|
| 66 |
},
|
| 67 |
"57525": {
|
| 68 |
-
"content": "<
|
| 69 |
"lstrip": false,
|
| 70 |
"normalized": true,
|
| 71 |
"rstrip": false,
|
|
@@ -73,7 +73,7 @@
|
|
| 73 |
"special": false
|
| 74 |
},
|
| 75 |
"57526": {
|
| 76 |
-
"content": "</
|
| 77 |
"lstrip": false,
|
| 78 |
"normalized": true,
|
| 79 |
"rstrip": false,
|
|
@@ -81,7 +81,7 @@
|
|
| 81 |
"special": false
|
| 82 |
},
|
| 83 |
"57527": {
|
| 84 |
-
"content": "<
|
| 85 |
"lstrip": false,
|
| 86 |
"normalized": true,
|
| 87 |
"rstrip": false,
|
|
@@ -89,7 +89,7 @@
|
|
| 89 |
"special": false
|
| 90 |
},
|
| 91 |
"57528": {
|
| 92 |
-
"content": "</
|
| 93 |
"lstrip": false,
|
| 94 |
"normalized": true,
|
| 95 |
"rstrip": false,
|
|
@@ -97,7 +97,7 @@
|
|
| 97 |
"special": false
|
| 98 |
},
|
| 99 |
"57529": {
|
| 100 |
-
"content": "<
|
| 101 |
"lstrip": false,
|
| 102 |
"normalized": true,
|
| 103 |
"rstrip": false,
|
|
@@ -105,7 +105,7 @@
|
|
| 105 |
"special": false
|
| 106 |
},
|
| 107 |
"57530": {
|
| 108 |
-
"content": "</
|
| 109 |
"lstrip": false,
|
| 110 |
"normalized": true,
|
| 111 |
"rstrip": false,
|
|
@@ -113,7 +113,7 @@
|
|
| 113 |
"special": false
|
| 114 |
},
|
| 115 |
"57531": {
|
| 116 |
-
"content": "<
|
| 117 |
"lstrip": false,
|
| 118 |
"normalized": true,
|
| 119 |
"rstrip": false,
|
|
@@ -121,7 +121,7 @@
|
|
| 121 |
"special": false
|
| 122 |
},
|
| 123 |
"57532": {
|
| 124 |
-
"content": "</
|
| 125 |
"lstrip": false,
|
| 126 |
"normalized": true,
|
| 127 |
"rstrip": false,
|
|
@@ -129,7 +129,7 @@
|
|
| 129 |
"special": false
|
| 130 |
},
|
| 131 |
"57533": {
|
| 132 |
-
"content": "<
|
| 133 |
"lstrip": false,
|
| 134 |
"normalized": true,
|
| 135 |
"rstrip": false,
|
|
@@ -137,7 +137,7 @@
|
|
| 137 |
"special": false
|
| 138 |
},
|
| 139 |
"57534": {
|
| 140 |
-
"content": "</
|
| 141 |
"lstrip": false,
|
| 142 |
"normalized": true,
|
| 143 |
"rstrip": false,
|
|
@@ -145,7 +145,7 @@
|
|
| 145 |
"special": false
|
| 146 |
},
|
| 147 |
"57535": {
|
| 148 |
-
"content": "<
|
| 149 |
"lstrip": false,
|
| 150 |
"normalized": true,
|
| 151 |
"rstrip": false,
|
|
@@ -153,7 +153,7 @@
|
|
| 153 |
"special": false
|
| 154 |
},
|
| 155 |
"57536": {
|
| 156 |
-
"content": "</
|
| 157 |
"lstrip": false,
|
| 158 |
"normalized": true,
|
| 159 |
"rstrip": false,
|
|
@@ -161,7 +161,7 @@
|
|
| 161 |
"special": false
|
| 162 |
},
|
| 163 |
"57537": {
|
| 164 |
-
"content": "<
|
| 165 |
"lstrip": false,
|
| 166 |
"normalized": true,
|
| 167 |
"rstrip": false,
|
|
@@ -169,7 +169,7 @@
|
|
| 169 |
"special": false
|
| 170 |
},
|
| 171 |
"57538": {
|
| 172 |
-
"content": "</
|
| 173 |
"lstrip": false,
|
| 174 |
"normalized": true,
|
| 175 |
"rstrip": false,
|
|
|
|
| 65 |
"special": true
|
| 66 |
},
|
| 67 |
"57525": {
|
| 68 |
+
"content": "<s_nationality>",
|
| 69 |
"lstrip": false,
|
| 70 |
"normalized": true,
|
| 71 |
"rstrip": false,
|
|
|
|
| 73 |
"special": false
|
| 74 |
},
|
| 75 |
"57526": {
|
| 76 |
+
"content": "</s_nationality>",
|
| 77 |
"lstrip": false,
|
| 78 |
"normalized": true,
|
| 79 |
"rstrip": false,
|
|
|
|
| 81 |
"special": false
|
| 82 |
},
|
| 83 |
"57527": {
|
| 84 |
+
"content": "<s_issue>",
|
| 85 |
"lstrip": false,
|
| 86 |
"normalized": true,
|
| 87 |
"rstrip": false,
|
|
|
|
| 89 |
"special": false
|
| 90 |
},
|
| 91 |
"57528": {
|
| 92 |
+
"content": "</s_issue>",
|
| 93 |
"lstrip": false,
|
| 94 |
"normalized": true,
|
| 95 |
"rstrip": false,
|
|
|
|
| 97 |
"special": false
|
| 98 |
},
|
| 99 |
"57529": {
|
| 100 |
+
"content": "<s_expiry>",
|
| 101 |
"lstrip": false,
|
| 102 |
"normalized": true,
|
| 103 |
"rstrip": false,
|
|
|
|
| 105 |
"special": false
|
| 106 |
},
|
| 107 |
"57530": {
|
| 108 |
+
"content": "</s_expiry>",
|
| 109 |
"lstrip": false,
|
| 110 |
"normalized": true,
|
| 111 |
"rstrip": false,
|
|
|
|
| 113 |
"special": false
|
| 114 |
},
|
| 115 |
"57531": {
|
| 116 |
+
"content": "<s_id>",
|
| 117 |
"lstrip": false,
|
| 118 |
"normalized": true,
|
| 119 |
"rstrip": false,
|
|
|
|
| 121 |
"special": false
|
| 122 |
},
|
| 123 |
"57532": {
|
| 124 |
+
"content": "</s_id>",
|
| 125 |
"lstrip": false,
|
| 126 |
"normalized": true,
|
| 127 |
"rstrip": false,
|
|
|
|
| 129 |
"special": false
|
| 130 |
},
|
| 131 |
"57533": {
|
| 132 |
+
"content": "<s_name>",
|
| 133 |
"lstrip": false,
|
| 134 |
"normalized": true,
|
| 135 |
"rstrip": false,
|
|
|
|
| 137 |
"special": false
|
| 138 |
},
|
| 139 |
"57534": {
|
| 140 |
+
"content": "</s_name>",
|
| 141 |
"lstrip": false,
|
| 142 |
"normalized": true,
|
| 143 |
"rstrip": false,
|
|
|
|
| 145 |
"special": false
|
| 146 |
},
|
| 147 |
"57535": {
|
| 148 |
+
"content": "<s_sex>",
|
| 149 |
"lstrip": false,
|
| 150 |
"normalized": true,
|
| 151 |
"rstrip": false,
|
|
|
|
| 153 |
"special": false
|
| 154 |
},
|
| 155 |
"57536": {
|
| 156 |
+
"content": "</s_sex>",
|
| 157 |
"lstrip": false,
|
| 158 |
"normalized": true,
|
| 159 |
"rstrip": false,
|
|
|
|
| 161 |
"special": false
|
| 162 |
},
|
| 163 |
"57537": {
|
| 164 |
+
"content": "<s_dob>",
|
| 165 |
"lstrip": false,
|
| 166 |
"normalized": true,
|
| 167 |
"rstrip": false,
|
|
|
|
| 169 |
"special": false
|
| 170 |
},
|
| 171 |
"57538": {
|
| 172 |
+
"content": "</s_dob>",
|
| 173 |
"lstrip": false,
|
| 174 |
"normalized": true,
|
| 175 |
"rstrip": false,
|