Upload tokenizer
Browse files- added_tokens.json +4 -1
- tokenizer.json +29 -2
- tokenizer_config.json +26 -2
added_tokens.json
CHANGED
|
@@ -1,4 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"<mask>": 20001,
|
| 3 |
-
"<unk>": 20000
|
|
|
|
|
|
|
|
|
|
| 4 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"<mask>": 20001,
|
| 3 |
+
"<unk>": 20000,
|
| 4 |
+
"<vi>": 20002,
|
| 5 |
+
"cv": 20004,
|
| 6 |
+
"zh": 20003
|
| 7 |
}
|
tokenizer.json
CHANGED
|
@@ -43,10 +43,37 @@
|
|
| 43 |
"id": 20001,
|
| 44 |
"content": "<mask>",
|
| 45 |
"single_word": false,
|
| 46 |
-
"lstrip":
|
| 47 |
"rstrip": false,
|
| 48 |
"normalized": true,
|
| 49 |
-
"special":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 50 |
}
|
| 51 |
],
|
| 52 |
"normalizer": null,
|
|
|
|
| 43 |
"id": 20001,
|
| 44 |
"content": "<mask>",
|
| 45 |
"single_word": false,
|
| 46 |
+
"lstrip": false,
|
| 47 |
"rstrip": false,
|
| 48 |
"normalized": true,
|
| 49 |
+
"special": false
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"id": 20002,
|
| 53 |
+
"content": "<vi>",
|
| 54 |
+
"single_word": false,
|
| 55 |
+
"lstrip": false,
|
| 56 |
+
"rstrip": false,
|
| 57 |
+
"normalized": true,
|
| 58 |
+
"special": false
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"id": 20003,
|
| 62 |
+
"content": "zh",
|
| 63 |
+
"single_word": false,
|
| 64 |
+
"lstrip": false,
|
| 65 |
+
"rstrip": false,
|
| 66 |
+
"normalized": true,
|
| 67 |
+
"special": false
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"id": 20004,
|
| 71 |
+
"content": "cv",
|
| 72 |
+
"single_word": false,
|
| 73 |
+
"lstrip": false,
|
| 74 |
+
"rstrip": false,
|
| 75 |
+
"normalized": true,
|
| 76 |
+
"special": false
|
| 77 |
}
|
| 78 |
],
|
| 79 |
"normalizer": null,
|
tokenizer_config.json
CHANGED
|
@@ -35,11 +35,35 @@
|
|
| 35 |
},
|
| 36 |
"20001": {
|
| 37 |
"content": "<mask>",
|
| 38 |
-
"lstrip":
|
| 39 |
"normalized": true,
|
| 40 |
"rstrip": false,
|
| 41 |
"single_word": false,
|
| 42 |
-
"special":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 43 |
}
|
| 44 |
},
|
| 45 |
"bos_token": "<s>",
|
|
|
|
| 35 |
},
|
| 36 |
"20001": {
|
| 37 |
"content": "<mask>",
|
| 38 |
+
"lstrip": false,
|
| 39 |
"normalized": true,
|
| 40 |
"rstrip": false,
|
| 41 |
"single_word": false,
|
| 42 |
+
"special": false
|
| 43 |
+
},
|
| 44 |
+
"20002": {
|
| 45 |
+
"content": "<vi>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": true,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false,
|
| 50 |
+
"special": false
|
| 51 |
+
},
|
| 52 |
+
"20003": {
|
| 53 |
+
"content": "zh",
|
| 54 |
+
"lstrip": false,
|
| 55 |
+
"normalized": true,
|
| 56 |
+
"rstrip": false,
|
| 57 |
+
"single_word": false,
|
| 58 |
+
"special": false
|
| 59 |
+
},
|
| 60 |
+
"20004": {
|
| 61 |
+
"content": "cv",
|
| 62 |
+
"lstrip": false,
|
| 63 |
+
"normalized": true,
|
| 64 |
+
"rstrip": false,
|
| 65 |
+
"single_word": false,
|
| 66 |
+
"special": false
|
| 67 |
}
|
| 68 |
},
|
| 69 |
"bos_token": "<s>",
|