infi commited on
Upload tokenizer
Browse files- special_tokens_map.json +28 -4
- tokenizer.json +4 -4
- tokenizer_config.json +4 -4
special_tokens_map.json
CHANGED
|
@@ -1,6 +1,30 @@
|
|
| 1 |
{
|
| 2 |
-
"bos_token":
|
| 3 |
-
|
| 4 |
-
|
| 5 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<|startbox|>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "<|endbox|>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": {
|
| 17 |
+
"content": "<pad>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"content": "<|unkbox|>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
}
|
| 30 |
}
|
tokenizer.json
CHANGED
|
@@ -54,8 +54,8 @@
|
|
| 54 |
"single_word": false,
|
| 55 |
"lstrip": false,
|
| 56 |
"rstrip": false,
|
| 57 |
-
"normalized":
|
| 58 |
-
"special":
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"id": 34005,
|
|
@@ -63,8 +63,8 @@
|
|
| 63 |
"single_word": false,
|
| 64 |
"lstrip": false,
|
| 65 |
"rstrip": false,
|
| 66 |
-
"normalized":
|
| 67 |
-
"special":
|
| 68 |
},
|
| 69 |
{
|
| 70 |
"id": 34006,
|
|
|
|
| 54 |
"single_word": false,
|
| 55 |
"lstrip": false,
|
| 56 |
"rstrip": false,
|
| 57 |
+
"normalized": true,
|
| 58 |
+
"special": false
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"id": 34005,
|
|
|
|
| 63 |
"single_word": false,
|
| 64 |
"lstrip": false,
|
| 65 |
"rstrip": false,
|
| 66 |
+
"normalized": true,
|
| 67 |
+
"special": false
|
| 68 |
},
|
| 69 |
{
|
| 70 |
"id": 34006,
|
tokenizer_config.json
CHANGED
|
@@ -44,18 +44,18 @@
|
|
| 44 |
"34004": {
|
| 45 |
"content": "<|startbox|>",
|
| 46 |
"lstrip": false,
|
| 47 |
-
"normalized":
|
| 48 |
"rstrip": false,
|
| 49 |
"single_word": false,
|
| 50 |
-
"special":
|
| 51 |
},
|
| 52 |
"34005": {
|
| 53 |
"content": "<|endbox|>",
|
| 54 |
"lstrip": false,
|
| 55 |
-
"normalized":
|
| 56 |
"rstrip": false,
|
| 57 |
"single_word": false,
|
| 58 |
-
"special":
|
| 59 |
},
|
| 60 |
"34006": {
|
| 61 |
"content": "<|para|>",
|
|
|
|
| 44 |
"34004": {
|
| 45 |
"content": "<|startbox|>",
|
| 46 |
"lstrip": false,
|
| 47 |
+
"normalized": true,
|
| 48 |
"rstrip": false,
|
| 49 |
"single_word": false,
|
| 50 |
+
"special": false
|
| 51 |
},
|
| 52 |
"34005": {
|
| 53 |
"content": "<|endbox|>",
|
| 54 |
"lstrip": false,
|
| 55 |
+
"normalized": true,
|
| 56 |
"rstrip": false,
|
| 57 |
"single_word": false,
|
| 58 |
+
"special": false
|
| 59 |
},
|
| 60 |
"34006": {
|
| 61 |
"content": "<|para|>",
|