update tokenizer
Browse files- tokenizer.json +21 -5
tokenizer.json
CHANGED
|
@@ -31,12 +31,28 @@
|
|
| 31 |
"special": true
|
| 32 |
}
|
| 33 |
],
|
| 34 |
-
"normalizer":
|
|
|
|
|
|
|
|
|
|
| 35 |
"pre_tokenizer": {
|
| 36 |
-
"type": "
|
| 37 |
-
"
|
| 38 |
-
|
| 39 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 40 |
},
|
| 41 |
"post_processor": {
|
| 42 |
"type": "TemplateProcessing",
|
|
|
|
| 31 |
"special": true
|
| 32 |
}
|
| 33 |
],
|
| 34 |
+
"normalizer": {
|
| 35 |
+
"type": "Prepend",
|
| 36 |
+
"prepend": " "
|
| 37 |
+
},
|
| 38 |
"pre_tokenizer": {
|
| 39 |
+
"type": "Sequence",
|
| 40 |
+
"pretokenizers": [
|
| 41 |
+
{
|
| 42 |
+
"type": "Split",
|
| 43 |
+
"pattern": {
|
| 44 |
+
"Regex": "'s|'t|'re|'ve|'m|'ll|'d| ?[\\p{L}\\p{M}]+| ?\\p{N}+| ?[^\\s\\p{L}\\p{N}]+|\\s+(?!\\S)|\\s+"
|
| 45 |
+
},
|
| 46 |
+
"behavior": "Removed",
|
| 47 |
+
"invert": true
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"type": "ByteLevel",
|
| 51 |
+
"add_prefix_space": false,
|
| 52 |
+
"trim_offsets": true,
|
| 53 |
+
"use_regex": false
|
| 54 |
+
}
|
| 55 |
+
]
|
| 56 |
},
|
| 57 |
"post_processor": {
|
| 58 |
"type": "TemplateProcessing",
|