guillermoruiz
commited on
Commit
•
6575a02
1
Parent(s):
5767724
Upload tokenizer
Browse files- tokenizer.json +57 -21
tokenizer.json
CHANGED
@@ -1,21 +1,7 @@
|
|
1 |
{
|
2 |
"version": "1.0",
|
3 |
-
"truncation":
|
4 |
-
|
5 |
-
"max_length": 280,
|
6 |
-
"strategy": "LongestFirst",
|
7 |
-
"stride": 0
|
8 |
-
},
|
9 |
-
"padding": {
|
10 |
-
"strategy": {
|
11 |
-
"Fixed": 280
|
12 |
-
},
|
13 |
-
"direction": "Right",
|
14 |
-
"pad_to_multiple_of": null,
|
15 |
-
"pad_id": 0,
|
16 |
-
"pad_type_id": 0,
|
17 |
-
"pad_token": "[PAD]"
|
18 |
-
},
|
19 |
"added_tokens": [
|
20 |
{
|
21 |
"id": 0,
|
@@ -64,11 +50,61 @@
|
|
64 |
}
|
65 |
],
|
66 |
"normalizer": {
|
67 |
-
"type": "
|
68 |
-
"
|
69 |
-
|
70 |
-
|
71 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
72 |
},
|
73 |
"pre_tokenizer": {
|
74 |
"type": "BertPreTokenizer"
|
|
|
1 |
{
|
2 |
"version": "1.0",
|
3 |
+
"truncation": null,
|
4 |
+
"padding": null,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
5 |
"added_tokens": [
|
6 |
{
|
7 |
"id": 0,
|
|
|
50 |
}
|
51 |
],
|
52 |
"normalizer": {
|
53 |
+
"type": "Sequence",
|
54 |
+
"normalizers": [
|
55 |
+
{
|
56 |
+
"type": "NFD"
|
57 |
+
},
|
58 |
+
{
|
59 |
+
"type": "Replace",
|
60 |
+
"pattern": {
|
61 |
+
"Regex": "@\\S+"
|
62 |
+
},
|
63 |
+
"content": "_USR"
|
64 |
+
},
|
65 |
+
{
|
66 |
+
"type": "Replace",
|
67 |
+
"pattern": {
|
68 |
+
"Regex": "(http|ftp|https)://\\S+"
|
69 |
+
},
|
70 |
+
"content": "_URL"
|
71 |
+
},
|
72 |
+
{
|
73 |
+
"type": "Replace",
|
74 |
+
"pattern": {
|
75 |
+
"Regex": " j(a|e|i)[jaei]+"
|
76 |
+
},
|
77 |
+
"content": " jaja"
|
78 |
+
},
|
79 |
+
{
|
80 |
+
"type": "Replace",
|
81 |
+
"pattern": {
|
82 |
+
"Regex": " h(a|e|i)[haei]+"
|
83 |
+
},
|
84 |
+
"content": " jaja"
|
85 |
+
},
|
86 |
+
{
|
87 |
+
"type": "Replace",
|
88 |
+
"pattern": {
|
89 |
+
"String": "&"
|
90 |
+
},
|
91 |
+
"content": "&"
|
92 |
+
},
|
93 |
+
{
|
94 |
+
"type": "Replace",
|
95 |
+
"pattern": {
|
96 |
+
"String": ">"
|
97 |
+
},
|
98 |
+
"content": ">"
|
99 |
+
},
|
100 |
+
{
|
101 |
+
"type": "Replace",
|
102 |
+
"pattern": {
|
103 |
+
"String": "<"
|
104 |
+
},
|
105 |
+
"content": "<"
|
106 |
+
}
|
107 |
+
]
|
108 |
},
|
109 |
"pre_tokenizer": {
|
110 |
"type": "BertPreTokenizer"
|