Upload tokenizer
Browse files- added_tokens.json +2 -3
- special_tokens_map.json +0 -7
- vocab.json +3 -4
added_tokens.json
CHANGED
@@ -1,5 +1,4 @@
|
|
1 |
{
|
2 |
-
"</s>":
|
3 |
-
"<s>":
|
4 |
-
"[PAD]": 147
|
5 |
}
|
|
|
1 |
{
|
2 |
+
"</s>": 145,
|
3 |
+
"<s>": 144
|
|
|
4 |
}
|
special_tokens_map.json
CHANGED
@@ -13,13 +13,6 @@
|
|
13 |
"normalized": true,
|
14 |
"rstrip": false,
|
15 |
"single_word": false
|
16 |
-
},
|
17 |
-
{
|
18 |
-
"content": "[PAD]",
|
19 |
-
"lstrip": false,
|
20 |
-
"normalized": true,
|
21 |
-
"rstrip": false,
|
22 |
-
"single_word": false
|
23 |
}
|
24 |
],
|
25 |
"bos_token": "<s>",
|
|
|
13 |
"normalized": true,
|
14 |
"rstrip": false,
|
15 |
"single_word": false
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
16 |
}
|
17 |
],
|
18 |
"bos_token": "<s>",
|
vocab.json
CHANGED
@@ -1,5 +1,4 @@
|
|
1 |
{
|
2 |
-
"!": 1,
|
3 |
"#": 2,
|
4 |
"&": 3,
|
5 |
"'": 4,
|
@@ -24,6 +23,8 @@
|
|
24 |
">": 23,
|
25 |
"?": 24,
|
26 |
"@": 25,
|
|
|
|
|
27 |
"_": 26,
|
28 |
"a": 27,
|
29 |
"b": 28,
|
@@ -136,12 +137,10 @@
|
|
136 |
"ỵ": 134,
|
137 |
"ỷ": 135,
|
138 |
"ỹ": 136,
|
139 |
-
"": 137,
|
140 |
"–": 138,
|
141 |
"‘": 139,
|
142 |
"’": 140,
|
143 |
"“": 141,
|
144 |
"”": 142,
|
145 |
-
"…": 143
|
146 |
-
"": 144
|
147 |
}
|
|
|
1 |
{
|
|
|
2 |
"#": 2,
|
3 |
"&": 3,
|
4 |
"'": 4,
|
|
|
23 |
">": 23,
|
24 |
"?": 24,
|
25 |
"@": 25,
|
26 |
+
"[PAD]": 146,
|
27 |
+
"[UNK]": 145,
|
28 |
"_": 26,
|
29 |
"a": 27,
|
30 |
"b": 28,
|
|
|
137 |
"ỵ": 134,
|
138 |
"ỷ": 135,
|
139 |
"ỹ": 136,
|
|
|
140 |
"–": 138,
|
141 |
"‘": 139,
|
142 |
"’": 140,
|
143 |
"“": 141,
|
144 |
"”": 142,
|
145 |
+
"…": 143
|
|
|
146 |
}
|