Upload pt-t.json
Browse files
pt-t.json
ADDED
@@ -0,0 +1,156 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"version": "1.0",
|
3 |
+
"truncation": null,
|
4 |
+
"padding": null,
|
5 |
+
"added_tokens": [
|
6 |
+
{
|
7 |
+
"id": 0,
|
8 |
+
"content": "[STOP]",
|
9 |
+
"single_word": false,
|
10 |
+
"lstrip": false,
|
11 |
+
"rstrip": false,
|
12 |
+
"normalized": false,
|
13 |
+
"special": true
|
14 |
+
},
|
15 |
+
{
|
16 |
+
"id": 1,
|
17 |
+
"content": "[UNK]",
|
18 |
+
"single_word": false,
|
19 |
+
"lstrip": false,
|
20 |
+
"rstrip": false,
|
21 |
+
"normalized": false,
|
22 |
+
"special": true
|
23 |
+
},
|
24 |
+
{
|
25 |
+
"id": 2,
|
26 |
+
"content": "[SPACE]",
|
27 |
+
"single_word": false,
|
28 |
+
"lstrip": false,
|
29 |
+
"rstrip": false,
|
30 |
+
"normalized": false,
|
31 |
+
"special": true
|
32 |
+
}
|
33 |
+
],
|
34 |
+
"normalizer": null,
|
35 |
+
"pre_tokenizer": {
|
36 |
+
"type": "Whitespace"
|
37 |
+
},
|
38 |
+
"post_processor": null,
|
39 |
+
"decoder": null,
|
40 |
+
"model": {
|
41 |
+
"type": "BPE",
|
42 |
+
"dropout": null,
|
43 |
+
"unk_token": "[UNK]",
|
44 |
+
"continuing_subword_prefix": null,
|
45 |
+
"end_of_word_suffix": null,
|
46 |
+
"fuse_unk": false,
|
47 |
+
"byte_fallback": false,
|
48 |
+
"vocab": {
|
49 |
+
"[STOP]": 0,
|
50 |
+
"[UNK]": 1,
|
51 |
+
"[SPACE]": 2,
|
52 |
+
"!":3,
|
53 |
+
"'":4,
|
54 |
+
"(":5,
|
55 |
+
")":6,
|
56 |
+
",":7,
|
57 |
+
"-":8,
|
58 |
+
".":9,
|
59 |
+
"/":10,
|
60 |
+
":":11,
|
61 |
+
";":12,
|
62 |
+
"?":13,
|
63 |
+
"a":14,
|
64 |
+
"b":15,
|
65 |
+
"c":16,
|
66 |
+
"d":17,
|
67 |
+
"e":18,
|
68 |
+
"f":19,
|
69 |
+
"g":20,
|
70 |
+
"h":21,
|
71 |
+
"i":22,
|
72 |
+
"j":23,
|
73 |
+
"k":24,
|
74 |
+
"l":25,
|
75 |
+
"m":26,
|
76 |
+
"n":27,
|
77 |
+
"o":28,
|
78 |
+
"p":29,
|
79 |
+
"q":30,
|
80 |
+
"r":31,
|
81 |
+
"s":32,
|
82 |
+
"t":33,
|
83 |
+
"u":34,
|
84 |
+
"v":35,
|
85 |
+
"w":36,
|
86 |
+
"x":37,
|
87 |
+
"y":38,
|
88 |
+
"z":39,
|
89 |
+
"¿": 40,
|
90 |
+
"À": 41,
|
91 |
+
"Á": 42,
|
92 |
+
"Â": 43,
|
93 |
+
"Ã": 44,
|
94 |
+
"Ä": 45,
|
95 |
+
"Å": 46,
|
96 |
+
"Æ": 47,
|
97 |
+
"Ç": 48,
|
98 |
+
"È": 49,
|
99 |
+
"É": 50,
|
100 |
+
"Ê": 51,
|
101 |
+
"Ë": 52,
|
102 |
+
"Ì": 53,
|
103 |
+
"Í": 54,
|
104 |
+
"Î": 55,
|
105 |
+
"Ï": 56,
|
106 |
+
"Ð": 57,
|
107 |
+
"Ñ": 58,
|
108 |
+
"Ò": 59,
|
109 |
+
"Ó": 60,
|
110 |
+
"Ô": 61,
|
111 |
+
"Õ": 62,
|
112 |
+
"Ö": 63,
|
113 |
+
"Ø": 64,
|
114 |
+
"Ù": 65,
|
115 |
+
"Ú": 66,
|
116 |
+
"Û": 67,
|
117 |
+
"Ü": 68,
|
118 |
+
"Ý": 69,
|
119 |
+
"Þ": 70,
|
120 |
+
"ß": 71,
|
121 |
+
"à": 72,
|
122 |
+
"á": 73,
|
123 |
+
"â": 74,
|
124 |
+
"ã": 75,
|
125 |
+
"ä": 76,
|
126 |
+
"å": 77,
|
127 |
+
"æ": 78,
|
128 |
+
"ç": 79,
|
129 |
+
"è": 80,
|
130 |
+
"é": 81,
|
131 |
+
"ê": 82,
|
132 |
+
"ë": 83,
|
133 |
+
"ì": 84,
|
134 |
+
"í": 85,
|
135 |
+
"î": 86,
|
136 |
+
"ï": 87,
|
137 |
+
"ð": 88,
|
138 |
+
"ñ": 89,
|
139 |
+
"ò": 90,
|
140 |
+
"ó": 91,
|
141 |
+
"ô": 92,
|
142 |
+
"õ": 93,
|
143 |
+
"ö": 94,
|
144 |
+
"ø": 95,
|
145 |
+
"ù": 96,
|
146 |
+
"ú": 97,
|
147 |
+
"û": 98,
|
148 |
+
"ü": 99,
|
149 |
+
"ý": 100,
|
150 |
+
"þ": 101,
|
151 |
+
"ÿ": 102,
|
152 |
+
"¡": 103
|
153 |
+
},
|
154 |
+
"merges": []
|
155 |
+
}
|
156 |
+
}
|