Upload 3 files
Browse files- special_tokens_map.json +4 -28
- tokenizer.json +8 -8
- tokenizer_config.json +5 -5
special_tokens_map.json
CHANGED
@@ -1,30 +1,6 @@
|
|
1 |
{
|
2 |
-
"bos_token":
|
3 |
-
|
4 |
-
|
5 |
-
|
6 |
-
"rstrip": false,
|
7 |
-
"single_word": false
|
8 |
-
},
|
9 |
-
"eos_token": {
|
10 |
-
"content": "<|eos|>",
|
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": "<|unknown|>",
|
25 |
-
"lstrip": false,
|
26 |
-
"normalized": false,
|
27 |
-
"rstrip": false,
|
28 |
-
"single_word": false
|
29 |
-
}
|
30 |
}
|
|
|
1 |
{
|
2 |
+
"bos_token": "<|bos|>",
|
3 |
+
"eos_token": "<|eos|>",
|
4 |
+
"pad_token": "<|pad|>",
|
5 |
+
"unk_token": "<|unknown|>"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
6 |
}
|
tokenizer.json
CHANGED
@@ -122,7 +122,7 @@
|
|
122 |
},
|
123 |
{
|
124 |
"id": 13,
|
125 |
-
"content": "<|
|
126 |
"single_word": false,
|
127 |
"lstrip": false,
|
128 |
"rstrip": false,
|
@@ -131,7 +131,7 @@
|
|
131 |
},
|
132 |
{
|
133 |
"id": 14,
|
134 |
-
"content": "<|
|
135 |
"single_word": false,
|
136 |
"lstrip": false,
|
137 |
"rstrip": false,
|
@@ -140,7 +140,7 @@
|
|
140 |
},
|
141 |
{
|
142 |
"id": 15,
|
143 |
-
"content": "<|
|
144 |
"single_word": false,
|
145 |
"lstrip": false,
|
146 |
"rstrip": false,
|
@@ -149,7 +149,7 @@
|
|
149 |
},
|
150 |
{
|
151 |
"id": 16,
|
152 |
-
"content": "<|
|
153 |
"single_word": false,
|
154 |
"lstrip": false,
|
155 |
"rstrip": false,
|
@@ -429,10 +429,10 @@
|
|
429 |
"<general>": 10,
|
430 |
"</general>": 11,
|
431 |
"<|input_end|>": 12,
|
432 |
-
"<|
|
433 |
-
"<|
|
434 |
-
"<|
|
435 |
-
"<|
|
436 |
"<|reserved_5|>": 17,
|
437 |
"<|reserved_6|>": 18,
|
438 |
"<|reserved_7|>": 19,
|
|
|
122 |
},
|
123 |
{
|
124 |
"id": 13,
|
125 |
+
"content": "<|very_vague|>",
|
126 |
"single_word": false,
|
127 |
"lstrip": false,
|
128 |
"rstrip": false,
|
|
|
131 |
},
|
132 |
{
|
133 |
"id": 14,
|
134 |
+
"content": "<|vague|>",
|
135 |
"single_word": false,
|
136 |
"lstrip": false,
|
137 |
"rstrip": false,
|
|
|
140 |
},
|
141 |
{
|
142 |
"id": 15,
|
143 |
+
"content": "<|detailed|>",
|
144 |
"single_word": false,
|
145 |
"lstrip": false,
|
146 |
"rstrip": false,
|
|
|
149 |
},
|
150 |
{
|
151 |
"id": 16,
|
152 |
+
"content": "<|very_detailed|>",
|
153 |
"single_word": false,
|
154 |
"lstrip": false,
|
155 |
"rstrip": false,
|
|
|
429 |
"<general>": 10,
|
430 |
"</general>": 11,
|
431 |
"<|input_end|>": 12,
|
432 |
+
"<|very_vague|>": 13,
|
433 |
+
"<|vague|>": 14,
|
434 |
+
"<|detailed|>": 15,
|
435 |
+
"<|very_detailed>": 16,
|
436 |
"<|reserved_5|>": 17,
|
437 |
"<|reserved_6|>": 18,
|
438 |
"<|reserved_7|>": 19,
|
tokenizer_config.json
CHANGED
@@ -1,4 +1,5 @@
|
|
1 |
{
|
|
|
2 |
"added_tokens_decoder": {
|
3 |
"0": {
|
4 |
"content": "<|bos|>",
|
@@ -105,7 +106,7 @@
|
|
105 |
"special": true
|
106 |
},
|
107 |
"13": {
|
108 |
-
"content": "<|
|
109 |
"lstrip": false,
|
110 |
"normalized": false,
|
111 |
"rstrip": false,
|
@@ -113,7 +114,7 @@
|
|
113 |
"special": true
|
114 |
},
|
115 |
"14": {
|
116 |
-
"content": "<|
|
117 |
"lstrip": false,
|
118 |
"normalized": false,
|
119 |
"rstrip": false,
|
@@ -121,7 +122,7 @@
|
|
121 |
"special": true
|
122 |
},
|
123 |
"15": {
|
124 |
-
"content": "<|
|
125 |
"lstrip": false,
|
126 |
"normalized": false,
|
127 |
"rstrip": false,
|
@@ -129,7 +130,7 @@
|
|
129 |
"special": true
|
130 |
},
|
131 |
"16": {
|
132 |
-
"content": "<|
|
133 |
"lstrip": false,
|
134 |
"normalized": false,
|
135 |
"rstrip": false,
|
@@ -362,6 +363,5 @@
|
|
362 |
"pad_token": "<|pad|>",
|
363 |
"pad_token_type_id": 0,
|
364 |
"padding_side": "right",
|
365 |
-
"tokenizer_class": "PreTrainedTokenizerFast",
|
366 |
"unk_token": "<|unknown|>"
|
367 |
}
|
|
|
1 |
{
|
2 |
+
"tokenizer_class": "PreTrainedTokenizerFast",
|
3 |
"added_tokens_decoder": {
|
4 |
"0": {
|
5 |
"content": "<|bos|>",
|
|
|
106 |
"special": true
|
107 |
},
|
108 |
"13": {
|
109 |
+
"content": "<|very_vague|>",
|
110 |
"lstrip": false,
|
111 |
"normalized": false,
|
112 |
"rstrip": false,
|
|
|
114 |
"special": true
|
115 |
},
|
116 |
"14": {
|
117 |
+
"content": "<|vague|>",
|
118 |
"lstrip": false,
|
119 |
"normalized": false,
|
120 |
"rstrip": false,
|
|
|
122 |
"special": true
|
123 |
},
|
124 |
"15": {
|
125 |
+
"content": "<|detailed|>",
|
126 |
"lstrip": false,
|
127 |
"normalized": false,
|
128 |
"rstrip": false,
|
|
|
130 |
"special": true
|
131 |
},
|
132 |
"16": {
|
133 |
+
"content": "<|very_detailed|>",
|
134 |
"lstrip": false,
|
135 |
"normalized": false,
|
136 |
"rstrip": false,
|
|
|
363 |
"pad_token": "<|pad|>",
|
364 |
"pad_token_type_id": 0,
|
365 |
"padding_side": "right",
|
|
|
366 |
"unk_token": "<|unknown|>"
|
367 |
}
|