Upload 4 files
Browse files- added_tokens.json +172 -0
- special_tokens_map.json +7 -0
- tokenizer_config.json +15 -0
- vocab.txt +0 -0
added_tokens.json
ADDED
@@ -0,0 +1,172 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"addmod": 30536,
|
3 |
+
"basefee": 30525,
|
4 |
+
"blockhash": 30585,
|
5 |
+
"bool": 30566,
|
6 |
+
"bytes1": 30552,
|
7 |
+
"bytes10": 30691,
|
8 |
+
"bytes11": 30545,
|
9 |
+
"bytes12": 30677,
|
10 |
+
"bytes13": 30639,
|
11 |
+
"bytes14": 30586,
|
12 |
+
"bytes15": 30542,
|
13 |
+
"bytes16": 30553,
|
14 |
+
"bytes17": 30609,
|
15 |
+
"bytes18": 30690,
|
16 |
+
"bytes19": 30526,
|
17 |
+
"bytes2": 30624,
|
18 |
+
"bytes20": 30631,
|
19 |
+
"bytes21": 30551,
|
20 |
+
"bytes22": 30572,
|
21 |
+
"bytes23": 30557,
|
22 |
+
"bytes24": 30588,
|
23 |
+
"bytes25": 30680,
|
24 |
+
"bytes26": 30686,
|
25 |
+
"bytes27": 30533,
|
26 |
+
"bytes28": 30665,
|
27 |
+
"bytes29": 30633,
|
28 |
+
"bytes3": 30637,
|
29 |
+
"bytes30": 30636,
|
30 |
+
"bytes31": 30587,
|
31 |
+
"bytes32": 30671,
|
32 |
+
"bytes4": 30598,
|
33 |
+
"bytes5": 30611,
|
34 |
+
"bytes6": 30664,
|
35 |
+
"bytes7": 30540,
|
36 |
+
"bytes8": 30576,
|
37 |
+
"bytes9": 30647,
|
38 |
+
"callcode": 30644,
|
39 |
+
"calldata": 30614,
|
40 |
+
"calldatacopy": 30555,
|
41 |
+
"calldataload": 30562,
|
42 |
+
"calldatasize": 30669,
|
43 |
+
"callvalue": 30615,
|
44 |
+
"chainid": 30537,
|
45 |
+
"coinbase": 30607,
|
46 |
+
"constructor": 30583,
|
47 |
+
"copyof": 30593,
|
48 |
+
"create2": 30597,
|
49 |
+
"delegatecall": 30685,
|
50 |
+
"delete": 30604,
|
51 |
+
"div": 30689,
|
52 |
+
"emit": 30653,
|
53 |
+
"enum": 30678,
|
54 |
+
"eq": 30596,
|
55 |
+
"exp": 30654,
|
56 |
+
"extcodecopy": 30640,
|
57 |
+
"extcodehash": 30663,
|
58 |
+
"extcodesize": 30656,
|
59 |
+
"fallback": 30617,
|
60 |
+
"gaslimit": 30621,
|
61 |
+
"gasprice": 30661,
|
62 |
+
"gwei": 30578,
|
63 |
+
"hex": 30523,
|
64 |
+
"immutable": 30655,
|
65 |
+
"int104": 30682,
|
66 |
+
"int112": 30613,
|
67 |
+
"int120": 30649,
|
68 |
+
"int128": 30595,
|
69 |
+
"int136": 30688,
|
70 |
+
"int144": 30538,
|
71 |
+
"int152": 30543,
|
72 |
+
"int16": 30676,
|
73 |
+
"int160": 30634,
|
74 |
+
"int168": 30623,
|
75 |
+
"int176": 30650,
|
76 |
+
"int184": 30556,
|
77 |
+
"int192": 30534,
|
78 |
+
"int200": 30546,
|
79 |
+
"int208": 30549,
|
80 |
+
"int216": 30584,
|
81 |
+
"int224": 30574,
|
82 |
+
"int232": 30619,
|
83 |
+
"int24": 30612,
|
84 |
+
"int240": 30675,
|
85 |
+
"int248": 30528,
|
86 |
+
"int256": 30558,
|
87 |
+
"int32": 30577,
|
88 |
+
"int40": 30573,
|
89 |
+
"int48": 30616,
|
90 |
+
"int56": 30629,
|
91 |
+
"int64": 30646,
|
92 |
+
"int72": 30610,
|
93 |
+
"int8": 30657,
|
94 |
+
"int80": 30628,
|
95 |
+
"int88": 30601,
|
96 |
+
"int96": 30565,
|
97 |
+
"iszero": 30550,
|
98 |
+
"keccak256": 30673,
|
99 |
+
"log0": 30548,
|
100 |
+
"log1": 30648,
|
101 |
+
"log2": 30651,
|
102 |
+
"log3": 30670,
|
103 |
+
"log4": 30539,
|
104 |
+
"mload": 30568,
|
105 |
+
"modifier": 30652,
|
106 |
+
"msg": 30641,
|
107 |
+
"msize": 30527,
|
108 |
+
"mstore": 30559,
|
109 |
+
"mstore8": 30571,
|
110 |
+
"mul": 30608,
|
111 |
+
"mulmod": 30643,
|
112 |
+
"mutable": 30532,
|
113 |
+
"override": 30589,
|
114 |
+
"payable": 30547,
|
115 |
+
"pragma": 30567,
|
116 |
+
"prevrandao": 30658,
|
117 |
+
"relocatable": 30580,
|
118 |
+
"returndatacopy": 30569,
|
119 |
+
"returndatasize": 30529,
|
120 |
+
"revert": 30599,
|
121 |
+
"sdiv": 30668,
|
122 |
+
"selfbalance": 30582,
|
123 |
+
"selfdestruct": 30575,
|
124 |
+
"shl": 30541,
|
125 |
+
"shr": 30630,
|
126 |
+
"signextend": 30524,
|
127 |
+
"sizeof": 30570,
|
128 |
+
"sload": 30581,
|
129 |
+
"slt": 30606,
|
130 |
+
"smod": 30554,
|
131 |
+
"sstore": 30679,
|
132 |
+
"staticcall": 30579,
|
133 |
+
"struct": 30683,
|
134 |
+
"timestamp": 30642,
|
135 |
+
"typedef": 30638,
|
136 |
+
"typeof": 30544,
|
137 |
+
"ufixed": 30600,
|
138 |
+
"uint": 30531,
|
139 |
+
"uint104": 30620,
|
140 |
+
"uint112": 30635,
|
141 |
+
"uint120": 30622,
|
142 |
+
"uint128": 30645,
|
143 |
+
"uint136": 30603,
|
144 |
+
"uint144": 30625,
|
145 |
+
"uint152": 30674,
|
146 |
+
"uint16": 30662,
|
147 |
+
"uint160": 30605,
|
148 |
+
"uint168": 30666,
|
149 |
+
"uint176": 30535,
|
150 |
+
"uint184": 30591,
|
151 |
+
"uint192": 30632,
|
152 |
+
"uint200": 30660,
|
153 |
+
"uint208": 30672,
|
154 |
+
"uint216": 30687,
|
155 |
+
"uint224": 30594,
|
156 |
+
"uint232": 30626,
|
157 |
+
"uint24": 30522,
|
158 |
+
"uint240": 30592,
|
159 |
+
"uint248": 30561,
|
160 |
+
"uint256": 30530,
|
161 |
+
"uint32": 30659,
|
162 |
+
"uint40": 30618,
|
163 |
+
"uint48": 30627,
|
164 |
+
"uint56": 30602,
|
165 |
+
"uint64": 30560,
|
166 |
+
"uint72": 30563,
|
167 |
+
"uint8": 30564,
|
168 |
+
"uint80": 30590,
|
169 |
+
"uint88": 30681,
|
170 |
+
"uint96": 30684,
|
171 |
+
"xor": 30667
|
172 |
+
}
|
special_tokens_map.json
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cls_token": "[CLS]",
|
3 |
+
"mask_token": "[MASK]",
|
4 |
+
"pad_token": "[PAD]",
|
5 |
+
"sep_token": "[SEP]",
|
6 |
+
"unk_token": "[UNK]"
|
7 |
+
}
|
tokenizer_config.json
ADDED
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"clean_up_tokenization_spaces": true,
|
3 |
+
"cls_token": "[CLS]",
|
4 |
+
"do_basic_tokenize": true,
|
5 |
+
"do_lower_case": true,
|
6 |
+
"mask_token": "[MASK]",
|
7 |
+
"model_max_length": 512,
|
8 |
+
"never_split": null,
|
9 |
+
"pad_token": "[PAD]",
|
10 |
+
"sep_token": "[SEP]",
|
11 |
+
"strip_accents": null,
|
12 |
+
"tokenize_chinese_chars": true,
|
13 |
+
"tokenizer_class": "BertTokenizer",
|
14 |
+
"unk_token": "[UNK]"
|
15 |
+
}
|
vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|