REGEX-1B / tokenizer.json
yzimmermann's picture
Upload tokenizer
465e89b verified
{
"version": "1.0",
"truncation": null,
"padding": null,
"added_tokens": [
{
"id": 0,
"content": "[BOS]",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": false,
"special": true
},
{
"id": 1,
"content": "[EOS]",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": false,
"special": true
},
{
"id": 2,
"content": "[PAD]",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": false,
"special": true
},
{
"id": 3,
"content": "[UNK]",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": false,
"special": true
},
{
"id": 4,
"content": "[MASK]",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": false,
"special": true
},
{
"id": 5,
"content": "[CLS]",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": false,
"special": true
},
{
"id": 6,
"content": "[SEP]",
"single_word": false,
"lstrip": false,
"rstrip": false,
"normalized": false,
"special": true
}
],
"normalizer": null,
"pre_tokenizer": {
"type": "Split",
"pattern": {
"Regex": "(\\[[^\\]]+]|Br?|Cl?|N|O|S|P|F|I|b|c|n|o|s|p|\\(|\\)|\\.|=|-|\\+|\\\\|\\/|:|~|@|\\?|>>?|\\*|\\$|\\%[0-9]{2}|[0-9])"
},
"behavior": "Isolated",
"invert": false
},
"post_processor": null,
"decoder": null,
"model": {
"type": "WordLevel",
"vocab": {
"[BOS]": 0,
"[EOS]": 1,
"[PAD]": 2,
"[UNK]": 3,
"[MASK]": 4,
"[CLS]": 5,
"[SEP]": 6,
"C": 7,
"c": 8,
"(": 9,
")": 10,
"O": 11,
"1": 12,
"2": 13,
"=": 14,
"N": 15,
"3": 16,
"n": 17,
"[C@H]": 18,
"[C@@H]": 19,
"4": 20,
"F": 21,
"[NH+]": 22,
"S": 23,
"o": 24,
"Cl": 25,
"s": 26,
"[nH]": 27,
"5": 28,
"[NH2+]": 29,
"#": 30,
"/": 31,
"Br": 32,
"[C@@]": 33,
"[C@]": 34,
"[O-]": 35,
"\\": 36,
"[nH+]": 37,
"[NH3+]": 38,
"[n-]": 39,
"6": 40,
"I": 41,
"[N-]": 42,
"-": 43,
"7": 44,
"[H]": 45,
"[Si]": 46,
"P": 47,
"[n+]": 48,
"[S-]": 49,
"[S@]": 50,
"[S@@]": 51,
"[N+]": 52,
"8": 53,
"B": 54,
"[CH]": 55,
"[C]": 56,
"9": 57,
"[P@@]": 58,
"[P@]": 59,
"[S+]": 60,
"[N@@+]": 61,
"[N@+]": 62,
"[CH2]": 63,
"[O]": 64,
"[s+]": 65,
"[Sn]": 66,
"[P+]": 67,
"[B-]": 68,
"[S@@+]": 69,
"[S@+]": 70,
"p": 71,
"[N]": 72,
"%10": 73,
"[C+]": 74,
"[o+]": 75,
"%11": 76,
"[N@]": 77,
"[P@@H]": 78,
"[n@]": 79,
"[C-]": 80,
"[c+]": 81,
"[IH2]": 82,
"%13": 83,
"[Si@@]": 84,
"%12": 85,
"[Si@]": 86,
"[N@@]": 87,
"[BH3-]": 88,
"[P@H]": 89,
"[CH-]": 90,
"[Sn@]": 91,
"[s@@]": 92,
"[s@]": 93,
"[P@+]": 94,
"[P@@+]": 95,
"[Sn@@]": 96,
"[c-]": 97,
"[17O]": 98,
"[BH-]": 99,
"[SnH4+2]": 100,
"[B@-]": 101,
"[B@@-]": 102,
"[cH-]": 103,
"[O+]": 104,
"[SnH2+]": 105,
"[SnH]": 106,
"%14": 107,
"[Sn+2]": 108,
"[I+]": 109,
"[P@@H+]": 110,
"%15": 111,
"%16": 112,
"%18": 113,
"[Br+]": 114,
"[NH]": 115,
"[Sn+]": 116,
"[n@@]": 117,
"%17": 118,
"%19": 119,
"%20": 120,
"%21": 121,
"%22": 122,
"[18OH]": 123,
"[BH2-]": 124,
"[S@@-]": 125,
"[S@@H]": 126,
"[Sn+3]": 127,
"[SnH2]": 128,
"[SnH6+3]": 129,
"[pH]": 130,
"[S@H]": 131,
"[SH3]": 132,
"[SiH2]": 133,
"[SiH3]": 134,
"[Sn-]": 135,
"[p+]": 136
},
"unk_token": "[UNK]"
}
}