| { |
| "version": "1.0", |
| "truncation": null, |
| "padding": null, |
| "added_tokens": [ |
| { |
| "id": 0, |
| "content": "<pad>", |
| "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": "<bos>", |
| "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": 140, |
| "content": "[PAD]", |
| "single_word": false, |
| "lstrip": false, |
| "rstrip": false, |
| "normalized": false, |
| "special": true |
| }, |
| { |
| "id": 141, |
| "content": "[EOS]", |
| "single_word": false, |
| "lstrip": false, |
| "rstrip": false, |
| "normalized": false, |
| "special": true |
| }, |
| { |
| "id": 142, |
| "content": "[BOS]", |
| "single_word": false, |
| "lstrip": false, |
| "rstrip": false, |
| "normalized": false, |
| "special": true |
| }, |
| { |
| "id": 143, |
| "content": "[UNK]", |
| "single_word": false, |
| "lstrip": false, |
| "rstrip": false, |
| "normalized": false, |
| "special": true |
| }, |
| { |
| "id": 144, |
| "content": "[MASK]", |
| "single_word": false, |
| "lstrip": false, |
| "rstrip": false, |
| "normalized": false, |
| "special": true |
| }, |
| { |
| "id": 145, |
| "content": "[CLS]", |
| "single_word": false, |
| "lstrip": false, |
| "rstrip": false, |
| "normalized": false, |
| "special": true |
| }, |
| { |
| "id": 146, |
| "content": "[SEP]", |
| "single_word": false, |
| "lstrip": false, |
| "rstrip": false, |
| "normalized": false, |
| "special": true |
| }, |
| { |
| "id": 147, |
| "content": "[PROMPT]", |
| "single_word": false, |
| "lstrip": false, |
| "rstrip": false, |
| "normalized": false, |
| "special": true |
| }, |
| { |
| "id": 148, |
| "content": "[TIMESTAMP]", |
| "single_word": false, |
| "lstrip": false, |
| "rstrip": false, |
| "normalized": false, |
| "special": true |
| } |
| ], |
| "normalizer": null, |
| "pre_tokenizer": { |
| "type": "Sequence", |
| "pretokenizers": [ |
| { |
| "type": "Split", |
| "pattern": { |
| "Regex": "(\\[[^\\]]+]|<[^>]+>|Br?|Cl?|N|O|S|P|F|H|I|b|c|n|o|s|p|\\(|\\)|\\.|=|#|-|\\+|\\\\\\\\|\\u005C|/|:|~|@|\\?|\\*|\\$|\\%[0-9]{2}|[0-9])" |
| }, |
| "behavior": "Isolated", |
| "invert": false |
| }, |
| { |
| "type": "Split", |
| "pattern": { |
| "Regex": "\\[|\\]|0|1|2|3|4|5|6|7|8|9|-|\\+|H|@" |
| }, |
| "behavior": "Isolated", |
| "invert": false |
| } |
| ] |
| }, |
| "post_processor": { |
| "type": "TemplateProcessing", |
| "single": [ |
| { |
| "SpecialToken": { |
| "id": "[CLS]", |
| "type_id": 0 |
| } |
| }, |
| { |
| "SpecialToken": { |
| "id": "[PROMPT]", |
| "type_id": 0 |
| } |
| }, |
| { |
| "SpecialToken": { |
| "id": "[TIMESTAMP]", |
| "type_id": 0 |
| } |
| }, |
| { |
| "SpecialToken": { |
| "id": "[BOS]", |
| "type_id": 0 |
| } |
| }, |
| { |
| "Sequence": { |
| "id": "A", |
| "type_id": 0 |
| } |
| }, |
| { |
| "SpecialToken": { |
| "id": "[EOS]", |
| "type_id": 0 |
| } |
| } |
| ], |
| "pair": [ |
| { |
| "SpecialToken": { |
| "id": "[CLS]", |
| "type_id": 0 |
| } |
| }, |
| { |
| "SpecialToken": { |
| "id": "[PROMPT]", |
| "type_id": 0 |
| } |
| }, |
| { |
| "SpecialToken": { |
| "id": "[TIMESTAMP]", |
| "type_id": 0 |
| } |
| }, |
| { |
| "SpecialToken": { |
| "id": "[BOS]", |
| "type_id": 0 |
| } |
| }, |
| { |
| "Sequence": { |
| "id": "A", |
| "type_id": 0 |
| } |
| }, |
| { |
| "SpecialToken": { |
| "id": "[SEP]", |
| "type_id": 0 |
| } |
| }, |
| { |
| "Sequence": { |
| "id": "B", |
| "type_id": 0 |
| } |
| }, |
| { |
| "SpecialToken": { |
| "id": "[EOS]", |
| "type_id": 0 |
| } |
| } |
| ], |
| "special_tokens": { |
| "[BOS]": { |
| "id": "[BOS]", |
| "ids": [ |
| 142 |
| ], |
| "tokens": [ |
| "[BOS]" |
| ] |
| }, |
| "[CLS]": { |
| "id": "[CLS]", |
| "ids": [ |
| 145 |
| ], |
| "tokens": [ |
| "[CLS]" |
| ] |
| }, |
| "[EOS]": { |
| "id": "[EOS]", |
| "ids": [ |
| 141 |
| ], |
| "tokens": [ |
| "[EOS]" |
| ] |
| }, |
| "[PROMPT]": { |
| "id": "[PROMPT]", |
| "ids": [ |
| 147 |
| ], |
| "tokens": [ |
| "[PROMPT]" |
| ] |
| }, |
| "[SEP]": { |
| "id": "[SEP]", |
| "ids": [ |
| 146 |
| ], |
| "tokens": [ |
| "[SEP]" |
| ] |
| }, |
| "[TIMESTAMP]": { |
| "id": "[TIMESTAMP]", |
| "ids": [ |
| 148 |
| ], |
| "tokens": [ |
| "[TIMESTAMP]" |
| ] |
| } |
| } |
| }, |
| "decoder": { |
| "type": "BPEDecoder", |
| "suffix": "</w>" |
| }, |
| "model": { |
| "type": "WordLevel", |
| "vocab": { |
| "<pad>": 0, |
| "<eos>": 1, |
| "<bos>": 2, |
| "<unk>": 3, |
| "<mask>": 4, |
| "C": 5, |
| "=": 6, |
| "(": 7, |
| ")": 8, |
| "O": 9, |
| "N": 10, |
| "1": 11, |
| "2": 12, |
| "3": 13, |
| "4": 14, |
| "F": 15, |
| "S": 16, |
| "<BASE>": 17, |
| "<SYST>": 18, |
| "<TRAD>": 19, |
| "5": 20, |
| "Cl": 21, |
| "[": 22, |
| "]": 23, |
| ".": 24, |
| "6": 25, |
| "7": 26, |
| "-": 27, |
| "+": 28, |
| "#": 29, |
| "Br": 30, |
| "8": 31, |
| "9": 32, |
| "P": 33, |
| "H": 34, |
| "I": 35, |
| "Si": 36, |
| "B": 37, |
| "Na": 38, |
| "Y": 39, |
| "Ir": 40, |
| "Pt": 41, |
| "K": 42, |
| "Se": 43, |
| "Li": 44, |
| "W": 45, |
| "Sn": 46, |
| "V": 47, |
| "Zr": 48, |
| "Zn": 49, |
| "Cu": 50, |
| "Fe": 51, |
| "%": 52, |
| "Ti": 53, |
| "Pd": 54, |
| "Co": 55, |
| "Mg": 56, |
| "Al": 57, |
| "Ni": 58, |
| "Ge": 59, |
| "Ru": 60, |
| "Ca": 61, |
| "U": 62, |
| "Mn": 63, |
| "Cr": 64, |
| "Au": 65, |
| "Ag": 66, |
| "As": 67, |
| "Te": 68, |
| "Mo": 69, |
| "Ac": 70, |
| "0": 71, |
| "Tb": 72, |
| "f": 73, |
| "Rh": 74, |
| "g": 75, |
| "Cs": 76, |
| "Rf": 77, |
| "Ar": 78, |
| "Sb": 79, |
| "Rb": 80, |
| "Ba": 81, |
| "Os": 82, |
| "Re": 83, |
| "Gd": 84, |
| "Cd": 85, |
| "Bi": 86, |
| "Pb": 87, |
| "In": 88, |
| "Ga": 89, |
| "Ce": 90, |
| "La": 91, |
| "Eu": 92, |
| "Tl": 93, |
| "Tc": 94, |
| "Nb": 95, |
| "Sr": 96, |
| "Ta": 97, |
| "Nd": 98, |
| "Pr": 99, |
| "Yb": 100, |
| "Sm": 101, |
| "Be": 102, |
| "Sc": 103, |
| "Dy": 104, |
| "Lu": 105, |
| "Fm": 106, |
| "Er": 107, |
| "Th": 108, |
| "o": 109, |
| "Tm": 110, |
| "At": 111, |
| "No": 112, |
| "Po": 113, |
| "Cm": 114, |
| "Sg": 115, |
| "Xe": 116, |
| "Np": 117, |
| "Lr": 118, |
| "Pu": 119, |
| "Pm": 120, |
| "Cf": 121, |
| "e": 122, |
| "Am": 123, |
| "Es": 124, |
| "Pa": 125, |
| "Ne": 126, |
| "Bk": 127, |
| "Db": 128, |
| "Mt": 129, |
| "Kr": 130, |
| "Rn": 131, |
| "s": 132, |
| "Bh": 133, |
| "Md": 134, |
| "Ra": 135, |
| "Fr": 136, |
| "@": 137, |
| "\\": 138, |
| "/": 139 |
| }, |
| "unk_token": "<unk>" |
| } |
| } |