{ "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": "[CLS]", "single_word": false, "lstrip": false, "rstrip": false, "normalized": false, "special": true }, { "id": 2, "content": "[SEP]", "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 } ], "normalizer": { "type": "Sequence", "normalizers": [ { "type": "NFD" }, { "type": "Lowercase" }, { "type": "StripAccents" } ] }, "pre_tokenizer": { "type": "Whitespace" }, "post_processor": { "type": "TemplateProcessing", "single": [ { "SpecialToken": { "id": "[CLS]", "type_id": 0 } }, { "Sequence": { "id": "A", "type_id": 0 } }, { "SpecialToken": { "id": "[SEP]", "type_id": 0 } } ], "pair": [ { "SpecialToken": { "id": "[CLS]", "type_id": 0 } }, { "Sequence": { "id": "A", "type_id": 0 } }, { "SpecialToken": { "id": "[SEP]", "type_id": 0 } }, { "Sequence": { "id": "B", "type_id": 1 } }, { "SpecialToken": { "id": "[SEP]", "type_id": 1 } } ], "special_tokens": { "[CLS]": { "id": "[CLS]", "ids": [ 1 ], "tokens": [ "[CLS]" ] }, "[SEP]": { "id": "[SEP]", "ids": [ 2 ], "tokens": [ "[SEP]" ] } } }, "decoder": null, "model": { "type": "WordPiece", "unk_token": "[UNK]", "continuing_subword_prefix": "##", "max_input_chars_per_word": 100, "vocab": { "[PAD]": 0, "[CLS]": 1, "[SEP]": 2, "[UNK]": 3, "[MASK]": 4, "\"": 5, "%": 6, "&": 7, "'": 8, "(": 9, ")": 10, "*": 11, "+": 12, ",": 13, "-": 14, ".": 15, "/": 16, "0": 17, "1": 18, "2": 19, "3": 20, "4": 21, "5": 22, "6": 23, "7": 24, "8": 25, "9": 26, ":": 27, ";": 28, "<": 29, "=": 30, ">": 31, "[": 32, "]": 33, "_": 34, "a": 35, "b": 36, "c": 37, "d": 38, "e": 39, "f": 40, "g": 41, "h": 42, "i": 43, "j": 44, "k": 45, "l": 46, "m": 47, "n": 48, "o": 49, "p": 50, "q": 51, "r": 52, "s": 53, "t": 54, "u": 55, "v": 56, "w": 57, "x": 58, "y": 59, "z": 60, "~": 61, "®": 62, "°": 63, "²": 64, "³": 65, "µ": 66, "ø": 67, "α": 68, "–": 69, "##u": 70, "##l": 71, "##i": 72, "##p": 73, "##w": 74, "##o": 75, "##d": 76, "##e": 77, "##s": 78, "##g": 79, "##n": 80, "##t": 81, "##c": 82, "##m": 83, "##h": 84, "##r": 85, "##v": 86, "##0": 87, "##a": 88, "##y": 89, "##7": 90, "##2": 91, "##1": 92, "##4": 93, "##b": 94, "##x": 95, "##6": 96, "##3": 97, "##5": 98, "##f": 99, "##j": 100, "##k": 101, "##;": 102, "##z": 103, "##µ": 104, "##8": 105, "##9": 106, "##%": 107, "##(": 108, "##q": 109, "##]": 110, "##)": 111, "##,": 112, "##_": 113, "##\"": 114, "##>": 115, "##/": 116, "##ø": 117, "##<": 118, "##=": 119, "##-": 120, "##er": 121, "##in": 122, "##on": 123, "##at": 124, "##ed": 125, "##ro": 126, "##en": 127, "##an": 128, "##ic": 129, "##as": 130, "##ar": 131, "##or": 132, "##al": 133, "##ct": 134, "##it": 135, "##ol": 136, "##il": 137, "##ion": 138, "##ing": 139, "##es": 140, "##et": 141, "##el": 142, "##id": 143, "in": 144, "##le": 145, "tr": 146, "##um": 147, "##yl": 148, "en": 149, "##ent": 150, "##te": 151, "##re": 152, "##ation": 153, "co": 154, "##st": 155, "##od": 156, "pl": 157, "##am": 158, "##rom": 159, "con": 160, "##ity": 161, "##80": 162, "##ric": 163, "##ul": 164, "di": 165, "from": 166, "##ene": 167, "##ig": 168, "##580": 169, "##5804": 170, "pro": 171, "ele": 172, "##ut": 173, "##ck": 174, "re": 175, "elect": 176, "##lo": 177, "##ur": 178, "##ate": 179, "a1": 180, "a3": 181, "##ad": 182, "al": 183, "##ix": 184, "##ated": 185, "pa": 186, "##igh": 187, "##ine": 188, "##du": 189, "##oly": 190, "mix": 191, "##ac": 192, "car": 193, "ste": 194, "at": 195, "##ood": 196, "st": 197, "electric": 198, "##ow": 199, "##im": 200, "##sp": 201, "##em": 202, "##ium": 203, "20": 204, "gr": 205, "poly": 206, "##ard": 207, "##eth": 208, "##ap": 209, "##ill": 210, "mm": 211, "electricity": 212, "##uck": 213, "truck": 214, "##ip": 215, "##ber": 216, "##ch": 217, "##duct": 218, "##15804": 219, "en15804": 220, "cont": 221, "##pp": 222, "##ort": 223, "##ide": 224, "##ight": 225, "plan": 226, "sh": 227, "of": 228, "##nd": 229, "##ast": 230, "product": 231, "##oad": 232, "##eat": 233, "##ge": 234, "##wood": 235, "##ased": 236, "pow": 237, "steel": 238, "##ver": 239, "##ess": 240, "ac": 241, "##ater": 242, "##00": 243, "##and": 244, "content": 245, "par": 246, "##ri": 247, "com": 248, "ro": 249, "was": 250, "##yload": 251, "payload": 252, "##th": 253, "ch": 254, "##ist": 255, "##ter": 256, "##of": 257, "fi": 258, "gas": 259, "##ass": 260, "waste": 261, "cap": 262, "##ylene": 263, "gro": 264, "##ith": 265, "##ane": 266, "power": 267, "tran": 268, "##cl": 269, "15804": 270, "app": 271, "based": 272, "##is": 273, "capac": 274, "##ure": 275, "th": 276, "and": 277, "water": 278, "##om": 279, "plant": 280, "for": 281, "un": 282, "##led": 283, "grid": 284, "##us": 285, "##eet": 286, "part": 287, "##yd": 288, "##2o": 289, "##est": 290, "##old": 291, "by": 292, "##co": 293, "oil": 294, "##eight": 295, "##ce": 296, "bo": 297, "##rox": 298, "res": 299, "##gy": 300, "##ss": 301, "##aw": 302, "sheet": 303, "h2o": 304, "us": 305, "##ly": 306, "with": 307, "##esel": 308, "transp": 309, "capacity": 310, "mo": 311, "##iner": 312, "transport": 313, "##uro": 314, "weight": 315, "##iz": 316, "),": 317, "##bon": 318, "bi": 319, "on": 320, "##ergy": 321, "diesel": 322, "energy": 323, "euro": 324, "##un": 325, "##lu": 326, "##rect": 327, "##age": 328, "approx": 329, "se": 330, "gross": 331, "carbon": 332, "cr": 333, "##imation": 334, "approximation": 335, "##cy": 336, "##ment": 337, "##fill": 338, "to": 339, "12": 340, "acid": 341, "land": 342, "##umin": 343, "op": 344, "##ug": 345, "##av": 346, "landfill": 347, "##ial": 348, "alumin": 349, "##mm": 350, "##put": 351, "##ural": 352, "allo": 353, "##cess": 354, "fiber": 355, "eth": 356, "powered": 357, "process": 358, "production": 359, "ex": 360, "vd": 361, "##act": 362, "moist": 363, "alloc": 364, "##iv": 365, "##cr": 366, "##ain": 367, "##lor": 368, "##ab": 369, "de": 370, "##mg": 371, "hyd": 372, "kw": 373, "pr": 374, "##ile": 375, "##one": 376, "moisture": 377, "hard": 378, "##amine": 379, "rol": 380, "open": 381, "##os": 382, "##all": 383, "te": 384, "##ne": 385, "##iler": 386, "##con": 387, "##for": 388, "kg": 389, "##erm": 390, "dri": 391, "input": 392, "comp": 393, "cf": 394, "so": 395, "cem": 396, "##ph": 397, "treat": 398, "treatment": 399, "ag": 400, "lc": 401, "##der": 402, "allocation": 403, "rolled": 404, "flo": 405, "mun": 406, "##icip": 407, "##ctor": 408, "plast": 409, "roof": 410, "municip": 411, "municipal": 412, "10": 413, "cold": 414, "##pe": 415, "##ay": 416, "##verage": 417, "average": 418, "##cycl": 419, "recycl": 420, "dried": 421, "meth": 422, "wood": 423, "ep": 424, "nat": 425, "##br": 426, "##og": 427, "60": 428, "##ance": 429, "201": 430, "natural": 431, "steam": 432, "aluminium": 433, "##ite": 434, "therm": 435, "high": 436, "wh": 437, "##di": 438, "##sion": 439, "reg": 440, "##ot": 441, "14": 442, "##uel": 443, "##bo": 444, "pan": 445, "ply": 446, "fuel": 447, "hydro": 448, "ha": 449, "mod": 450, "sol": 451, "##ub": 452, "##ust": 453, "##lass": 454, "##ir": 455, "conne": 456, "connector": 457, "c3": 458, "lum": 459, "inc": 460, "chlor": 461, "region": 462, "eg": 463, "sof": 464, "##sul": 465, "##ry": 466, "coated": 467, "floor": 468, "soft": 469, "cor": 470, "light": 471, "comb": 472, "board": 473, "lumber": 474, "c4": 475, "##kv": 476, "kil": 477, "mel": 478, "kiln": 479, "bf": 480, "coat": 481, "a5": 482, "recycling": 483, "##x1": 484, "##ical": 485, "inciner": 486, "##mer": 487, "haul": 488, "##ens": 489, "##ode": 490, "cut": 491, "fl": 492, "he": 493, "pet": 494, "up": 495, "##anul": 496, "direct": 497, "granul": 498, "incineration": 499, "heat": 500, "melamine": 501, "but": 502, "softwood": 503, "rene": 504, "insul": 505, "200": 506, "vda": 507, "cement": 508, "renew": 509, "e1": 510, "##ethane": 511, "15": 512, "pin": 513, "##uf": 514, "##dust": 515, "25": 516, "hd": 517, "sm": 518, "sy": 519, "slu": 520, "##ia": 521, "##op": 522, "##dge": 523, "##yp": 524, "##int": 525, "##ingle": 526, "dist": 527, "roofing": 528, "granulate": 529, "sludge": 530, "mill": 531, "sp": 532, "##ding": 533, "electr": 534, "##board": 535, "##ue": 536, "##if": 537, "##ong": 538, "flooring": 539, "22": 540, "fo": 541, "pe": 542, "##infor": 543, "reinfor": 544, "reinforc": 545, "##ied": 546, "##ox": 547, "##tr": 548, "mem": 549, "##ive": 550, "nm": 551, "single": 552, "##per": 553, "##fin": 554, "##icle": 555, "incl": 556, "##sph": 557, "tex": 558, "plywood": 559, "model": 560, "16": 561, "18": 562, "man": 563, "ph": 564, "saw": 565, "##oline": 566, "indi": 567, "gasoline": 568, "indirect": 569, "dw": 570, "##yr": 571, "##ete": 572, "##ell": 573, "resist": 574, "hardwood": 575, "sil": 576, "##anol": 577, "ba": 578, "coating": 579, "000": 580, "pip": 581, "##ool": 582, "boiler": 583, "sul": 584, "indust": 585, "membr": 586, "##ous": 587, "##ides": 588, "##crete": 589, "membrane": 590, "resin": 591, "insulation": 592, "32": 593, "econ": 594, "ic": 595, "econom": 596, "pv": 597, "##pylene": 598, "##hyd": 599 } } }