{ "added_tokens_decoder": { "0": { "content": "", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "1": { "content": "", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "2": { "content": "", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "3": { "content": "", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270447": { "content": "", "lstrip": true, "normalized": true, "rstrip": false, "single_word": false, "special": true }, "270448": { "content": "ace_Arab", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270449": { "content": "ace_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270450": { "content": "acm_Arab", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270451": { "content": "acq_Arab", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270452": { "content": "aeb_Arab", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270453": { "content": "afr_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270454": { "content": "ajp_Arab", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270455": { "content": "aka_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270456": { "content": "als_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270457": { "content": "amh_Ethi", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270458": { "content": "apc_Arab", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270459": { "content": "arb_Arab", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270460": { "content": "ars_Arab", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270461": { "content": "ary_Arab", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270462": { "content": "arz_Arab", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270463": { "content": "asm_Beng", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270464": { "content": "ast_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270465": { "content": "awa_Deva", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270466": { "content": "ayr_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270467": { "content": "azb_Arab", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270468": { "content": "azj_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270469": { "content": "bak_Cyrl", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270470": { "content": "bam_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270471": { "content": "ban_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270472": { "content": "bel_Cyrl", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270473": { "content": "bem_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270474": { "content": "ben_Beng", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270475": { "content": "bho_Deva", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270476": { "content": "bjn_Arab", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270477": { "content": "bjn_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270478": { "content": "bod_Tibt", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270479": { "content": "bos_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270480": { "content": "bug_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270481": { "content": "bul_Cyrl", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270482": { "content": "cat_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270483": { "content": "ceb_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270484": { "content": "ces_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270485": { "content": "cjk_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270486": { "content": "ckb_Arab", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270487": { "content": "crh_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270488": { "content": "cym_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270489": { "content": "dan_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270490": { "content": "deu_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270491": { "content": "dik_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270492": { "content": "dyu_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270493": { "content": "dzo_Tibt", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270494": { "content": "ell_Grek", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270495": { "content": "eng_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270496": { "content": "epo_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270497": { "content": "est_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270498": { "content": "eus_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270499": { "content": "ewe_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270500": { "content": "fao_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270501": { "content": "fij_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270502": { "content": "fin_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270503": { "content": "fon_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270504": { "content": "fra_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270505": { "content": "fur_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270506": { "content": "fuv_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270507": { "content": "gaz_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270508": { "content": "gla_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270509": { "content": "gle_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270510": { "content": "glg_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270511": { "content": "grn_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270512": { "content": "guj_Gujr", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270513": { "content": "hat_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270514": { "content": "hau_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270515": { "content": "heb_Hebr", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270516": { "content": "hin_Deva", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270517": { "content": "hne_Deva", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270518": { "content": "hrv_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270519": { "content": "hun_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270520": { "content": "hye_Armn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270521": { "content": "ibo_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270522": { "content": "ilo_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270523": { "content": "ind_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270524": { "content": "isl_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270525": { "content": "ita_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270526": { "content": "jav_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270527": { "content": "jpn_Jpan", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270528": { "content": "kab_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270529": { "content": "kac_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270530": { "content": "kam_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270531": { "content": "kan_Knda", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270532": { "content": "kas_Arab", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270533": { "content": "kas_Deva", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270534": { "content": "kat_Geor", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270535": { "content": "kaz_Cyrl", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270536": { "content": "kbp_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270537": { "content": "kea_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270538": { "content": "khk_Cyrl", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270539": { "content": "khm_Khmr", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270540": { "content": "kik_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270541": { "content": "kin_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270542": { "content": "kir_Cyrl", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270543": { "content": "kmb_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270544": { "content": "kmr_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270545": { "content": "knc_Arab", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270546": { "content": "knc_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270547": { "content": "kon_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270548": { "content": "kor_Hang", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270549": { "content": "lao_Laoo", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270550": { "content": "lez_Cyrl", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270551": { "content": "lij_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270552": { "content": "lim_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270553": { "content": "lin_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270554": { "content": "lit_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270555": { "content": "lmo_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270556": { "content": "ltg_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270557": { "content": "ltz_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270558": { "content": "lua_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270559": { "content": "lug_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270560": { "content": "luo_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270561": { "content": "lus_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270562": { "content": "lvs_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270563": { "content": "mag_Deva", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270564": { "content": "mai_Deva", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270565": { "content": "mal_Mlym", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270566": { "content": "mar_Deva", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270567": { "content": "min_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270568": { "content": "mkd_Cyrl", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270569": { "content": "mlt_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270570": { "content": "mni_Beng", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270571": { "content": "mos_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270572": { "content": "mri_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270573": { "content": "mya_Mymr", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270574": { "content": "nld_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270575": { "content": "nno_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270576": { "content": "nob_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270577": { "content": "npi_Deva", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270578": { "content": "nso_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270579": { "content": "nus_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270580": { "content": "nya_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270581": { "content": "oci_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270582": { "content": "ory_Orya", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270583": { "content": "pag_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270584": { "content": "pan_Guru", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270585": { "content": "pap_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270586": { "content": "pbt_Arab", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270587": { "content": "pes_Arab", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270588": { "content": "plt_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270589": { "content": "pol_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270590": { "content": "por_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270591": { "content": "prs_Arab", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270592": { "content": "quy_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270593": { "content": "ron_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270594": { "content": "run_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270595": { "content": "rus_Cyrl", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270596": { "content": "sag_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270597": { "content": "san_Deva", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270598": { "content": "sat_Beng", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270599": { "content": "scn_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270600": { "content": "shn_Mymr", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270601": { "content": "sin_Sinh", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270602": { "content": "slk_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270603": { "content": "slv_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270604": { "content": "smo_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270605": { "content": "sna_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270606": { "content": "snd_Arab", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270607": { "content": "som_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270608": { "content": "sot_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270609": { "content": "spa_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270610": { "content": "srd_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270611": { "content": "srp_Cyrl", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270612": { "content": "ssw_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270613": { "content": "sun_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270614": { "content": "swe_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270615": { "content": "swh_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270616": { "content": "szl_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270617": { "content": "tam_Taml", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270618": { "content": "taq_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270619": { "content": "taq_Tfng", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270620": { "content": "tat_Cyrl", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270621": { "content": "tel_Telu", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270622": { "content": "tgk_Cyrl", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270623": { "content": "tgl_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270624": { "content": "tha_Thai", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270625": { "content": "tir_Ethi", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270626": { "content": "tpi_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270627": { "content": "tsn_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270628": { "content": "tso_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270629": { "content": "tuk_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270630": { "content": "tum_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270631": { "content": "tur_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270632": { "content": "twi_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270633": { "content": "tzm_Tfng", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270634": { "content": "uig_Arab", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270635": { "content": "ukr_Cyrl", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270636": { "content": "umb_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270637": { "content": "urd_Arab", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270638": { "content": "uzn_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270639": { "content": "vec_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270640": { "content": "vie_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270641": { "content": "war_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270642": { "content": "wol_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270643": { "content": "xho_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270644": { "content": "ydd_Hebr", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270645": { "content": "yor_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270646": { "content": "yue_Hant", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270647": { "content": "zho_Hans", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270648": { "content": "zho_Hant", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270649": { "content": "zsm_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true }, "270650": { "content": "zul_Latn", "lstrip": false, "normalized": false, "rstrip": false, "single_word": false, "special": true } }, "additional_special_tokens": [ "ace_Arab", "ace_Latn", "acm_Arab", "acq_Arab", "aeb_Arab", "afr_Latn", "ajp_Arab", "aka_Latn", "als_Latn", "amh_Ethi", "apc_Arab", "arb_Arab", "ars_Arab", "ary_Arab", "arz_Arab", "asm_Beng", "ast_Latn", "awa_Deva", "ayr_Latn", "azb_Arab", "azj_Latn", "bak_Cyrl", "bam_Latn", "ban_Latn", "bel_Cyrl", "bem_Latn", "ben_Beng", "bho_Deva", "bjn_Arab", "bjn_Latn", "bod_Tibt", "bos_Latn", "bug_Latn", "bul_Cyrl", "cat_Latn", "ceb_Latn", "ces_Latn", "cjk_Latn", "ckb_Arab", "crh_Latn", "cym_Latn", "dan_Latn", "deu_Latn", "dik_Latn", "dyu_Latn", "dzo_Tibt", "ell_Grek", "eng_Latn", "epo_Latn", "est_Latn", "eus_Latn", "ewe_Latn", "fao_Latn", "fij_Latn", "fin_Latn", "fon_Latn", "fra_Latn", "fur_Latn", "fuv_Latn", "gaz_Latn", "gla_Latn", "gle_Latn", "glg_Latn", "grn_Latn", "guj_Gujr", "hat_Latn", "hau_Latn", "heb_Hebr", "hin_Deva", "hne_Deva", "hrv_Latn", "hun_Latn", "hye_Armn", "ibo_Latn", "ilo_Latn", "ind_Latn", "isl_Latn", "ita_Latn", "jav_Latn", "jpn_Jpan", "kab_Latn", "kac_Latn", "kam_Latn", "kan_Knda", "kas_Arab", "kas_Deva", "kat_Geor", "kaz_Cyrl", "kbp_Latn", "kea_Latn", "khk_Cyrl", "khm_Khmr", "kik_Latn", "kin_Latn", "kir_Cyrl", "kmb_Latn", "kmr_Latn", "knc_Arab", "knc_Latn", "kon_Latn", "kor_Hang", "lao_Laoo", "lez_Cyrl", "lij_Latn", "lim_Latn", "lin_Latn", "lit_Latn", "lmo_Latn", "ltg_Latn", "ltz_Latn", "lua_Latn", "lug_Latn", "luo_Latn", "lus_Latn", "lvs_Latn", "mag_Deva", "mai_Deva", "mal_Mlym", "mar_Deva", "min_Latn", "mkd_Cyrl", "mlt_Latn", "mni_Beng", "mos_Latn", "mri_Latn", "mya_Mymr", "nld_Latn", "nno_Latn", "nob_Latn", "npi_Deva", "nso_Latn", "nus_Latn", "nya_Latn", "oci_Latn", "ory_Orya", "pag_Latn", "pan_Guru", "pap_Latn", "pbt_Arab", "pes_Arab", "plt_Latn", "pol_Latn", "por_Latn", "prs_Arab", "quy_Latn", "ron_Latn", "run_Latn", "rus_Cyrl", "sag_Latn", "san_Deva", "sat_Beng", "scn_Latn", "shn_Mymr", "sin_Sinh", "slk_Latn", "slv_Latn", "smo_Latn", "sna_Latn", "snd_Arab", "som_Latn", "sot_Latn", "spa_Latn", "srd_Latn", "srp_Cyrl", "ssw_Latn", "sun_Latn", "swe_Latn", "swh_Latn", "szl_Latn", "tam_Taml", "taq_Latn", "taq_Tfng", "tat_Cyrl", "tel_Telu", "tgk_Cyrl", "tgl_Latn", "tha_Thai", "tir_Ethi", "tpi_Latn", "tsn_Latn", "tso_Latn", "tuk_Latn", "tum_Latn", "tur_Latn", "twi_Latn", "tzm_Tfng", "uig_Arab", "ukr_Cyrl", "umb_Latn", "urd_Arab", "uzn_Latn", "vec_Latn", "vie_Latn", "war_Latn", "wol_Latn", "xho_Latn", "ydd_Hebr", "yor_Latn", "yue_Hant", "zho_Hans", "zho_Hant", "zsm_Latn", "zul_Latn" ], "bos_token": "", "clean_up_tokenization_spaces": true, "cls_token": "", "eos_token": "", "legacy_behaviour": false, "mask_token": "", "model_max_length": 1024, "pad_token": "", "sep_token": "", "sp_model_kwargs": {}, "src_lang": "lez_Cyrl", "tgt_lang": "rus_Cyrl", "tokenizer_class": "NllbTokenizer", "unk_token": "" }