IndicBERT_WR / tokenizer_config.json
contactsharma's picture
uploaded model with rationale with learning rate 2e-5
31c7642 verified
{
"added_tokens_decoder": {
"0": {
"content": "[UNK]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"1": {
"content": "[CLS]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"2": {
"content": "[SEP]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"3": {
"content": "[PAD]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"4": {
"content": "[MASK]",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"5": {
"content": "<as>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"6": {
"content": "<bd>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"7": {
"content": "<bn>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"8": {
"content": "<dg>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"9": {
"content": "<en>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"10": {
"content": "<gom>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"11": {
"content": "<gu>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"12": {
"content": "<hi>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"13": {
"content": "<kha>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"14": {
"content": "<kn>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"15": {
"content": "<ks>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"16": {
"content": "<mai>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"17": {
"content": "<ml>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"18": {
"content": "<mni>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"19": {
"content": "<mr>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"20": {
"content": "<ne>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"21": {
"content": "<or>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"22": {
"content": "<pa>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"23": {
"content": "<sa>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24": {
"content": "<sd>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"25": {
"content": "<sat>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"26": {
"content": "<ta>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"27": {
"content": "<te>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"28": {
"content": "<ur>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250000": {
"content": "πŸ™‰",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250001": {
"content": "πŸ“¦",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250002": {
"content": "🍾",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250003": {
"content": "πŸ₯²",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250004": {
"content": "πŸ’",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250005": {
"content": "πŸ—",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250006": {
"content": "🦾",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250007": {
"content": "🀫",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250008": {
"content": "🍱",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250009": {
"content": "πŸ’Έ",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250010": {
"content": "πŸ“",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250011": {
"content": "πŸ₯ƒ",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250012": {
"content": "🫑",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250013": {
"content": "🍰",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250014": {
"content": "πŸ₯±",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250015": {
"content": "🍚",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250016": {
"content": "πŸ₯΅",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250017": {
"content": "πŸ›",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250018": {
"content": "🎏",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250019": {
"content": "πŸͺ“",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250020": {
"content": "β›ͺ",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250021": {
"content": "πŸ₯Ά",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250022": {
"content": "πŸ₯š",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250023": {
"content": "πŸ›•",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250024": {
"content": "✝",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250025": {
"content": "πŸƒ",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250026": {
"content": "🚘",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250027": {
"content": "πŸ–",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250028": {
"content": "πŸ‘ƒ",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250029": {
"content": "πŸ•Ž",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250030": {
"content": "πŸ«‚",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250031": {
"content": "🦬",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250032": {
"content": "πŸ’Ί",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250033": {
"content": "🐏",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250034": {
"content": "🍨",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250035": {
"content": "🫒",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250036": {
"content": "🀠",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250037": {
"content": "⛏",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250038": {
"content": "πŸ‘Ί",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250039": {
"content": "🦰",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250040": {
"content": "πŸ•",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250041": {
"content": "🫢",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250042": {
"content": "πŸš—",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250043": {
"content": "🏰",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250044": {
"content": "🍡",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250045": {
"content": "πŸ₯›",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250046": {
"content": "πŸ€‘",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
}
},
"clean_up_tokenization_spaces": false,
"cls_token": "[CLS]",
"mask_token": "[MASK]",
"model_max_length": 1000000000000000019884624838656,
"pad_token": "[PAD]",
"sep_token": "[SEP]",
"tokenizer_class": "PreTrainedTokenizerFast",
"unk_token": "[UNK]"
}