DrishtiSharma commited on
Commit
e4c3b57
1 Parent(s): 2b3d238

add tokenizer

Browse files
Files changed (3) hide show
  1. special_tokens_map.json +1 -0
  2. tokenizer_config.json +1 -0
  3. vocab.json +1 -0
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
1
+ {"bos_token": "<s>", "eos_token": "</s>", "unk_token": "[UNK]", "pad_token": "[PAD]"}
tokenizer_config.json ADDED
@@ -0,0 +1 @@
 
1
+ {"unk_token": "[UNK]", "bos_token": "<s>", "eos_token": "</s>", "pad_token": "[PAD]", "do_lower_case": false, "word_delimiter_token": "|", "tokenizer_class": "Wav2Vec2CTCTokenizer"}
vocab.json ADDED
@@ -0,0 +1 @@
 
1
+ {"«": 1, "»": 2, "а": 3, "б": 4, "в": 5, "г": 6, "д": 7, "е": 8, "ж": 9, "з": 10, "и": 11, "й": 12, "к": 13, "л": 14, "м": 15, "н": 16, "о": 17, "п": 18, "р": 19, "с": 20, "т": 21, "у": 22, "ф": 23, "х": 24, "ц": 25, "ч": 26, "ш": 27, "щ": 28, "ъ": 29, "ы": 30, "ь": 31, "э": 32, "ю": 33, "я": 34, "ё": 35, "—": 36, "•": 37, "|": 0, "[UNK]": 38, "[PAD]": 39}