Upload processor

special_tokens_map.json

@@ -0,0 +1,6 @@
+{
+  "bos_token": "<s>",
+  "eos_token": "</s>",
+  "pad_token": "[PAD]",
+  "unk_token": "[UNK]"
+}

tokenizer_config.json

@@ -0,0 +1,13 @@
+{
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": true,
+  "do_lower_case": false,
+  "eos_token": "</s>",
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "[PAD]",
+  "processor_class": "Wav2Vec2Processor",
+  "replace_word_delimiter_char": " ",
+  "tokenizer_class": "Wav2Vec2CTCTokenizer",
+  "unk_token": "[UNK]",
+  "word_delimiter_token": "|"
+}

vocab.json

@@ -0,0 +1,110 @@
+{
+  "[PAD]": 107,
+  "[UNK]": 106,
+  "a": 30,
+  "b": 55,
+  "c": 80,
+  "d": 85,
+  "e": 86,
+  "f": 71,
+  "g": 41,
+  "h": 37,
+  "i": 72,
+  "j": 77,
+  "k": 51,
+  "l": 16,
+  "m": 95,
+  "n": 4,
+  "o": 17,
+  "p": 53,
+  "r": 52,
+  "s": 33,
+  "t": 10,
+  "u": 87,
+  "v": 27,
+  "w": 46,
+  "y": 76,
+  "z": 13,
+  "|": 3,
+  "་": 78,
+  "༌": 90,
+  "།": 39,
+  "༑": 81,
+  "༠": 98,
+  "༡": 29,
+  "༢": 21,
+  "༣": 94,
+  "༤": 79,
+  "༥": 15,
+  "༦": 47,
+  "༧": 20,
+  "༨": 62,
+  "༩": 45,
+  "ཀ": 31,
+  "ཁ": 96,
+  "ག": 92,
+  "ང": 74,
+  "ཅ": 61,
+  "ཆ": 67,
+  "ཇ": 5,
+  "ཉ": 23,
+  "ཊ": 59,
+  "ཌ": 40,
+  "ཎ": 100,
+  "ཏ": 102,
+  "ཐ": 44,
+  "ད": 88,
+  "ན": 104,
+  "པ": 35,
+  "ཕ": 65,
+  "བ": 34,
+  "མ": 18,
+  "ཙ": 64,
+  "ཚ": 66,
+  "ཛ": 69,
+  "ཝ": 50,
+  "ཞ": 7,
+  "ཟ": 82,
+  "འ": 26,
+  "ཡ": 19,
+  "ར": 36,
+  "ལ": 68,
+  "ཤ": 22,
+  "ས": 97,
+  "ཧ": 56,
+  "ཨ": 14,
+  "ཱ": 63,
+  "ི": 42,
+  "ུ": 54,
+  "ེ": 70,
+  "ོ": 60,
+  "ྐ": 38,
+  "ྒ": 0,
+  "ྔ": 11,
+  "ྕ": 48,
+  "ྗ": 105,
+  "ྙ": 1,
+  "ྜ": 89,
+  "ྟ": 28,
+  "ྡ": 12,
+  "ྣ": 91,
+  "ྤ": 99,
+  "ྥ": 73,
+  "ྦ": 93,
+  "ྨ": 57,
+  "ྩ": 25,
+  "ྫ": 8,
+  "ྭ": 6,
+  "ྱ": 101,
+  "ྲ": 43,
+  "ླ": 58,
+  "ྷ": 83,
+  "“": 75,
+  "”": 2,
+  "五": 24,
+  "傅": 103,
+  "师": 32,
+  "康": 9,
+  "液": 84,
+  "粮": 49
+}