Upload processor

preprocessor_config.json

@@ -0,0 +1,10 @@
+{
+  "do_normalize": true,
+  "feature_extractor_type": "Wav2Vec2FeatureExtractor",
+  "feature_size": 1,
+  "padding_side": "right",
+  "padding_value": 0.0,
+  "processor_class": "Wav2Vec2Processor",
+  "return_attention_mask": false,
+  "sampling_rate": 16000
+}

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 +1,14 @@
+{
+  "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": " ",
+  "target_lang": null,
+  "tokenizer_class": "Wav2Vec2CTCTokenizer",
+  "unk_token": "[UNK]",
+  "word_delimiter_token": "|"
+}

vocab.json

@@ -1 +1,112 @@
-{"c": 0, "\u1eef": 1, "i": 2, "\u1eb7": 3, "\u1eed": 4, "\u00e8": 5, "\u1edb": 6, "\u1ef3": 7, "\u1ea3": 8, "o": 9, "\u00e2": 10, "\u00fd": 11, "\u00e3": 12, "\u0300": 13, "t": 14, "\u1ebf": 15, "\u1eb1": 16, "\u1ef5": 17, "\u1ecd": 18, "\u00f9": 19, "\u01a1": 20, "\u1ecf": 21, "r": 22, "d": 23, "q": 24, "\u00f5": 25, "\u1eaf": 26, "h": 27, "\u1eeb": 28, "\u00f2": 29, "\u1ed5": 30, "5": 31, "a": 32, "z": 33, "\u1ead": 34, "\u1ef1": 35, "s": 36, "w": 37, "g": 38, "\u1edd": 39, "\u1ea1": 40, "\u1eb9": 41, "e": 42, "4": 43, "\u0323": 44, "\u0443": 45, "\u1ed1": 46, "\u1ed7": 47, "\u1ef7": 48, "\u00ec": 49, "\u01b0": 50, "3": 51, "\u1ebd": 52, "p": 53, "\u1ed3": 54, "\u1ec9": 55, "\u1ecb": 56, "\u1ea5": 57, "y": 58, "\u1ec1": 59, "x": 60, "\u1ee3": 61, "\u1ea9": 62, "\u0129": 63, "n": 64, "1": 65, "\u1eab": 66, "6": 67, "\u1edf": 68, "\u1ee1": 69, "9": 70, "\u00e1": 71, "v": 72, "\u00f3": 73, "\u0111": 74, "l": 75, "\u1ea7": 76, "\u0169": 77, "f": 78, "\u1eb3": 79, "8": 80, "\u00e9": 81, "0": 82, "\u1ebb": 83, "\u00ea": 84, "\u00fa": 85, "\u1ec5": 86, "\u1ee7": 87, "\u1ef9": 88, "\u0301": 89, "\u00ed": 90, "\u00e0": 91, "\u1ee9": 92, "2": 93, "%": 94, "\u0103": 95, "k": 96, "\u1ee5": 97, "\u1ec7": 98, "b": 100, "u": 101, "\u00f4": 102, "\u1eb5": 103, "\u1ec3": 104, "m": 105, "\u1ed9": 106, "7": 107, "|": 99, "[UNK]": 108, "[PAD]": 109}
+{
+  "%": 94,
+  "0": 82,
+  "1": 65,
+  "2": 93,
+  "3": 51,
+  "4": 43,
+  "5": 31,
+  "6": 67,
+  "7": 107,
+  "8": 80,
+  "9": 70,
+  "[PAD]": 109,
+  "[UNK]": 108,
+  "a": 32,
+  "b": 100,
+  "c": 0,
+  "d": 23,
+  "e": 42,
+  "f": 78,
+  "g": 38,
+  "h": 27,
+  "i": 2,
+  "k": 96,
+  "l": 75,
+  "m": 105,
+  "n": 64,
+  "o": 9,
+  "p": 53,
+  "q": 24,
+  "r": 22,
+  "s": 36,
+  "t": 14,
+  "u": 101,
+  "v": 72,
+  "w": 37,
+  "x": 60,
+  "y": 58,
+  "z": 33,
+  "|": 99,
+  "à": 91,
+  "á": 71,
+  "â": 10,
+  "ã": 12,
+  "è": 5,
+  "é": 81,
+  "ê": 84,
+  "ì": 49,
+  "í": 90,
+  "ò": 29,
+  "ó": 73,
+  "ô": 102,
+  "õ": 25,
+  "ù": 19,
+  "ú": 85,
+  "ý": 11,
+  "ă": 95,
+  "đ": 74,
+  "ĩ": 63,
+  "ũ": 77,
+  "ơ": 20,
+  "ư": 50,
+  "̀": 13,
+  "́": 89,
+  "̣": 44,
+  "у": 45,
+  "ạ": 40,
+  "ả": 8,
+  "ấ": 57,
+  "ầ": 76,
+  "ẩ": 62,
+  "ẫ": 66,
+  "ậ": 34,
+  "ắ": 26,
+  "ằ": 16,
+  "ẳ": 79,
+  "ẵ": 103,
+  "ặ": 3,
+  "ẹ": 41,
+  "ẻ": 83,
+  "ẽ": 52,
+  "ế": 15,
+  "ề": 59,
+  "ể": 104,
+  "ễ": 86,
+  "ệ": 98,
+  "ỉ": 55,
+  "ị": 56,
+  "ọ": 18,
+  "ỏ": 21,
+  "ố": 46,
+  "ồ": 54,
+  "ổ": 30,
+  "ỗ": 47,
+  "ộ": 106,
+  "ớ": 6,
+  "ờ": 39,
+  "ở": 68,
+  "ỡ": 69,
+  "ợ": 61,
+  "ụ": 97,
+  "ủ": 87,
+  "ứ": 92,
+  "ừ": 28,
+  "ử": 4,
+  "ữ": 1,
+  "ự": 35,
+  "ỳ": 7,
+  "ỵ": 17,
+  "ỷ": 48,
+  "ỹ": 88
+}