Upload tokenizer
Browse files- added_tokens.json +4 -0
- special_tokens_map.json +22 -0
- tokenizer_config.json +13 -0
- vocab.json +79 -0
added_tokens.json
ADDED
|
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"</s>": 76,
|
| 3 |
+
"<s>": 75
|
| 4 |
+
}
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
{
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": true,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false
|
| 9 |
+
},
|
| 10 |
+
{
|
| 11 |
+
"content": "</s>",
|
| 12 |
+
"lstrip": false,
|
| 13 |
+
"normalized": true,
|
| 14 |
+
"rstrip": false,
|
| 15 |
+
"single_word": false
|
| 16 |
+
}
|
| 17 |
+
],
|
| 18 |
+
"bos_token": "<s>",
|
| 19 |
+
"eos_token": "</s>",
|
| 20 |
+
"pad_token": "[PAD]",
|
| 21 |
+
"unk_token": "[UNK]"
|
| 22 |
+
}
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"clean_up_tokenization_spaces": true,
|
| 4 |
+
"do_lower_case": false,
|
| 5 |
+
"eos_token": "</s>",
|
| 6 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 7 |
+
"pad_token": "[PAD]",
|
| 8 |
+
"replace_word_delimiter_char": " ",
|
| 9 |
+
"target_lang": "ara",
|
| 10 |
+
"tokenizer_class": "Wav2Vec2CTCTokenizer",
|
| 11 |
+
"unk_token": "[UNK]",
|
| 12 |
+
"word_delimiter_token": "|"
|
| 13 |
+
}
|
vocab.json
ADDED
|
@@ -0,0 +1,79 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"ara": {
|
| 3 |
+
"[PAD]": 74,
|
| 4 |
+
"[UNK]": 73,
|
| 5 |
+
"a": 1,
|
| 6 |
+
"b": 2,
|
| 7 |
+
"c": 3,
|
| 8 |
+
"d": 4,
|
| 9 |
+
"e": 5,
|
| 10 |
+
"f": 6,
|
| 11 |
+
"g": 7,
|
| 12 |
+
"h": 8,
|
| 13 |
+
"i": 9,
|
| 14 |
+
"j": 10,
|
| 15 |
+
"k": 11,
|
| 16 |
+
"l": 12,
|
| 17 |
+
"m": 13,
|
| 18 |
+
"n": 14,
|
| 19 |
+
"o": 15,
|
| 20 |
+
"p": 16,
|
| 21 |
+
"q": 17,
|
| 22 |
+
"r": 18,
|
| 23 |
+
"s": 19,
|
| 24 |
+
"t": 20,
|
| 25 |
+
"u": 21,
|
| 26 |
+
"v": 22,
|
| 27 |
+
"w": 23,
|
| 28 |
+
"x": 24,
|
| 29 |
+
"y": 25,
|
| 30 |
+
"z": 26,
|
| 31 |
+
"|": 0,
|
| 32 |
+
"،": 27,
|
| 33 |
+
"؟": 28,
|
| 34 |
+
"ء": 29,
|
| 35 |
+
"آ": 30,
|
| 36 |
+
"أ": 31,
|
| 37 |
+
"ؤ": 32,
|
| 38 |
+
"إ": 33,
|
| 39 |
+
"ئ": 34,
|
| 40 |
+
"ا": 35,
|
| 41 |
+
"ب": 36,
|
| 42 |
+
"ة": 37,
|
| 43 |
+
"ت": 38,
|
| 44 |
+
"ث": 39,
|
| 45 |
+
"ج": 40,
|
| 46 |
+
"ح": 41,
|
| 47 |
+
"خ": 42,
|
| 48 |
+
"د": 43,
|
| 49 |
+
"ذ": 44,
|
| 50 |
+
"ر": 45,
|
| 51 |
+
"ز": 46,
|
| 52 |
+
"س": 47,
|
| 53 |
+
"ش": 48,
|
| 54 |
+
"ص": 49,
|
| 55 |
+
"ض": 50,
|
| 56 |
+
"ط": 51,
|
| 57 |
+
"ظ": 52,
|
| 58 |
+
"ع": 53,
|
| 59 |
+
"غ": 54,
|
| 60 |
+
"ف": 55,
|
| 61 |
+
"ق": 56,
|
| 62 |
+
"ك": 57,
|
| 63 |
+
"ل": 58,
|
| 64 |
+
"م": 59,
|
| 65 |
+
"ن": 60,
|
| 66 |
+
"ه": 61,
|
| 67 |
+
"و": 62,
|
| 68 |
+
"ى": 63,
|
| 69 |
+
"ي": 64,
|
| 70 |
+
"ً": 65,
|
| 71 |
+
"ٌ": 66,
|
| 72 |
+
"ٍ": 67,
|
| 73 |
+
"َ": 68,
|
| 74 |
+
"ُ": 69,
|
| 75 |
+
"ِ": 70,
|
| 76 |
+
"ّ": 71,
|
| 77 |
+
"ْ": 72
|
| 78 |
+
}
|
| 79 |
+
}
|