Upload tokenizer
Browse files- tokenizer_config.json +1 -0
- vocab.json +37 -37
tokenizer_config.json
CHANGED
|
@@ -6,6 +6,7 @@
|
|
| 6 |
"model_max_length": 1000000000000000019884624838656,
|
| 7 |
"pad_token": "[PAD]",
|
| 8 |
"replace_word_delimiter_char": " ",
|
|
|
|
| 9 |
"tokenizer_class": "Wav2Vec2CTCTokenizer",
|
| 10 |
"unk_token": "[UNK]",
|
| 11 |
"word_delimiter_token": "|"
|
|
|
|
| 6 |
"model_max_length": 1000000000000000019884624838656,
|
| 7 |
"pad_token": "[PAD]",
|
| 8 |
"replace_word_delimiter_char": " ",
|
| 9 |
+
"target_lang": null,
|
| 10 |
"tokenizer_class": "Wav2Vec2CTCTokenizer",
|
| 11 |
"unk_token": "[UNK]",
|
| 12 |
"word_delimiter_token": "|"
|
vocab.json
CHANGED
|
@@ -1,42 +1,42 @@
|
|
| 1 |
{
|
| 2 |
-
"'":
|
| 3 |
"[PAD]": 39,
|
| 4 |
"[UNK]": 38,
|
| 5 |
-
"a":
|
| 6 |
-
"b":
|
| 7 |
-
"c":
|
| 8 |
-
"d":
|
| 9 |
-
"e":
|
| 10 |
-
"f":
|
| 11 |
-
"g":
|
| 12 |
-
"h":
|
| 13 |
-
"i":
|
| 14 |
-
"j":
|
| 15 |
-
"k":
|
| 16 |
-
"l":
|
| 17 |
-
"m":
|
| 18 |
-
"n":
|
| 19 |
-
"o":
|
| 20 |
-
"p":
|
| 21 |
-
"r":
|
| 22 |
-
"s":
|
| 23 |
-
"t":
|
| 24 |
-
"u":
|
| 25 |
-
"v":
|
| 26 |
-
"w":
|
| 27 |
-
"y":
|
| 28 |
-
"z":
|
| 29 |
-
"|":
|
| 30 |
-
"á":
|
| 31 |
-
"â":
|
| 32 |
-
"ä":
|
| 33 |
-
"è":
|
| 34 |
-
"é":
|
| 35 |
-
"ê":
|
| 36 |
-
"ë":
|
| 37 |
"ï": 32,
|
| 38 |
-
"ô":
|
| 39 |
-
"ú":
|
| 40 |
-
"û":
|
| 41 |
-
"ü":
|
| 42 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"'": 33,
|
| 3 |
"[PAD]": 39,
|
| 4 |
"[UNK]": 38,
|
| 5 |
+
"a": 25,
|
| 6 |
+
"b": 23,
|
| 7 |
+
"c": 30,
|
| 8 |
+
"d": 37,
|
| 9 |
+
"e": 31,
|
| 10 |
+
"f": 8,
|
| 11 |
+
"g": 1,
|
| 12 |
+
"h": 15,
|
| 13 |
+
"i": 7,
|
| 14 |
+
"j": 35,
|
| 15 |
+
"k": 12,
|
| 16 |
+
"l": 24,
|
| 17 |
+
"m": 34,
|
| 18 |
+
"n": 27,
|
| 19 |
+
"o": 21,
|
| 20 |
+
"p": 6,
|
| 21 |
+
"r": 36,
|
| 22 |
+
"s": 9,
|
| 23 |
+
"t": 18,
|
| 24 |
+
"u": 10,
|
| 25 |
+
"v": 5,
|
| 26 |
+
"w": 13,
|
| 27 |
+
"y": 11,
|
| 28 |
+
"z": 3,
|
| 29 |
+
"|": 0,
|
| 30 |
+
"á": 22,
|
| 31 |
+
"â": 19,
|
| 32 |
+
"ä": 20,
|
| 33 |
+
"è": 4,
|
| 34 |
+
"é": 2,
|
| 35 |
+
"ê": 14,
|
| 36 |
+
"ë": 28,
|
| 37 |
"ï": 32,
|
| 38 |
+
"ô": 26,
|
| 39 |
+
"ú": 29,
|
| 40 |
+
"û": 17,
|
| 41 |
+
"ü": 16
|
| 42 |
}
|