Commit
·
5a46cf4
1
Parent(s):
fbe6273
Upload tokenizer
Browse files- special_tokens_map.json +6 -0
- tokenizer_config.json +17 -0
- vocab.json +48 -0
special_tokens_map.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"eos_token": "</s>",
|
| 4 |
+
"pad_token": "<pad>",
|
| 5 |
+
"unk_token": "<unk>"
|
| 6 |
+
}
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 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 |
+
"reserved_token1": "?reserved2?",
|
| 10 |
+
"reserved_token2": "?reserved3?",
|
| 11 |
+
"silence_token": "<sil>",
|
| 12 |
+
"spoken_noise_token": "<spn>",
|
| 13 |
+
"tokenizer_class": "Wav2Vec2CTCTokenizer",
|
| 14 |
+
"unk_token": "<unk>",
|
| 15 |
+
"vocab_size": 46,
|
| 16 |
+
"word_delimiter_token": "|"
|
| 17 |
+
}
|
vocab.json
ADDED
|
@@ -0,0 +1,48 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<pad>": 32,
|
| 3 |
+
"<sil>": 4,
|
| 4 |
+
"<spn>": 23,
|
| 5 |
+
"<unk>": 42,
|
| 6 |
+
"?reserved2?": 8,
|
| 7 |
+
"?reserved3?": 21,
|
| 8 |
+
"AA": 20,
|
| 9 |
+
"AE": 33,
|
| 10 |
+
"AH": 2,
|
| 11 |
+
"AO": 7,
|
| 12 |
+
"AW": 44,
|
| 13 |
+
"AY": 41,
|
| 14 |
+
"B": 12,
|
| 15 |
+
"CH": 31,
|
| 16 |
+
"D": 26,
|
| 17 |
+
"DH": 6,
|
| 18 |
+
"DX": 35,
|
| 19 |
+
"EH": 27,
|
| 20 |
+
"ER": 18,
|
| 21 |
+
"EY": 34,
|
| 22 |
+
"F": 30,
|
| 23 |
+
"G": 19,
|
| 24 |
+
"HH": 22,
|
| 25 |
+
"IH": 43,
|
| 26 |
+
"IY": 10,
|
| 27 |
+
"JH": 36,
|
| 28 |
+
"K": 14,
|
| 29 |
+
"L": 29,
|
| 30 |
+
"M": 11,
|
| 31 |
+
"N": 24,
|
| 32 |
+
"NG": 40,
|
| 33 |
+
"OW": 0,
|
| 34 |
+
"OY": 16,
|
| 35 |
+
"P": 28,
|
| 36 |
+
"R": 13,
|
| 37 |
+
"S": 25,
|
| 38 |
+
"SH": 39,
|
| 39 |
+
"T": 17,
|
| 40 |
+
"TH": 38,
|
| 41 |
+
"UH": 45,
|
| 42 |
+
"UW": 1,
|
| 43 |
+
"V": 9,
|
| 44 |
+
"W": 5,
|
| 45 |
+
"Y": 37,
|
| 46 |
+
"Z": 3,
|
| 47 |
+
"ZH": 15
|
| 48 |
+
}
|