snu-nia-12
commited on
Commit
•
424743d
1
Parent(s):
28843b5
add model
Browse files- tokenizer_config.json +2 -2
- vocab.json +41 -41
tokenizer_config.json
CHANGED
@@ -2,11 +2,11 @@
|
|
2 |
"bos_token": "<s>",
|
3 |
"do_lower_case": false,
|
4 |
"eos_token": "</s>",
|
5 |
-
"name_or_path": "speech31/wav2vec2-large-TIMIT-
|
6 |
"pad_token": "[PAD]",
|
7 |
"processor_class": "Wav2Vec2Processor",
|
8 |
"replace_word_delimiter_char": " ",
|
9 |
-
"special_tokens_map_file": "/home/eunjung/.cache/huggingface/transformers/
|
10 |
"tokenizer_class": "Wav2Vec2CTCTokenizer",
|
11 |
"unk_token": "[UNK]",
|
12 |
"word_delimiter_token": "|"
|
|
|
2 |
"bos_token": "<s>",
|
3 |
"do_lower_case": false,
|
4 |
"eos_token": "</s>",
|
5 |
+
"name_or_path": "speech31/wav2vec2-large-TIMIT-IPA2",
|
6 |
"pad_token": "[PAD]",
|
7 |
"processor_class": "Wav2Vec2Processor",
|
8 |
"replace_word_delimiter_char": " ",
|
9 |
+
"special_tokens_map_file": "/home/eunjung/.cache/huggingface/transformers/1041fd19ccb8e52014df301635f1aceb10d678cce8443796811c57e878df9da4.fea372b8528a479b7415f13ca4e27a2f5f3782cbb3f15b4d19bb3cbe734e8137",
|
10 |
"tokenizer_class": "Wav2Vec2CTCTokenizer",
|
11 |
"unk_token": "[UNK]",
|
12 |
"word_delimiter_token": "|"
|
vocab.json
CHANGED
@@ -1,46 +1,46 @@
|
|
1 |
{
|
2 |
-
"'":
|
3 |
"[PAD]": 43,
|
4 |
"[UNK]": 42,
|
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 |
-
"q":
|
22 |
-
"r":
|
23 |
-
"s":
|
24 |
-
"t":
|
25 |
-
"u":
|
26 |
"v": 41,
|
27 |
-
"w":
|
28 |
-
"x":
|
29 |
-
"y":
|
30 |
-
"z":
|
31 |
-
"|":
|
32 |
-
"æ":
|
33 |
-
"ð":
|
34 |
-
"ŋ":
|
35 |
-
"ɑ":
|
36 |
-
"ɔ":
|
37 |
-
"ə":
|
38 |
-
"ɛ":
|
39 |
-
"ɪ":
|
40 |
-
"ʃ":
|
41 |
-
"ʊ":
|
42 |
-
"ʒ":
|
43 |
-
"ʤ":
|
44 |
-
"ʧ":
|
45 |
-
"θ":
|
46 |
}
|
|
|
1 |
{
|
2 |
+
"'": 38,
|
3 |
"[PAD]": 43,
|
4 |
"[UNK]": 42,
|
5 |
+
"a": 32,
|
6 |
+
"b": 40,
|
7 |
+
"c": 29,
|
8 |
+
"d": 21,
|
9 |
+
"e": 14,
|
10 |
+
"f": 36,
|
11 |
+
"g": 3,
|
12 |
+
"h": 30,
|
13 |
+
"i": 13,
|
14 |
+
"j": 4,
|
15 |
+
"k": 24,
|
16 |
+
"l": 35,
|
17 |
+
"m": 17,
|
18 |
+
"n": 9,
|
19 |
+
"o": 2,
|
20 |
+
"p": 5,
|
21 |
+
"q": 23,
|
22 |
+
"r": 8,
|
23 |
+
"s": 27,
|
24 |
+
"t": 6,
|
25 |
+
"u": 18,
|
26 |
"v": 41,
|
27 |
+
"w": 34,
|
28 |
+
"x": 25,
|
29 |
+
"y": 11,
|
30 |
+
"z": 22,
|
31 |
+
"|": 1,
|
32 |
+
"æ": 20,
|
33 |
+
"ð": 31,
|
34 |
+
"ŋ": 33,
|
35 |
+
"ɑ": 28,
|
36 |
+
"ɔ": 39,
|
37 |
+
"ə": 0,
|
38 |
+
"ɛ": 19,
|
39 |
+
"ɪ": 15,
|
40 |
+
"ʃ": 12,
|
41 |
+
"ʊ": 10,
|
42 |
+
"ʒ": 7,
|
43 |
+
"ʤ": 16,
|
44 |
+
"ʧ": 26,
|
45 |
+
"θ": 37
|
46 |
}
|