Update README.md
Browse files
README.md
CHANGED
@@ -7,6 +7,8 @@ language:
|
|
7 |
size_categories:
|
8 |
- 100K<n<1M
|
9 |
---
|
|
|
|
|
10 |
|
11 |
### Original Dataset can be found in <https://www.openslr.org/145/>
|
12 |
|
@@ -16,7 +18,7 @@ size_categories:
|
|
16 |
|
17 |
### How to Use
|
18 |
|
19 |
-
- Almost the same with [Librispeech](https://huggingface.co/datasets/openslr/librispeech_asr) dataset module since i refered to
|
20 |
|
21 |
- three types of transcripts are given
|
22 |
- `text_normalized` : the trascript from Librispeech ASR
|
@@ -31,17 +33,22 @@ import aiohttp
|
|
31 |
import os
|
32 |
|
33 |
# set huggingface cache directory for the extracted raw files and huggingface-cli token
|
34 |
-
|
35 |
-
|
|
|
36 |
|
37 |
|
38 |
# download dataset
|
39 |
# if already have librispeech_asr in the cache_dir it will use the same audio files.
|
40 |
-
libripc = load_dataset("yoom618/librispeech_pc",
|
41 |
-
"all",
|
42 |
-
cache_dir="/
|
43 |
trust_remote_code=True,
|
44 |
-
# storage_options={
|
|
|
|
|
|
|
|
|
45 |
|
46 |
|
47 |
# check dataset info
|
@@ -55,6 +62,30 @@ pprint(libripc['train.clean.100'][0]['audio'])
|
|
55 |
```
|
56 |
|
57 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
58 |
|
59 |
### The number of samples in Librispeech-PC
|
60 |
|
@@ -62,7 +93,7 @@ pprint(libripc['train.clean.100'][0]['audio'])
|
|
62 |
- `train.clean.100` : 26,041 ( 2,498 out of 28,539 were dropped )
|
63 |
- `train.clean.360` : 95,404 ( 8,610 out of 104,014 were dropped )
|
64 |
- `train.other.500` : 134,679 ( 14,009 out of 148,688 were dropped )
|
65 |
-
- dev
|
66 |
- `dev.clean` : 2,530 ( 173 out of 2,703 were dropped )
|
67 |
- `dev.other` : 2,728 ( 136 out of 2,864 were dropped )
|
68 |
- test
|
|
|
7 |
size_categories:
|
8 |
- 100K<n<1M
|
9 |
---
|
10 |
+
# Librispeech-PC (punctuation and capitalization restored)
|
11 |
+
|
12 |
|
13 |
### Original Dataset can be found in <https://www.openslr.org/145/>
|
14 |
|
|
|
18 |
|
19 |
### How to Use
|
20 |
|
21 |
+
- Almost the same with [Librispeech](https://huggingface.co/datasets/openslr/librispeech_asr) dataset module since i refered to its [source code](https://huggingface.co/datasets/openslr/librispeech_asr/blob/main/librispeech_asr.py).
|
22 |
|
23 |
- three types of transcripts are given
|
24 |
- `text_normalized` : the trascript from Librispeech ASR
|
|
|
33 |
import os
|
34 |
|
35 |
# set huggingface cache directory for the extracted raw files and huggingface-cli token
|
36 |
+
# if don't, you might download the raw tar.gz file in home cache dir even if you set `cache_dir` param
|
37 |
+
os.environ['HF_HOME'] = "/data_dir/to/download"
|
38 |
+
!export HF_HOME="/data_dir/to/download"
|
39 |
|
40 |
|
41 |
# download dataset
|
42 |
# if already have librispeech_asr in the cache_dir it will use the same audio files.
|
43 |
+
libripc = load_dataset("yoom618/librispeech_pc",
|
44 |
+
"all", # all, clean, other
|
45 |
+
cache_dir="/data_dir/to/download",
|
46 |
trust_remote_code=True,
|
47 |
+
# storage_options={
|
48 |
+
# # add if you need to increase the timeout for openslr download
|
49 |
+
# 'client_kwargs': {'timeout': aiohttp.ClientTimeout(total=7200)}
|
50 |
+
# },
|
51 |
+
)
|
52 |
|
53 |
|
54 |
# check dataset info
|
|
|
62 |
```
|
63 |
|
64 |
|
65 |
+
- Data Sample
|
66 |
+
|
67 |
+
```raw
|
68 |
+
{'audio': {'array': array([ 7.01904297e-04, 7.32421875e-04, 7.32421875e-04, ...,
|
69 |
+
-2.74658203e-04, -1.83105469e-04, -3.05175781e-05]),
|
70 |
+
'path': '/data_dir/to/download/downloads/extracted/.../374-180298-0000.flac',
|
71 |
+
'sampling_rate': 16000},
|
72 |
+
'chapter_id': 180298,
|
73 |
+
'duration': 14.529999732971191,
|
74 |
+
'file': '/data_dir/to/download/downloads/extracted/.../374-180298-0000.flac',
|
75 |
+
'id': '374-180298-0000',
|
76 |
+
'speaker_id': 374,
|
77 |
+
'text': 'Chapter sixteen I might have told you of the beginning of this '
|
78 |
+
'liaison in a few lines, but I wanted you to see every step by which '
|
79 |
+
'we came, I to agree to whatever Marguerite wished,',
|
80 |
+
'text_normalized': 'CHAPTER SIXTEEN I MIGHT HAVE TOLD YOU OF THE BEGINNING OF '
|
81 |
+
'THIS LIAISON IN A FEW LINES BUT I WANTED YOU TO SEE EVERY '
|
82 |
+
'STEP BY WHICH WE CAME I TO AGREE TO WHATEVER MARGUERITE '
|
83 |
+
'WISHED',
|
84 |
+
'text_raw': 'Chapter sixteen I might have told you of the beginning of this '
|
85 |
+
'liaison in a few lines, but I wanted you to see every step by '
|
86 |
+
'which we came, I to agree to whatever Marguerite wished,'}
|
87 |
+
```
|
88 |
+
|
89 |
|
90 |
### The number of samples in Librispeech-PC
|
91 |
|
|
|
93 |
- `train.clean.100` : 26,041 ( 2,498 out of 28,539 were dropped )
|
94 |
- `train.clean.360` : 95,404 ( 8,610 out of 104,014 were dropped )
|
95 |
- `train.other.500` : 134,679 ( 14,009 out of 148,688 were dropped )
|
96 |
+
- dev (validation)
|
97 |
- `dev.clean` : 2,530 ( 173 out of 2,703 were dropped )
|
98 |
- `dev.other` : 2,728 ( 136 out of 2,864 were dropped )
|
99 |
- test
|