TingChen-ppmc
commited on
Commit
•
a6cbb0b
1
Parent(s):
e238a0a
Update README.md
Browse files
README.md
CHANGED
@@ -21,6 +21,48 @@ configs:
|
|
21 |
- split: train
|
22 |
path: data/train-*
|
23 |
---
|
24 |
-
#
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
25 |
|
26 |
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
|
|
21 |
- split: train
|
22 |
path: data/train-*
|
23 |
---
|
24 |
+
# Corpus
|
25 |
+
|
26 |
+
This dataset is built from Magicdata [ASR-CZDIACSC: A CHINESE ZHENGZHOU DIALECT CONVERSATIONAL SPEECH CORPUS](https://magichub.com/datasets/zhengzhou-dialect-conversational-speech-corpus/)
|
27 |
+
|
28 |
+
This corpus is licensed under a [Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License](http://creativecommons.org/licenses/by-nc-nd/4.0/). Please refer to the license for further information.
|
29 |
+
|
30 |
+
Modifications: The audio is split in sentences based on the time span on the transcription file. Sentences that span less than 1 second is discarded. Topics of conversation is removed.
|
31 |
+
|
32 |
+
# Usage
|
33 |
+
|
34 |
+
To load this dataset, use
|
35 |
+
|
36 |
+
```python
|
37 |
+
from datasets import load_dataset
|
38 |
+
dialect_corpus = load_dataset("TingChen-ppmc/Zhengzhou_Dialect_Conversational_Speech_Corpus")
|
39 |
+
```
|
40 |
+
|
41 |
+
This dataset only has train split. To split out a test split, use
|
42 |
+
|
43 |
+
```python
|
44 |
+
from datasets import load_dataset
|
45 |
+
train_split = load_dataset("TingChen-ppmc/Zhengzhou_Dialect_Conversational_Speech_Corpus", split="train")
|
46 |
+
# where test=0.5 denotes 0.5 of the dataset will be split to test split
|
47 |
+
corpus = train_split.train_test_split(test=0.5)
|
48 |
+
```
|
49 |
+
|
50 |
+
A sample data would be
|
51 |
+
|
52 |
+
```python
|
53 |
+
# note this data is from the Nanchang Dialect corpus, the data format is shared
|
54 |
+
{'audio':
|
55 |
+
{'path': 'A0001_S001_0_G0001_0.WAV',
|
56 |
+
'array': array([-0.00030518, -0.00039673,
|
57 |
+
-0.00036621, ..., -0.00064087,
|
58 |
+
-0.00015259, -0.00042725]),
|
59 |
+
'sampling_rate': 16000},
|
60 |
+
'gender': '女',
|
61 |
+
'speaker_id': 'G0001',
|
62 |
+
'transcription': '北京爱数智慧语音采集'
|
63 |
+
}
|
64 |
+
```
|
65 |
+
|
66 |
+
|
67 |
|
68 |
[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|