gigant commited on
Commit
ee23015
1 Parent(s): 879f7b8

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +54 -3
README.md CHANGED
@@ -1,3 +1,54 @@
1
- ---
2
- license: cc
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc
3
+ pretty_name: M-AILABS Speech Dataset (French)
4
+ languages:
5
+ - fr
6
+ task_categories:
7
+ - speech-processing
8
+ task_ids:
9
+ - automatic-speech-recognition
10
+ size_categories:
11
+ fr:
12
+ - 1K<n<10K
13
+ ---
14
+
15
+ ## Dataset Description
16
+ - **Homepage:** https://www.caito.de/2019/01/the-m-ailabs-speech-dataset/
17
+
18
+ ### Dataset Summary
19
+
20
+ The M-AILABS Speech Dataset is the first large dataset that we are providing free-of-charge, freely usable as training data for speech recognition and speech synthesis.
21
+
22
+ Most of the data is based on LibriVox and Project Gutenberg. The training data consist of nearly thousand hours of audio and the text-files in prepared format.
23
+
24
+ A transcription is provided for each clip. Clips vary in length from 1 to 20 seconds and have a total length of approximately shown in the list (and in the respective info.txt-files) below.
25
+
26
+
27
+ The texts were published between 1884 and 1964, and are in the public domain. The audio was recorded by the LibriVox project and is also in the public domain – except for Ukrainian.
28
+
29
+ Ukrainian audio was kindly provided either by Nash Format or Gwara Media for machine learning purposes only (please check the data info.txt files for details).
30
+
31
+
32
+ ### Languages
33
+
34
+ French
35
+
36
+ ## Dataset Structure
37
+ ### Data Instances
38
+ A typical data point comprises the path to the audio file, called audio and its sentence.
39
+
40
+
41
+ ### Data Fields
42
+
43
+ - audio: A dictionary containing the path to the downloaded audio file, the decoded audio array, and the sampling rate. Note that when accessing the audio column: `dataset[0]["audio"]` the audio file is automatically decoded and resampled to `dataset.features["audio"].sampling_rate`. Decoding and resampling of a large number of audio files might take a significant amount of time. Thus it is important to first query the sample index before the `"audio"` column, *i.e.* `dataset[0]["audio"]` should **always** be preferred over `dataset["audio"][0]`.
44
+
45
+ - sentence: The sentence the user was prompted to speak
46
+
47
+ ### Data Splits
48
+ The speech material has been subdivided into portions for train and test.
49
+ The train split consists of [TODO] audio clips and the related sentences.
50
+ The test split consists of [TODO] audio clips and the related sentences.
51
+
52
+
53
+ ### Contributions
54
+ [@gigant](https://huggingface.co/gigant) added this dataset.