Datasets:

Languages:
Arabic
Multilinguality:
monolingual
Size Categories:
1K<n<10K
Language Creators:
crowdsourced
Annotations Creators:
expert-generated
Source Datasets:
original
Tags:
License:
lhoestq HF staff commited on
Commit
8d7c2fd
1 Parent(s): 4cb1641

add dataset_info in dataset metadata

Browse files
Files changed (1) hide show
  1. README.md +25 -1
README.md CHANGED
@@ -33,6 +33,30 @@ train-eval-index:
33
  name: WER
34
  - type: cer
35
  name: CER
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
36
  ---
37
 
38
  # Dataset Card for Arabic Speech Corpus
@@ -204,4 +228,4 @@ This dataset was created by:
204
  * Nawar Halabi [@nawarhalabi](https://github.com/nawarhalabi) main creator and annotator.
205
  * Two anonymous Arabic langauge teachers as annotators.
206
  * One anonymous voice talent.
207
- * Thanks to [@zaidalyafeai](https://github.com/zaidalyafeai) for adding this dataset.
33
  name: WER
34
  - type: cer
35
  name: CER
36
+ dataset_info:
37
+ features:
38
+ - name: file
39
+ dtype: string
40
+ - name: text
41
+ dtype: string
42
+ - name: audio
43
+ dtype:
44
+ audio:
45
+ sampling_rate: 48000
46
+ - name: phonetic
47
+ dtype: string
48
+ - name: orthographic
49
+ dtype: string
50
+ config_name: clean
51
+ splits:
52
+ - name: test
53
+ num_bytes: 65784
54
+ num_examples: 100
55
+ - name: train
56
+ num_bytes: 1002365
57
+ num_examples: 1813
58
+ download_size: 1192302846
59
+ dataset_size: 1068149
60
  ---
61
 
62
  # Dataset Card for Arabic Speech Corpus
228
  * Nawar Halabi [@nawarhalabi](https://github.com/nawarhalabi) main creator and annotator.
229
  * Two anonymous Arabic langauge teachers as annotators.
230
  * One anonymous voice talent.
231
+ * Thanks to [@zaidalyafeai](https://github.com/zaidalyafeai) for adding this dataset.