readjust readme

README.md

@@ -1,13 +1,22 @@
 ---
-license: other
 dataset_info:
   features:
-  - name: audio
-    sampling_rate: 16000
-    type: Audio
-  - name: sentence
-    dtype: string
-    type: value
+    - name: input_features
+      sequence:
+        sequence: float32
+    - name: input_length
+      dtype: float64
+    - name: labels
+      sequence: int64
+  splits:
+    - name: train
+      num_bytes: 877085901.0485437
+      num_examples: 912
+    - name: test
+      num_bytes: 217335736.68965518
+      num_examples: 226
+  download_size: 888280535
+  dataset_size: 1094421637.7381988
 language:
 - ar
 tags:
@@ -19,6 +28,8 @@ tags:
 pretty_name: 'Egyptian Arabic dialect automatic speech recognition '
 size_categories:
 - 1K<n<10K
+task_categories:
+- automatic-speech-recognition
 ---
 # Egyptian Arabic dialect automatic speech recognition 
 
@@ -122,5 +133,4 @@ the resulting dataset was uploaded to huggingface.
 [Speech Recognition Challenge in the Wild: Arabic MGB-3](https://arxiv.org/abs/1709.07276)
 
 
-### Contributions
-
+### Contributions