Convert dataset to Parquet

Convert dataset to Parquet.

README.md

@@ -1,20 +1,22 @@
 ---
 annotations_creators:
 - expert-generated
-language:
-- mt
 language_creators:
 - other
+language:
+- mt
 license: cc-by-nc-sa-4.0
 multilinguality:
 - monolingual
-pretty_name: >-
-  MASRI-DEV CORPUS: Audio and Transcriptions in Maltese extracted from the
-  YouTube channel of the University of Malta.
 size_categories:
 - n<1K
 source_datasets:
 - original
+task_categories:
+- automatic-speech-recognition
+task_ids: []
+pretty_name: 'MASRI-DEV CORPUS: Audio and Transcriptions in Maltese extracted from
+  the YouTube channel of the University of Malta.'
 tags:
 - masri
 - maltese
@@ -23,9 +25,35 @@ tags:
 - evaluation corpus
 - development corpus
 - dev
-task_categories:
-- automatic-speech-recognition
-task_ids: []
+dataset_info:
+  config_name: masri_dev
+  features:
+  - name: audio_id
+    dtype: string
+  - name: audio
+    dtype:
+      audio:
+        sampling_rate: 16000
+  - name: speaker_id
+    dtype: string
+  - name: gender
+    dtype: string
+  - name: duration
+    dtype: float32
+  - name: normalized_text
+    dtype: string
+  splits:
+  - name: validation
+    num_bytes: 72810026.0
+    num_examples: 669
+  download_size: 72798422
+  dataset_size: 72810026.0
+configs:
+- config_name: masri_dev
+  data_files:
+  - split: validation
+    path: masri_dev/validation-*
+  default: true
 ---
 # Dataset Card for masri_dev
 ## Table of Contents

masri_dev/validation-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:365b16f48b8496c65841dd6f1f35d30cd84dca0a8f1d2e1f7837071233d48b33
+size 72798422