Update README.md

README.md

@@ -1,4 +1,4 @@
-readme_content = """---
+---
 dataset_info:
   features:
     - name: audio
@@ -8,18 +8,35 @@ dataset_info:
   splits:
     - name: train
       num_examples: 1984
+download_size: 1790000000
+dataset_size: 1790000000
 tags:
   - tts
+  - speech
+  - audio
   - tunisian-arabic
   - derja
-language: ar-TN
+  - arabic
+language:
+  - ar
+language_bcp47:
+  - ar-tn
+license: mit
+pretty_name: TunArTTS Mini HF
 ---
 
-Tunisian Arabic TTS Dataset (mini) from DerjaNinja.
-- 1984 audio clips
-- Transcriptions in Tunisian Arabic
-"""
-with open("/content/tunartts_mini_hf/README.md", "w") as f:
-    f.write(readme_content)
+# TunArTTS Mini HF
+
+A mini Tunisian Arabic (Derja) TTS dataset scraped from DerjaNinja.
+
+- **1984 high-quality audio clips** with human pronunciations
+- Full transcriptions in Tunisian Arabic script
+- Sampling rate: 22050 Hz (standard for many TTS models)
+- Perfect for fine-tuning TTS models (e.g. YourTTS, VITS, Bark, Tortoise)
+
+### Columns
+- `audio`: the audio file
+- `text`: the transcription in Tunisian Arabic
 
-# Dann upload_folder wie vorher
+Source: DerjaNinja (public domain audio recordings)
+License: MIT (permissive for research and commercial use)