Upload README.md with huggingface_hub

README.md

@@ -10,7 +10,9 @@ size_categories:
 
 # CTIM-Bench: Benchmark for Traditional Chinese Music Generation
 
-**CTIM-Bench** is a curated evaluation dataset consisting of 500 text-audio pairs, specifically designed to benchmark music generation models on traditional Chinese instruments. It serves as the test set for the **CTIM-Gen** project.
+**CTIM-Bench** is a curated evaluation dataset consisting of **500 expert-validated high-fidelity samples**, specifically designed to benchmark music generation models on traditional Chinese instruments. It serves as the test set for the **CTIM-Gen** project.
+
+The dataset features **42.6% minority ethnic instruments** (e.g., Matouqin, Hulusi), reflecting the rich diversity of Chinese musical culture.
 
 ## ⚖️ Data Source & Usage Policy
 
@@ -22,4 +24,4 @@ The audio data in this benchmark was collected from [China Music Network (www.ch
 
 - **audio/**: Ground truth audio files (30s clips).
 - **test_prompts.json**: Metadata containing prompts and file references.
-- **Instruments**: Guqin, Guzheng, Matouqin, Hulusi, Pipa.
+- **Instruments**: Guqin, Guzheng, Matouqin, Hulusi, Pipa, Erhu, Yangqin.