Update README.md

README.md

@@ -14,14 +14,9 @@ size_categories:
 ## 📂 Dataset Structure
 The dataset is organized into three folders, with **4,000 audio files per model**:
 
-- `dia2/`: Generated using **Dia2**.
-- `maya1/`: Generated using **Maya1**.
-- `melotts/`: Generated using **MeloTTS**.
-
-## 🤖 Source Models
-- **Dia2**: [https://huggingface.co/nari-labs/Dia2-2B/tree/main]
-- **Maya1**: [https://huggingface.co/maya-research/maya1]
-- **MeloTTS**: [https://huggingface.co/myshell-ai/MeloTTS-English]
+- `dia2/`: Generated using [nari-labs/Dia2-2B](https://huggingface.co/nari-labs/Dia2-2B/tree/main)
+- `maya1/`: Generated using [maya-research/maya1](https://huggingface.co/maya-research/maya1)
+- `melotts/`: Generated using [myshell-ai/MeloTTS-English](https://huggingface.co/myshell-ai/MeloTTS-English)
 
 ## 🎯 Use Cases
 - **AI Audio Detection:** Training classifiers to detect synthetic voices.
@@ -29,6 +24,6 @@ The dataset is organized into three folders, with **4,000 audio files per model*
 - **TTS Evaluation:** Benchmarking different architectures on conversational text.
 
 ## 📜 Citation
-If you use this dataset, please cite this repository and the original model creators.
+If you use this dataset, the [DailyDialog](https://huggingface.co/datasets/roskoN/dailydialog) please cite this repository and the original model creators.
 
 This project was provided with computing AI and storage resources by GENCI at IDRIS thanks to the grant 2025-AD011016076 on the supercomputer Jean Zay's V100 partition .