Update dataset card: Refine metadata, add paper and GitHub links, and add relevant tags

This PR improves the Balalaika dataset card by:
- Correcting the license in the metadata to `cc-by-nc-nd-4.0`, aligning with the dataset's specific license as stated in the content.
- Adding a direct link to the associated paper on Hugging Face Papers.
- Adding a link to the GitHub repository for easier access.
- Including relevant tags (`speech-synthesis`, `speech-enhancement`, `russian`) to enhance discoverability, based on the paper's abstract and the dataset's language.

README.md

@@ -1,14 +1,21 @@
 ---
-license: cc-by-nc-sa-4.0
 language:
 - ru
+license: cc-by-nc-nd-4.0
 task_categories:
 - text-to-speech
 pretty_name: Balalaika
+tags:
+- speech-synthesis
+- speech-enhancement
+- russian
 ---
 
 # A Data-Centric Framework for Addressing Phonetic and Prosodic Challenges in Russian Speech Generative Models
 
+Paper: [A Data-Centric Framework for Addressing Phonetic and Prosodic Challenges in Russian Speech Generative Models](https://huggingface.co/papers/2507.13563)
+GitHub: [https://github.com/mtuciru/balalaika](https://github.com/mtuciru/balalaika)
+
 Russian speech synthesis presents distinctive challenges, including vowel reduction, consonant devoicing, variable stress patterns, homograph ambiguity, and unnatural intonation. This paper introduces Balalaika, a novel dataset comprising more than 2,000 hours of studio-quality Russian speech with comprehensive textual annotations, including punctuation and stress markings. Experimental results show that models trained on Balalaika significantly outperform those trained on existing datasets in both speech synthesis and enhancement tasks.
 
 ---