docs: rename + Vaani attribution

README.md

@@ -47,7 +47,7 @@ This dataset trains and evaluates the paired ASR model
 — a character-level CTC adapter over a frozen
 [`facebook/mms-1b-all`](https://huggingface.co/facebook/mms-1b-all) encoder.
 An augmented training variant is available at
-[`sulabhkatiyar/ne-asr-nri-aug`](https://huggingface.co/datasets/sulabhkatiyar/ne-asr-nri-aug).
+[`sulabhkatiyar/ne-asr-dataset-nri-aug`](https://huggingface.co/datasets/sulabhkatiyar/ne-asr-dataset-nri-aug).
 
 Part of the **[NE Speech AI V1](https://huggingface.co/collections/sulabhkatiyar/ne-speech-ai-v1-6a8efa7593551e32d63b876b)** collection.
 
@@ -59,4 +59,4 @@ Part of the **[NE Speech AI V1](https://huggingface.co/collections/sulabhkatiyar
 
 ## License
 
-`cc-by-4.0`. Please retain attribution and comply with the license terms when using this dataset.
+`cc-by-4.0`, following the ARTPARK-IISc **Vaani** source dataset (https://vaani.iisc.ac.in/), from which this data is derived. Please retain that attribution and comply with the license terms when using this dataset.