Update README.md

README.md

@@ -4,9 +4,9 @@ configs:
 - config_name: 22khz
   data_files:
   - split: utterancces
-    path: "manifest_22khz.json"
+    path: "22khz/manifest_22khz.json"
   - split: chapters
-    path: "chapters_22khz.json"
+    path: "22khz/chapters_22khz.json"
 ---
 
 # HiFiTTS-2: A Large-Scale High Bandwidth Speech Dataset
@@ -21,6 +21,14 @@ padding: 0;
 
 ## Dataset Description
 
+This repository contains the metadata for HiFiTTS-2, a large scale speech dataset derived from LibriVox audiobooks. For more details, please refer to our paper. [TODO: Add ArXiv link]
+
+The dataset contains metadata for approximately 36.7k hours of audio that can be downloaded from LibriVox at a 48 kHz sampling rate.
+
+The metadata contains estimated bandwidth, which can be used to infer the original sampling rate the audio was recorded at. The base dataset is filtered for a bandwidth appropriate for training speech models at *22 kHz*. We also provide a precomputed subset with 31.7k hours appropraite for *44 kHz* training. Users can modify the download script to use any sampling rate and bandwidth threshold which might be more appropriate for their work.
+
+Due to legal and ethical considerations, we do not redistribute LibriVox audiobooks on Hugging Face. All audio in the dataset can be downloaded from LibriVox, following the instructions below.
+
 ## Download Instructions
 
 1. Download the *manifet.json* file and *chapter.json* files corresponding to your desired sampling rate from this Hugging Face repository. Copy these into a workspace directory (in this example */home/hifitts2*).
@@ -50,8 +58,6 @@ python /home/NeMo-speech-data-processor/main.py \
     max_workers=8
 ```
 
-Due to legal and ethical considerations, we do not redistribute LibriVox audiobooks to download from Hugging Face.
-
 ## Dataset Owner(s)
 
 NVIDIA Corporation