Upload 2 files

dataset_info.json

@@ -1,5 +1,6 @@
 {
-  "description": "This dataset is designed for speech-to-text tasks and contains audio files stored in tar archives along with corresponding transcript files in TSV format. The data is organized for the Uzbek language and split into train, test, and validation sets.",
+  "description": "This dataset is designed for speech-to-text (STT) tasks with a focus on the Uzbek language. It contains audio recordings stored as tar archives along with their corresponding transcript files in TSV format.",
+  "citation": "@misc{dataset_stt2025, title={Dataset_STT}, author={Your Name}, year={2025}}",
   "homepage": "https://huggingface.co/datasets/Elyordev/Dataset_STT",
   "license": "MIT",
   "splits": {
@@ -20,13 +21,16 @@
     }
   },
   "features": {
+    "id": "string",
     "audio": {
-      "dtype": "string",
-      "description": "Audio files are stored as tar archives containing the actual audio data."
+      "type": "audio",
+      "sampling_rate": 16000
     },
-    "transcript": {
-      "dtype": "string",
-      "description": "Transcript files in TSV format containing the corresponding transcriptions for the audio data."
-    }
+    "sentence": "string",
+    "duration": "float",
+    "age": "string",
+    "gender": "string",
+    "accents": "string",
+    "locale": "string"
   }
 }

readme.md

@@ -22,27 +22,40 @@ Dataset_STT/
 │   │   └── validation/
 │   │       └── validation.tsv
 
+
 ## Files Description
 
-- **Audio Files (`*.tar`):**  
-  The audio files are provided as tar archives containing the actual audio data.
-  - `audio/uz/train/train.tar`: Training audio files.
-  - `audio/uz/test/test.tar`: Testing audio files.
-  - `audio/uz/validation/validation.tar`: Validation audio files.
+- **Audio Files:**  
+  The audio files are stored as tar archives for each data split (train, test, and validation). Each tar archive contains the actual audio recordings (e.g., in MP3 format).
+
+- **Transcript Files:**  
+  The transcript files are provided in TSV format with the following columns:
+  - `id`
+  - `path` (the filename of the audio file within the tar archive)
+  - `sentence` (the transcription)
+  - `duration` (audio duration in seconds)
+  - `age`
+  - `gender`
+  - `accents`
+  - `locale`
 
-- **Transcript Files (`*.tsv`):**  
-  The transcript files are in TSV (Tab-Separated Values) format and contain the text transcriptions corresponding to the audio data.
-  - `transcript/uz/train/train.tsv`: Training transcripts.
-  - `transcript/uz/test/test.tsv`: Testing transcripts.
-  - `transcript/uz/validation/validation.tsv`: Validation transcripts.
+- **Custom Loader (`dataset_stt.py`):**  
+  This script extracts audio files from the tar archives and pairs them with their metadata from the transcript TSV files. It returns the audio data using the `datasets.Audio` feature (with a sampling rate of 16000 Hz), which enables interactive playback in the Hugging Face dataset viewer.
 
 ## How to Load the Dataset
 
-You can load this dataset using the Hugging Face [datasets](https://huggingface.co/docs/datasets/) library. For example:
+You can load the dataset using the Hugging Face `datasets` library. For example:
 
 ```python
 from datasets import load_dataset
 
-# Load the dataset from the Hub
-dataset = load_dataset("Elyordev/Dataset_STT")
-print(dataset)
+data_files = {
+    "train": {"audio": "audio/uz/train/train.tar", "transcript": "transcript/uz/train/train.tsv"},
+    "test": {"audio": "audio/uz/test/test.tar", "transcript": "transcript/uz/test/test.tsv"},
+    "validation": {"audio": "audio/uz/validation/validation.tar", "transcript": "transcript/uz/validation/validation.tsv"}
+}
+
+dataset = load_dataset("Elyordev/Dataset_STT", data_files=data_files)
+print(dataset)
+
+