Datasets:
audio dict | transcription stringclasses 1
value | label stringclasses 1
value |
|---|---|---|
{
"bytes": [
82,
73,
70,
70,
36,
125,
0,
0,
87,
65,
86,
69,
102,
109,
116,
32,
16,
0,
0,
0,
1,
0,
1,
0,
128,
62,
0,
0,
0,
125,
0,
0,
2,
0,
16,
0,
100,
97,
116,
... | ا | alif |
{
"bytes": [
82,
73,
70,
70,
36,
125,
0,
0,
87,
65,
86,
69,
102,
109,
116,
32,
16,
0,
0,
0,
1,
0,
1,
0,
128,
62,
0,
0,
0,
125,
0,
0,
2,
0,
16,
0,
100,
97,
116,
... | ا | alif |
{
"bytes": [
82,
73,
70,
70,
36,
125,
0,
0,
87,
65,
86,
69,
102,
109,
116,
32,
16,
0,
0,
0,
1,
0,
1,
0,
128,
62,
0,
0,
0,
125,
0,
0,
2,
0,
16,
0,
100,
97,
116,
... | ا | alif |
{
"bytes": [
82,
73,
70,
70,
36,
125,
0,
0,
87,
65,
86,
69,
102,
109,
116,
32,
16,
0,
0,
0,
1,
0,
1,
0,
128,
62,
0,
0,
0,
125,
0,
0,
2,
0,
16,
0,
100,
97,
116,
... | ا | alif |
{"bytes":"UklGRiR9AABXQVZFZm10IBAAAAABAAEAgD4AAAB9AAACABAAZGF0YQB9AAD7/woA+v8IABUA9v8JAAIA1f/9/zAAGg(...TRUNCATED) | ا | alif |
{"bytes":"UklGRiR9AABXQVZFZm10IBAAAAABAAEAgD4AAAB9AAACABAAZGF0YQB9AAAHAP3/BQAFAA4A+//w/wYAEAAXAPv/DA(...TRUNCATED) | ا | alif |
{"bytes":"UklGRiR9AABXQVZFZm10IBAAAAABAAEAgD4AAAB9AAACABAAZGF0YQB9AADr//D/GgDm/+//+f/1//j/AADt//v/9P(...TRUNCATED) | ا | alif |
{"bytes":"UklGRiR9AABXQVZFZm10IBAAAAABAAEAgD4AAAB9AAACABAAZGF0YQB9AAAMAP//FgDd/+3/+f/X/+r/6v/5/xoAAA(...TRUNCATED) | ا | alif |
{"bytes":"UklGRiR9AABXQVZFZm10IBAAAAABAAEAgD4AAAB9AAACABAAZGF0YQB9AAAKAOj/JAAfAOf/CQD//+3/8v8DAAEA9f(...TRUNCATED) | ا | alif |
{"bytes":"UklGRiR9AABXQVZFZm10IBAAAAABAAEAgD4AAAB9AAACABAAZGF0YQB9AADg//L/8//6/+X/AgDe/wsAEQAiAAMA5f(...TRUNCATED) | ا | alif |
End of preview. Expand in Data Studio
Alif Dataset — Arabic Letter Speech Recognition
Single-letter Arabic speech dataset containing recordings of the letter Alif (ا).
Dataset Details
| Property | Value |
|---|---|
| Language | Arabic (ar) |
| Audio format | WAV, 16 kHz, mono |
| Duration per sample | ~1 second |
| Total samples | 78 |
| Train / Val / Test | 62 / 8 / 8 |
| Transcription | ا |
| Label | alif |
Usage
from datasets import load_dataset, Audio
ds = load_dataset("YOUR_HF_USERNAME/alif-dataset")
# Cast audio column to 16 kHz
ds = ds.cast_column("audio", Audio(sampling_rate=16_000))
Note:
metadata.csvcontains asplitcolumn (train/validation/test).
Filter by split after loading, or use the prepare script in the accompanying Colab notebook.
Intended Use
Fine-tune openai/whisper-small (or any Whisper variant) for Arabic letter / isolated-word recognition.
Citation
@dataset{alif_dataset_2026,
author = {Masum Billah},
title = {Alif Arabic Letter Speech Dataset},
year = {2026},
publisher = {Hugging Face},
}
- Downloads last month
- 298