---
language:
  - fa
task_categories:
  - text-to-speech
  - automatic-speech-recognition
---

# filimo-farsi

Persian speech dataset. Audio resampled to **16000 Hz**.

## Splits

| Split | Samples | Total Duration (H:MM:SS) | Volume |
|-------|--------:|:------------------------:|-------:|
| train | 118206 | 85:17:53 | 1.546 GB |
| test | 16788 | 12:02:12 | 0.218 GB |
| **total** | **134994** | **97:20:05** | **1.764 GB** |