---
license: cc-by-4.0
task_categories:
- automatic-speech-recognition
language:
- tr
pretty_name: FLEURS (google/fleurs, tr_tr) — Turkish test split
size_categories:
- 1K<n<10K
---

# fleurs-tr-test

Turkish **test** split of **FLEURS (google/fleurs, tr_tr)**, re-hosted for Turkish STT benchmarking.

- Rows: **743**
- Columns: `id, num_samples, path, audio, transcription, raw_transcription, gender, lang_id, language, lang_group_id`
- Source: https://huggingface.co/datasets/google/fleurs
- License: `cc-by-4.0` (inherited from source)

Only the Turkish `test` split is included, extracted as-is from the source dataset.