README.md

metadata

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.