Nav Train Gnjr S T T: Persian Speech Audio, 221,455 Samples

221,455 audio samples of Persian speech, totaling 753 hours and 52 minutes of audio. The dataset was uploaded by user 'argoveziriii' to Hugging Face and was last updated on May 25, 2026. Audio files have been resampled to 16000 Hz.

Use Cases

• Train Persian speech recognition models based on the large volume of audio samples.
• Fine-tune speech-to-text systems on the provided train and test splits.
• Benchmark audio processing pipelines using the 16000 Hz resampled audio format.
• Develop voice activity detection or speaker diarization models for Persian speech.

Strengths

• Contains 221,455 audio samples, providing a substantial data volume.
• Offers a total duration of 753 hours and 52 minutes of speech.
• Includes a predefined split of 160,931 training and 60,524 test samples.
• Audio is preprocessed to a standard 16000 Hz sampling rate.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• The source, collection method, and speaker demographics are unknown, which may indicate bias.

Provenance

Source

huggingface

Freshness

Last updated 2026-05-25 10:23:07; freshness should be verified.