Neyshekar V3 Asr Aligned: A Curated Persian Speech Recognition Dataset

A repaired subset of the Neyshekar v3 dataset for Persian automatic speech recognition. The dataset contains real audio clips and transcripts, curated by matching multiple ASR hypotheses back to the original transcript pool to ensure alignment. It was created by Peacockery and last updated on 2026-05-13.

Use Cases

• Training Persian ASR models based on real audio and transcript pairs.
• Evaluating ASR model performance using a curated candidate set.
• Fine-tuning speech recognition systems on verified Persian speech data.

Strengths

• Contains real audio and real transcripts, as stated in the description.
• Curated to ensure alignment between audio clips and their transcripts.

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.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

huggingface

Collection Method

Repaired subset of Neyshekar v3, with transcripts recovered by matching multiple ASR hypotheses.

Freshness

Last updated 2026-05-13 03:19:40; freshness should be verified.