README.md

metadata

license: cc-by-4.0
tags:
  - audio-diarization
  - speaker-segmentation
language:
  - kk

Diarizations Dataset

Aggregated speaker segmentation outputs.

Videos

ecx7ywj89m4

• Segments: 54
• Speakers: SPEAKER_00, SPEAKER_01
• Duration: 2369.59s

Config

• Trimmed: none
• Diarize model: pyannote/speaker-diarization
• ASR model: akuzdeuov/whisper-base.kk chunk 30s, lang kk
• Batch: 16

FWmr-zrGK_w

• Segments: 206
• Speakers: SPEAKER_00, SPEAKER_01, SPEAKER_02
• Duration: 3071.74s

Config

• Trimmed: none
• Diarize model: pyannote/speaker-diarization
• ASR model: akuzdeuov/whisper-base.kk chunk 30s, lang kk
• Batch: 16

FzeUEbA6j4E

• Segments: 535
• Speakers: SPEAKER_00, SPEAKER_01, SPEAKER_02, SPEAKER_03
• Duration: 4425.93s

Config

• Trimmed: none
• Diarize model: pyannote/speaker-diarization
• ASR model: akuzdeuov/whisper-base.kk chunk 30s, lang kk
• Batch: 16

WC_8foLpFbE

• Segments: 158
• Speakers: SPEAKER_00, SPEAKER_01, SPEAKER_02
• Duration: 3605.37s

Config

• Trimmed: none
• Diarize model: pyannote/speaker-diarization
• ASR model: akuzdeuov/whisper-base.kk chunk 30s, lang kk
• Batch: 16

Usage

from datasets import load_dataset
ds = load_dataset('pushthetempo/diarizations')
print(ds)