/jvs_huggingface_datasets

Primary LanguageJupyter Notebook

JVS Corups for ๐Ÿค— datasets

Preparation

  1. Download the JVS Corpus from the official web page and extract it to this directory
  2. Run fix.sh to fix wrong file names
    • If zip file you downloaded is md5 hash: 2987778b0ee830914bfebb97783d0c3e
  3. Run prepare.py to generate all_transcripts.txt file
  4. Add from jvs_datasets import jvs_datasets in your python scripts
    • Need to include this directory in your PYTHONPATH

Usage

In [1]: from jvs_datasets import jvs_datasets
Using custom data configuration default-f40f9d93b88f3f56
Reusing dataset csv (/Users/yag_ays/.cache/huggingface/datasets/csv/default-f40f9d93b88f3f56/0.0.0/6b9057d9e23d9d8a2f05b985917a0da84d70c5dae3d22ddd8a3f22fb01c69d9e)
100%|โ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆ| 1/1 [00:00<00:00, 508.09it/s]

In [2]: jvs_datasets[0]
Out[2]:
{'file_id': 'VOICEACTRESS100_001',
 'sentence': 'ใพใŸใ€ๆฑๅฏบใฎใ‚ˆใ†ใซใ€ไบ”ๅคงๆ˜Ž็Ž‹ใจๅ‘ผใฐใ‚Œใ‚‹ใ€ไธป่ฆใชๆ˜Ž็Ž‹ใฎ**ใซ้…ใ•ใ‚Œใ‚‹ใ“ใจใ‚‚ๅคšใ„ใ€‚',
 'path': '/Users/yag_ays/dev/speech-to-text/jvs_datasets/jvs_ver1/jvs096/parallel100/wav24kHz16bit/VOICEACTRESS100_001.wav'}