Nkonya_nko: 10.01 h voice-converted (6,103 clips)

Nkonya_nko/{train-00000-of-00004.parquet → train-00000-of-00005.parquet}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:98a4729d28c24a3ac7bddf6a251f73894a7534425309b7b52f0abb20b7c09272
-size 162350096
+oid sha256:7f97d28083ca5564b6c206d7a6447771a56c05dc0b077a75c05a5494b1e38b1f
+size 148727669

Nkonya_nko/{train-00001-of-00004.parquet → train-00001-of-00005.parquet}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6ed4e458942740c2241add51c472cc534360ee386fa26f280010e9178acc96d7
-size 168161375
+oid sha256:a02b4358eb757f71262a246fe6a6673d390c7dcf7e52db7ba3df446c25c9c9b8
+size 145352342

Nkonya_nko/{train-00002-of-00004.parquet → train-00002-of-00005.parquet}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c114d673fe5ed36abc12f071bb77f54d913de3f92372541fb60d350cf8ec5a95
-size 168811890
+oid sha256:0a1b87fc39493f1a92d34a3e4a9e722f6762e501d890d4ed445590fda36103b0
+size 148305938

Nkonya_nko/{train-00003-of-00004.parquet → train-00003-of-00005.parquet}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:443fccc010c47d0177d8c95e9a3338b97d896175d94521e710e6f147a2b6a861
-size 164961797
+oid sha256:350688ed45fccacb480f089af4ade8d6ea2ad2f1e4602b1e7dabd318dcf9e782
+size 149912468

Nkonya_nko/train-00004-of-00005.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e55069b6d1a76c426db42d6f1fbc8ec67bd1c3d8defe5366b73a9459eb61323d
+size 146286066

README.md

@@ -373,10 +373,10 @@ dataset_info:
   features: *id001
   splits:
   - name: train
-    num_bytes: 664285158
-    num_examples: 5500
-  download_size: 664285158
-  dataset_size: 664285158
+    num_bytes: 738584483
+    num_examples: 6103
+  download_size: 738584483
+  dataset_size: 738584483
 - config_name: Nzema_nzi
   features: *id001
   splits:
@@ -602,9 +602,9 @@ configs:
 
 # Ghana Speech - Female 22kHz Voice Converted (Scripture / Bible)
 
-This dataset contains **381.02 hours** of natural Ghanaian speech converted into a high-fidelity African female voice at **22.05 kHz**, derived from scripture and Bible recordings in **[ghananlpcommunity/ghana-speech](https://huggingface.co/datasets/ghananlpcommunity/ghana-speech)**.
+This dataset contains **382.04 hours** of natural Ghanaian speech converted into a high-fidelity African female voice at **22.05 kHz**, derived from scripture and Bible recordings in **[ghananlpcommunity/ghana-speech](https://huggingface.co/datasets/ghananlpcommunity/ghana-speech)**.
 
-- **Total Scale**: **381.02 hours** across **35 Ghanaian language subsets** (178,495 clips, 26.73 GB).
+- **Total Scale**: **382.04 hours** across **35 Ghanaian language subsets** (179,098 clips, 26.80 GB).
 - **Target Voice**: `clear-high-slow` (native Twi female voice, pitch ~261 Hz, clear timbre, lively intonation) from **[AfriSpeech/africa-female-vc](https://github.com/AfriSpeech/africa-female-vc)**.
 - **Audio Format**: **22,050 Hz (22.05 kHz)** mono 16-bit FLAC.
 - **Model**: Seed-VC (`seed-uvit-whisper-small-wavenet`) + NVIDIA BigVGAN 22kHz neural vocoder.
@@ -640,7 +640,7 @@ This dataset contains **381.02 hours** of natural Ghanaian speech converted into
 | Mampruli | `Mampruli_maw` | `maw` | 3,847 | 10.00h | 5 | 672.4 MB |
 | Nawuri | `Nawuri_naw` | `naw` | 4,789 | 10.00h | 5 | 685.6 MB |
 | Ninkare (Frafra / Gurene) | `Ninkare_gur` | `gur` | 4,946 | 10.00h | 5 | 668.6 MB |
-| Nkonya | `Nkonya_nko` | `nko` | 5,500 | 8.99h | 4 | 633.5 MB |
+| Nkonya | `Nkonya_nko` | `nko` | 6,103 | 10.01h | 5 | 704.4 MB |
 | Nzema | `Nzema_nzi` | `nzi` | 781 | 1.50h | 1 | 108.9 MB |
 | Sehwi | `Sehwi_sfw` | `sfw` | 7,704 | 16.20h | 8 | 1129.1 MB |
 | Sekpele | `Sekpele_lip` | `lip` | 3,809 | 10.00h | 5 | 652.8 MB |
@@ -651,9 +651,9 @@ This dataset contains **381.02 hours** of natural Ghanaian speech converted into
 | Tem | `Tem_kdh` | `kdh` | 1,000 | 2.11h | 1 | 139.8 MB |
 | Tuwuli | `Tuwuli_bov` | `bov` | 500 | 0.92h | 1 | 63.2 MB |
 | Vagla | `Vagla_vag` | `vag` | 5,060 | 10.00h | 5 | 657.2 MB |
-| **Total** | **35 languages** | — | **178,495** | **381.02h** | **164** | **26.73 GB** |
+| **Total** | **35 languages** | — | **179,098** | **382.04h** | **165** | **26.80 GB** |
 
-Most subsets hold 10-20 h. Ewe (6.7 h), Fante (6.7 h), Nkonya (9.0 h), Nzema (1.5 h), Siwu (2.5 h), Tampulma (2.5 h), Tem (2.1 h), Tuwuli (0.9 h) hold less: their conversion was stopped early, or (Ewe, Fante, Nzema) less scripture audio was converted.
+Most subsets hold 10-20 h. Ewe (6.7 h), Fante (6.7 h), Nzema (1.5 h), Siwu (2.5 h), Tampulma (2.5 h), Tem (2.1 h), Tuwuli (0.9 h) hold less: their ghana-speech scripture audio ran out (or is still being converted).
 
 ---