Datasets:
Fix audio schema to struct<bytes,path> + add text transcripts + HF feature metadata
Browse files- README.md +8 -6
- data/shard-000000.parquet +2 -2
- data/shard-000001.parquet +2 -2
- data/shard-000002.parquet +2 -2
- data/shard-000003.parquet +2 -2
- manifest.json +2 -2
README.md
CHANGED
|
@@ -47,7 +47,8 @@ A curated pool of **2,000 English reference voices**, each a single 12–30 seco
|
|
| 47 |
| `sub_pool` | string | Internal grouping (`controller`) |
|
| 48 |
| `gender_hint` | string | Null for this release |
|
| 49 |
| `language` | string | BCP-47 language tag (`en`) |
|
| 50 |
-
| `
|
|
|
|
| 51 |
|
| 52 |
## Selection procedure
|
| 53 |
|
|
@@ -64,10 +65,11 @@ from datasets import load_dataset
|
|
| 64 |
|
| 65 |
ds = load_dataset("twangodev/radiotalk-voices-2k", split="train")
|
| 66 |
row = ds[0]
|
| 67 |
-
print(row["voice_id"], row["duration_s"])
|
| 68 |
-
#
|
| 69 |
-
|
| 70 |
-
|
|
|
|
| 71 |
```
|
| 72 |
|
| 73 |
## License & attribution
|
|
@@ -85,4 +87,4 @@ If you use this dataset, please cite the upstream corpus:
|
|
| 85 |
}
|
| 86 |
```
|
| 87 |
|
| 88 |
-
See `NOTICE.md` for the full source attribution.
|
|
|
|
| 47 |
| `sub_pool` | string | Internal grouping (`controller`) |
|
| 48 |
| `gender_hint` | string | Null for this release |
|
| 49 |
| `language` | string | BCP-47 language tag (`en`) |
|
| 50 |
+
| `text` | string | Normalized transcript from LibriTTS-R (`text_normalized`) |
|
| 51 |
+
| `audio` | Audio | FLAC-encoded mono at 24 kHz (Hub viewer plays it inline) |
|
| 52 |
|
| 53 |
## Selection procedure
|
| 54 |
|
|
|
|
| 65 |
|
| 66 |
ds = load_dataset("twangodev/radiotalk-voices-2k", split="train")
|
| 67 |
row = ds[0]
|
| 68 |
+
print(row["voice_id"], row["duration_s"], row["text"][:60])
|
| 69 |
+
# Audio is already decoded by `datasets` into {array, sampling_rate, path}
|
| 70 |
+
audio = row["audio"]
|
| 71 |
+
import soundfile as sf
|
| 72 |
+
sf.write("out.wav", audio["array"], audio["sampling_rate"])
|
| 73 |
```
|
| 74 |
|
| 75 |
## License & attribution
|
|
|
|
| 87 |
}
|
| 88 |
```
|
| 89 |
|
| 90 |
+
See `NOTICE.md` for the full source attribution.
|
data/shard-000000.parquet
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1245ff344a35ec70fd983cca6254cdead0a103db7065b3852ec6f36f80deb819
|
| 3 |
+
size 276411499
|
data/shard-000001.parquet
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ee508e14a0e1d7cbe0e21108e0eeaff8cc48dde1789deb27959cbff9d94cfbc0
|
| 3 |
+
size 276799628
|
data/shard-000002.parquet
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:33aa295f2c2925e5bc43e3941c05b7abfb6908cd642eb578b48d4c05ece9a640
|
| 3 |
+
size 279608627
|
data/shard-000003.parquet
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f496dd85f6d58228484184301ab9dce9b286015756b5fcbc3c7c1e3a4bf6cf8b
|
| 3 |
+
size 279796367
|
manifest.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"total_rows": 2000,
|
| 3 |
"last_shard_index": 3,
|
| 4 |
-
"started_at": "2026-04-
|
| 5 |
-
"updated_at": "2026-04-
|
| 6 |
"seed": 42,
|
| 7 |
"target": 2000,
|
| 8 |
"sources": [
|
|
|
|
| 1 |
{
|
| 2 |
"total_rows": 2000,
|
| 3 |
"last_shard_index": 3,
|
| 4 |
+
"started_at": "2026-04-24T03:02:35+00:00",
|
| 5 |
+
"updated_at": "2026-04-24T03:03:56+00:00",
|
| 6 |
"seed": 42,
|
| 7 |
"target": 2000,
|
| 8 |
"sources": [
|