Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -37,10 +37,38 @@ size_categories:
|
|
| 37 |
>
|
| 38 |
> To license the full **9,773.2-hour commercial vault** ($29,500 flat enterprise fee; zero royalties; instant AWS S3 / GCP bucket transfer), proceed to our live portal:
|
| 39 |
>
|
| 40 |
-
> 👉 [CLICK HERE TO LICENSE THE FULL 9,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 41 |
|
| 42 |
---
|
| 43 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 44 |
## 📊 Full Dataset Overview (9,773.2 Total Hours)
|
| 45 |
|
| 46 |
The **Taskmaster OmniVox 10K Vault** is a multi-modal, commercially cleared speech dataset spanning **60+ global languages and regional dialects**, optimized for enterprise ASR, TTS voice cloning, and code-switching models.
|
|
|
|
| 37 |
>
|
| 38 |
> To license the full **9,773.2-hour commercial vault** ($29,500 flat enterprise fee; zero royalties; instant AWS S3 / GCP bucket transfer), proceed to our live portal:
|
| 39 |
>
|
| 40 |
+
> 👉 [CLICK HERE TO LICENSE THE FULL 9,773.2-HOUR VAULT ($29,500)](https://buy.stripe.com/7sY28sb3Baaud82czX9oc00)
|
| 41 |
+
---
|
| 42 |
+
|
| 43 |
+
## 📊 Master Manifest Schema Preview
|
| 44 |
+
|
| 45 |
+
All 9,773.2+ hours follow this standardized JSON/CSV manifest schema:
|
| 46 |
+
|
| 47 |
+
| Column Name | Data Type | Example Value | Description |
|
| 48 |
+
| :--- | :--- | :--- | :--- |
|
| 49 |
+
| `file_name` | string | `fleurs_cmn_hans_cn_train_00001.wav` | Relative path to 16 kHz / 16-bit WAV |
|
| 50 |
+
| `transcript` | string | `北京是中华人民共和国 offered 的首都` | Normalized transcription |
|
| 51 |
+
| `locale` | string | `cmn_hans_cn` | BCP-47 language tag |
|
| 52 |
+
| `language` | string | `Mandarin_China` | Full primary language name |
|
| 53 |
+
| `sample_rate` | integer | `16000` | Standardized audio sample rate (Hz) |
|
| 54 |
+
| `license_status` | string | `CC-BY 4.0` | Verified commercial clearance tag |
|
| 55 |
+
| `attribution_required` | string | `Google LLC / FLEURS Corpus` | Provenance / legal attribution string |
|
| 56 |
|
| 57 |
---
|
| 58 |
|
| 59 |
+
## 🔊 Verified Audio Previews
|
| 60 |
+
|
| 61 |
+
<details>
|
| 62 |
+
<summary><b>Click to expand audio samples</b></summary>
|
| 63 |
+
|
| 64 |
+
### Mandarin Chinese (`cmn_hans_cn`) — CC-BY 4.0
|
| 65 |
+
<audio controls src="https://huggingface.co/datasets/TaskmasterVirtualAlly/Taskmaster-OmniVox-9K-Vault-Preview/resolve/main/samples/mandarin_sample.wav"></audio>
|
| 66 |
+
|
| 67 |
+
### Standard Arabic (`ar_eg`) — CC-BY 4.0
|
| 68 |
+
<audio controls src="https://huggingface.co/datasets/TaskmasterVirtualAlly/Taskmaster-OmniVox-9K-Vault-Preview/resolve/main/samples/arabic_sample.wav"></audio>
|
| 69 |
+
|
| 70 |
+
</details>
|
| 71 |
+
|
| 72 |
## 📊 Full Dataset Overview (9,773.2 Total Hours)
|
| 73 |
|
| 74 |
The **Taskmaster OmniVox 10K Vault** is a multi-modal, commercially cleared speech dataset spanning **60+ global languages and regional dialects**, optimized for enterprise ASR, TTS voice cloning, and code-switching models.
|