Loading…
![]() String updated in the repository |
|
![]() String updated in the repository |
|
![]() String updated in the repository |
|
![]() String updated in the repository |
|
Things to check
English | Chinese (Traditional) | ||
---|---|---|---|
Scripty | Glossary | ||
This is the name of the bot, it should not be translated as a proper noun |
Key
data-storage-embed-descriptionSource string description
data_storage commandFlags
fluent-parts, fluent-references, fluent-source-inner-html, fluent-source-syntax, fluent-target-inner-html, fluent-target-syntax, fluent-type:Message, ignore-fluent-target-inner-html, ignore-xml-invalid, ignore-xml-tags
That said, here goes.
Scripty requires a lot of audio and text data to train a proper speech-to-text model. Not everyone is able to donate or buy premium to help us out, so a big way you can help out is by allowing us to store your data like audio and messages for training a model.
We understand this data can be extremely personal, so this is entirely opt-in and will not affect your experience in any way.
Here's what we'd do with it:
{ "*" } With stored messages, we would feed them into a scorer targeted to your language. This scorer would allow the algorithm to select the most likely words for a given set of sounds. Although immensely helpful, this isn't as important as audio. Note that this message data is encrypted with AES 256-bit encryption.
{ "*" } With stored audio, we would feed it and the transcript of it into a model to increase the accuracy of the speech-to-text model. This is insanely helpful, even if you have a poor microphone and lots of background noise: in fact, the more noise, the better, as long as a human can still make out what you are saying.
If you are opted in, and you decide later to opt out, your data is still stored, but you can request deletion of your voice data by running `{ $contextPrefix }delete_all_data`. However, it is impossible to delete your message data. This is because we do not store a link of what user sent what message.
Your data is stored on servers that are locked down tightly. It would be extremely difficult for anyone attempting to gain access to successfully do so.
You can toggle your choices using the below buttons.
內容如下。
Scripty 需要大量音頻和文本數據來訓練正確的語音到文本模型。 並非每個人都能夠捐贈或購買高級等級來幫助我們,因此你可以提供幫助的一個重要方式是允許我們存儲你的數據,例如用於訓練模型的音頻和信息。
我們了解這些數據可能非常個人,因此這完全是選擇性的,不會以任何方式影響你的體驗。
下面是我們將如何使用它:
{ "*" } 通過存儲的信息,我們會將它們提供給針對你的語言的記分器。 該記分器將允許算法為一組音頻選擇最可能的單詞。 儘管非常有用,但這並不像音頻那麼重要。
{ "*" } 使用存儲的音頻,我們會將其及其轉錄內容輸入到模型中,以提高語音到文本模型的準確性。 這非常有用,即使你的麥克風很差並且背景噪音很大:事實上,噪音越多越好,只要人類仍然可以聽清你在說什麼。
如果你選擇加入,並且稍後決定退出,你的數據仍會被存儲,但你可以通過聯繫支持服務器中的核心開發人員請求刪除:{ $supportServerInvite }。 我們將永久擦除你的所有數據。
你的數據存儲在核心開發人員擁有的硬件上,並被鎖定。 對於任何試圖成功訪問的人來說,這將是極其困難的。
你可以使用以下按鈕切換你的選擇。