AsyncFakeYou

異關係

AsyncFakeyou是一個異步的Python庫，用於與假您的文本到語音API進行交互。它允許您使用Asyncio，AIOHTTP和AIOFILES以異步方式提交TTS請求，跟踪其狀態並下載所得的音頻文件。

特徵

• 異步請求偽造您的TTS API。
• 並行執行控制生成
• 用會話cookie處理用戶身份驗證

安裝

 pip install asyncfakeyou

簡單用法

您可以獲取生成的音頻文件的直接URL。

 from asyncfakeyou import AsyncAudioGen
import asyncio


async def text_to_speech ():
    gen = AsyncAudioGen ( cookies = "your_cookie_string" )
    audio_url = await gen . fetch_audio ( "model_token_here" , "sample_text" )
    print ( audio_url )


asyncio . run ( text_to_speech ())

或者，您可以將生成的音頻文件下載到指定的目錄。

 from asyncfakeyou import AsyncAudioGen
import asyncio


async def text_to_speech ():
    gen = AsyncAudioGen ( cookies = "your_cookie_string" )
    audio_url = await gen . fetch_and_save_audio ( "model_token_here" , "sample_text" ,
                                               output_path = "./audio" ,
                                               filename = "generated_audio.wav" )
    print ( audio_url )


asyncio . run ( text_to_speech ())

如何獲得餅乾？

AsyncAudioGen類別類別的cookies參數是可選的。但是，我建議設置它，因為即使沒有付費訂閱，它也會為您提供更高的隊列優先級。如果您有高級帳戶，則可以通過此API利用其所有利益。

 from asyncfakeyou import receive_cookies
import asyncio


async def get_my_cookies ():
    cookies = await receive_cookies ( "your_username_or_email" , "your_password" )
    print ( cookies )


asyncio . run ( get_my_cookies ())

並行執行

如果需要生成多個音頻文件，則可以使用fetch_tasks和fetch_and_save_tasks加快該過程。這些方法並行處理固定數量的任務（ concurrent_tasks參數），並自動重試失敗的請求。 concurrent_tasks的默認值為3，但是您可以調整此參數。

您可以通過直接URL迭代到生成的音頻文件。

 from asyncfakeyou import AsyncAudioGen
import asyncio


async def multiple_text_to_speech ():
    gen = AsyncAudioGen ( cookies = "your_cookie_string" )
    audio_tasks = [
        ( "model_token_1" , "sample_text_1" ),
        ( "model_token_2" , "sample_text_2" ),
        ( "model_token_3" , "sample_text_3" )
    ]
    async for audio_url in gen . fetch_tasks ( audio_tasks ):
        print ( audio_url )


asyncio . run ( multiple_text_to_speech ())

或者，您可以將生成的音頻文件下載到指定的目錄。

 from asyncfakeyou import AsyncAudioGen
import asyncio


async def multiple_text_to_speech ():
    gen = AsyncAudioGen ( cookies = "your_cookie_string" )
    audio_tasks = [
        ( "model_token_1" , "sample_text_1" , "filename1.wav" ),
        ( "model_token_2" , "sample_text_2" , "filename2.wav" ),
        ( "model_token_3" , "sample_text_3" , "filename3.wav" )
    ]
    await gen . fetch_and_save_tasks ( audio_tasks , output_path = "./audio" )


asyncio . run ( multiple_text_to_speech ())

貢獻

歡迎捐款！請分配存儲​​庫，並通過改進來創建拉動請求。