PVT Sakura Version 1 is a simple real-time translation for advanced users. The following operation panel and video will help you easily translate audio on your computer into the specified language. Set up and start translating in the following order.

 

A series of operating procedures are shown in the video below.

1. How to obtain an API key?

If you search, you will find a detailed explanation on how to obtain an OpenAI API key. The following is for reference only, but please register an account and obtain an API key from the URL below while referring to several sites. The API key you obtain will be issued only once. In case you lost it, you need to reissue it again.

 

ref. How to obtain an API key?

The following is not the official OpenAI website, but a third party carefully explains how to obtain the information. For reference.

ref. Link1:How to obtain an API key?

ref. Link2:How to obtain an API key? (日本語)

 

2. How to obtain Faster-whisper model

  Faster-whisper is available in several models. However, the models that can run lightly on CPUs are almost always tiny or base. Depending on the GPU, large-v3 may work with high accuracy on the latest RTX30,40 series PCs.

Below are the specifications of the CUDA environment. (For advanced users)

 

 CUDA Toolkit 12.x and cuDNN 9.x are recommended. 

 

Ref.:CUDA Toolkit Archive | NVIDIA Developer

           CuDNN Archive