Starting from version 1.92, ByteDance VolcEngine Ark models can be used for translation within the software. Here's a detailed guide on how to enable and use them:

Step One: Log In & Register

Go to https://console.volcengine.com/auth/login/. If you don't have an account, register first; otherwise, log in directly.

Step Two: Real-Name Verification

This step is unavoidable; it's a mandatory legal requirement for domestic services in China. If you want to use the service, you must comply. After logging in, open your account center at the following address and click "Real-Name Verification":

https://console.volcengine.com/user/authentication/detail/

You can choose between face verification or bank card verification.

Step Three: Activate Services & Claim Benefits

After successful real-name verification, navigate to https://console.volcengine.com/ark/region:ark+cn-beijing/openManagement until you reach the "Service Management" page.

Then, click on any "Activate Service" button. On the page that opens, select all models and confirm activation.

Click "Submit" to continue.

Once done, return to the "Service Management" page and refresh. You should see that all services have been successfully activated.

Now that the services are activated, how do you use them in your code? Let's continue.

Create an API Key

Below the "Service Management" section, you'll find the "API key" menu. Click on it.

One last step remains: creating an "Inference Endpoint." You will use this endpoint name where a model needs to be specified.

Create a Model Inference Endpoint

From the same location, glance upwards, and you'll see the "Model Inference" menu. Click on it.

After clicking "Create Inference Endpoint," the following interface will pop up. Fill in the "Endpoint Name" as you wish, then click "Add Model." A list of available models will appear.

As shown below, in the model selection, it's recommended to only choose models that are present in your "Service Management," such as those starting with Doubao-lite, Doubao-pro, or Moon. After selecting, several model versions will appear on the right. You need to choose a specific one; here, it's recommended to select versions with numerical names, like 240515.

Here are my selections:

After creation, return to the "Model Inference" list page. Note the string of text starting with ep-; that is the "Endpoint Name" you'll use in your code.

In your code, this endpoint name and your API key are the two required parameters.

How to Use in pyVideoTrans

https://github.com/jianchang512/pyvideotrans

1. Upgrade pyVideoTrans to version 1.92+ (address above).
2. Open the videotrans/set.ini file, search for zijiehuoshan_model=, and enter your endpoint name after the equals sign. Note that this name must be the automatically generated one.

After filling it in, it will look similar to zijiehuoshan_model=ep-20240615073625-zrlbj,ep-41521-auyed. If you have multiple endpoints, separate them with English commas.

3. Open the software, go to the "Settings" menu, then "ByteDance VolcEngine," fill in your API key, and select the endpoint(s) you wish to use.
4. If testing shows no issues, you can now use it from the main interface.

Important Usage Notes

For well-known reasons, when using this model, you must be careful not to send or inquire about any sensitive questions, content containing sensitive keywords, or unhealthy content. Minor violations may result in a "I cannot answer this question" response, while severe violations could lead to account suspension.