FAQ
Frequently Asked Questions
This section answers common questions about the project's functionalities and usage, helping users resolve frequently encountered issues.
If you have any suggestions or feedback on this process, please create an issue on the GitHub repository to help us improve the user experience.
Common Questions
- Why do I need to configure the model service API key?
- Is the application free to use?
- Why cannot we find the chat history when I open different workspace?
General Questions
Why do I need to configure the model service API key?
API keys are required for authentication and authorization when connecting to model services. Each model service has its own API key, which is used to verify the user's identity and access permissions.
Currently, I cannot offer a shared API key for all model services, so you will need to configure the API key for each service you intend to use.
How do I obtain an API key for a model service?
To get an API key for a model service, you can just simply click the Get API Key link next to the input field for the specific model service you want to configure.
If the link is invalid, please let us know as fast as possible.
Is the application free to use?
Yes, the application is free to use.
As far as I maintain the project, I will keep it free to use and open-source.
Why cannot we find the chat history when I open different workspace?
We current use the workspace .vscode folder to store the chat history, so if you open a different workspace, you will not find the chat history.
Please let us know if you do need to share the chat history between different workspaces. I may consider adding a shared chat history feature.
Model Related Questions
Why do we only allow local model service for auto-completion?
The auto-completion feature is based on the local model service, which is more secure and faster than the remote model service. Also, it can run more frequently than the remote model service. We also plan to apply LoRa tuning to the local model service to improve the auto-completion performance.
Otherwise, you will cost a lot of payment for the remote model service.