Model Settings
This page guides you on how to view and switch between available AI models, as well as how to configure and adjust model parameters.
I. Switch Chat Models
Open the app and click the blue "API Connection" button
OR Open the app and click the "Menu"
Click the "API Connection" button
On the API Connection page, click the menu icon next to the API you wish to use
Set your preferred API as the default
II. Model Parameter Configuration
If you need precise control over the model's generation effects, behavior, and resource usage, you can access the configuration interface through the following two main paths to make adjustments.
Path 1: Adjust via Global Settings
This path is suitable for setting global default parameters or making quick adjustments.
Operation Path:
◦ Open the left sidebar.
◦ Click the More option, find and select Settings in the menu.
◦ After entering the settings interface, select the Model Settings tab.Adjustable Parameters: On this page, you can globally adjust the following key model parameters:
◦ Context Memory Length: Controls the length of conversation history the model can remember.
◦ Reply Token Limit: Limits the maximum length of a single reply generated by the model.
◦ Temperature: Controls the randomness and creativity of replies. Higher values yield more diverse replies; lower values produce more deterministic replies.
◦ Top-P: Lower values result in more consistent personality traits; higher values lead to more multifaceted personality expressions.
◦ Top-K: Lower values make the dialogue style more fixed; higher values introduce more variability in the dialogue style.
◦ Streaming Output Toggle: Controls whether replies are displayed in real-time character by character (enabled) or displayed in full at once (disabled).
Path 2: Adjust via API Configuration Page
1. On the API configuration page, locate and click the "Parameter Settings" button.
2. You will enter the model parameter configuration interface.
Configuration Instructions
- Reference Interface Examples: The interface provides example values and effect descriptions for main parameters (e.g., temperature, maximum generation length) to help you quickly understand.
- Consult Official Documentation: For optimal results, it is recommended to refer to the official API documentation of each model platform (e.g., OpenAI, Anthropic) to understand the exact definitions and value ranges of parameters.
- Custom Configuration: Based on the above information and your actual needs, fill in or adjust the values of each parameter.
Note: Some parameters may exist in both interfaces. Parameter settings on the API configuration page usually have higher priority and will override the same parameter in global settings.