Skip to main content

Settings

The settings section in Ayushma allows users to customize their experience and preferences for interacting with the AI assistant.

Accessing Settings​

To access the settings, navigate to the settings page within Ayushma's web interface. The exact location and method may vary depending on the specific implementation, but it is typically accessible through a dedicated "Settings" or "Preferences" menu option.

Available Settings​

  • OpenAI API Key:
    • If you have been granted the "Access Key Allowed" permission, you can enter your own OpenAI API key in this field. This allows you to use your own OpenAI account for conversations, giving you more control over the AI model used and the associated costs.
  • Override Key:
    • If you have entered your own OpenAI API key, you can enable the "Override Key" option to use your key instead of Ayushma's default key for conversations.
  • Show stats for nerds:
    • Enable this option to display detailed statistics about each AI response, such as response time, translation time, and token usage. This is useful for understanding the performance and behavior of the AI model.
  • Autoplay responses text to speech:
    • Turn on this option to automatically play the AI's responses as speech using text-to-speech technology. This is helpful for users who prefer to listen to the responses rather than read them.
  • Show Original English Responses:
    • If you are using Ayushma in a language other than English, enabling this option will display the original English version of the AI's responses alongside the translated version.
  • Chat Temperature:
    • Adjust the "Chat Temperature" slider to control the creativity and randomness of the AI's responses. A lower temperature will result in more factual and deterministic responses, while a higher temperature will lead to more creative and diverse outputs.
  • References:
    • Use the "References" slider to control the number of reference documents considered by the AI when generating responses. A higher value will provide the AI with more context but may increase response time.