AI Model
Last updated
Last updated
The AI model action block is the perfect solution for providing intelligent and informative responses to open-ended queries from visitors.
By utilizing OpenAI APIs and leveraging trained data, this powerful tool enables you to deliver smarter replies that will impress and engage your audience.
With AI model block, you can confidently address any question / inquiry accuracy and efficiency.
Here you will see all the AI Models which you have created in the AI Studio. Depending on the AI model that is selected, the chatbot will provide responses to inquiries. It is important to ensure that the AI model has been properly trained with accurate information.
In order to obtain a response from the AI model, it is necessary to submit the question to it.
Prior to the "AI model" action block, it is important to utilize a "Collect input" action block to inquire about the question and store it in a variable.
The variable holding the question should be selected here.
Create a persona text that serves as the basis for the AI model's response. This allows you to receive answers to your questions in a format that is easily understandable for you.
Some of the examples of persona are as follows:
Answer the questions asked in a simple and easy to understand manner.
Answer these questions as a 5-year old.
Give me answers considering me as a technical person.
For the duration of this conversation, please fully immerse yourself in the persona of Albert Einstein.
When using the AI model, you have the option to specify the desired length of the responses when asking a question.
This length is determined based on tokens, which can be considered as fragments of words.
Specifically, 1 token is equivalent to 4 characters.
Users have the option to customize their experience by choosing whether they prefer unique responses for each question or a consistent response for repeated inquiries.
This can be achieved by defining a number between 0 and 1. A value closer to 1 provides distinct responses, while a value closer to 0 delivers fixed responses.
By default, the system is set at 0.5, striking a balance between variety and consistency.
The objective of constructing an AI model is to have the chatbot respond exclusively to queries derived from a particular domain or set of files.
This setup grants you the ability to determine whether you want the chatbot to address inquiries using the broader knowledge of the global ChatGPT model, or restrict its responses solely to the information it has been trained on within that specific AI model.
As a default setting, this configuration is deactivated.
After responding to a query, the chatbot will prompt the user for feedback by providing options for a thumbs up or thumbs down.
The feedback received will be displayed on the AI studio page, allowing for visibility and assessment of performance.