Chuyển tới nội dung chính

Generate Chat Completion

RCA.Activities.AIServices.GenerateChatCompletion

Description​

The Generate Chat Completion activity generates a chat completion response using an OpenAI model based on user prompts. This activity must be used inside an AI Scope with Provider Type set to OpenAI.

chat-open-ai

(* is mandatory)

In the body of activity​

  • Prompt - The input prompt used to generate the chat completion.

Properties​

Common

  • Continue On Error (Boolean) - A Boolean variable has two possible values: True or False
    - True : allows the rest of the process to continue the execution even an error occurs within the activity.
    - False : blocks the process from continuing the execution.
  • Timeout MS (Int32) - The maximum amount of time (in milliseconds) to wait for the activity to complete before an error is thrown. Default value: 30000 (milliseconds).
    E.g: 30000

Input

  • Prompt (String)* - The input prompt for the chat completion.

Options

  • Best Of (Int32?) - Generates multiple completions server-side and returns the best one. Must be greater than N when used with N. Defaults to 1.
  • Echo (Boolean?) - Echoes back the prompt in addition to the completion. Defaults to False.
  • Frequency Penalty (Double?) - Number between -2.0 and 2.0. Positive values reduce repeated text. Defaults to 0.
  • Instruction (String) - A system-level instruction for the AI model to follow throughout the conversation. It is added once at the beginning of the AI Scope conversation history.
  • Log Probabilities (Int32?) - Includes the log probabilities on the most likely output tokens.
  • Max Tokens (Int32) - The maximum number of tokens allowed for the prompt and generated answer. Default value: 1920.
  • N (Int32?) - Number of completion choices to return. Defaults to 1.
  • Presence Penalty (Double?) - Number between -2.0 and 2.0. Positive values increase the likelihood of new topics. Defaults to 0.
  • Stop (String[]) - Up to 4 sequences where the API will stop generating further tokens.
  • Stream (Boolean?) - Whether to stream back partial progress. Defaults to False.
  • Suffix (String) - The suffix that comes after a completion of inserted text.
  • Summarization Message Threshold (Int32) - Message count threshold that triggers conversation summarization. Default value: 15.
  • Summarization Token Threshold (Int32) - Token threshold that triggers conversation summarization to save tokens for future turns. Default value: 3000.
  • Temperature (Double?) - A number between 0 and 2. Higher values produce more random output, while lower values make it more focused. Defaults to 1.
  • Top P (Double?) - A number between 0 and 1. The lower the number, the fewer tokens are considered. Defaults to 1.

Output

  • Full Response (JObject) - The full response from the service.
  • Result (String) - The result of the chat completion.

Misc

  • Public (Checkbox) - Check if you want to public the activity. Remember to consider data security requirement before using this property.
  • Display Name (String) - The name of this activity. You can edit the name of the activity to organize and structure your code better.
    E.g: Generate Chat Completion