Google Vertex AI

Category: Chat ModelPackage: TruBot.OpenAI.Activities

Purpose

Enables to interact in a conversational dialogue form and provides responses in efficient and accurate form that imitates human conversation.

Description

The Google Vertex AI component enables to interact in conversational dialogue form and provides responses in efficient and accurate form that imitates human conversational form. In the given example, a question has been put forth to the Google Vertex AI chat model and the response is assigned to the Result property.

Properties

Property	Data Type	Description
General
Display Name*	String	Display name of the component.
Enable Pause	-NA-	Option to pause the bot execution based upon the signal received from the Cockpit application. The checkbox is selected by default.
Input
Conversation Request*	String	Conversation request text either in plain English text format or in required Question format to get the response from Google Vertex AI.
Google Credential*	SecureString	User credentials for accessing the Google Vertex AI chat service.
Location*	String	URL for accessing the Google Vertex AI chat service
Model Name*	String	Model name of the Google Vertex AI chat model.
Project Id*	String	The unique Id of the Google Cloud project that is to be used for managing the resources.
Provider*	-NA-	Provider for chat embedding model.
Version*	String	Version of the chat model.
Misc
Is Reserved	-NA-	Option to disable data tracing related to the component.
Enable Bookmark	-NA-	Option to set a bookmark.
Multimodal
Image Paths	ImagePaths(Dictionary<String, GeminiImageMimeType>)	List of path for images that are to be analyzed.
Mime Type	-NA-	The format in which the video that is to be analyzed is available.
Video Path	String	Te video that is to be analyzed.
Output
Connection	IChatModel	Stored the Connection object of Google Vertex AI
Result	AIResponse	Stores the returned response of Google Vertex AI where response will be either required output or exceptions
Preferences
Existing Conversation	Boolean	Allows to use the past conversation requests that were done with Google Vertex AI using TruBot.
Maximum Tokens*	Int32	Decides the maximum number of tokens to generate.
Top K*	Double	TopK determines how many of the highest probability tokens are considered during the model's output generation, influencing response randomness; a lower value yields less randomness, while a higher value increases randomness.
Top P*	Double	Controls diversity via nucleus sampling. 0.5 means half of all likelihood-weighted options are considered. Supports Top Probabilities range between 0 to 1.
Temperature*	Double	Controls the randomness. Lowering results in less random completions. As the temperature approaches 0, the model will become deterministic and repetitive. Supports Temperature range between 0 to 1.
Stop Sequence	List<String>	List of strings that, if found in the response, prompt the model to stop generating text. The response truncates at the first occurrence of any specified string. Case-sensitive
Retriever	IVectorStore	It represents the data that needs to be searched for the query represented by the Conversation Request property. Note: If the value is not provided then the model will use its own database to answer the query.
Safety Settings
Harassment	-NA-	Preference about the content that contains references to sexual acts or other lewd content.
Hate speech	-NA-	Preference about the content that contains negative or harmful comments targeting identity and/or protected attributes.
Sexually explicit	-NA-	Preference about the content that contains references to sexual acts or other lewd content.
Dangerous	-NA-	Preference about the content that contains references to sexual acts or other lewd content.
Settings
Timeout	Int32	Specific amount of time in seconds to wait for a response from the server before throwing an error
Note: Fields marked with * sign are mandatory.