PromptVertexAI 2025.3.28.13-SNAPSHOT¶
BUNDLE¶
com.snowflake.openflow.runtime | runtime-vertexai-nar
DESCRIPTION¶
Sends a prompt to VertexAI, writing the response either as a FlowFile attribute or to the contents of the incoming FlowFile. The prompt may consist of pure text interaction or may include multimedia.
INPUT REQUIREMENT¶
Supports Sensitive Dynamic Properties¶
false
PROPERTIES¶
Property |
Description |
|---|---|
GCP Credentials Service |
The Controller Service used to obtain Google Cloud Platform credentials. |
GCP Location |
The location to configure the Vertex client with |
GCP Project ID |
The project ID to configure the Vertex client with |
Max File Size |
The maximum size of a FlowFile that can be sent to Vertex as an image. If the FlowFile is larger than this, it will be routed to ‘failure’. |
Max Tokens |
The maximum number of tokens to generate |
Media MIME Type |
The MIME type of the media in the FlowFile content. Supported media types are listed here: https://firebase.google.com/docs/vertex-ai/input-file-requirements |
Model Name |
The name of the Vertex model |
Output Strategy |
Determines response output destination |
Prompt Type |
The type of prompt to send to Vertex. Text to send a simple prompt. Media to send a multimedia type first followed by a text prompt. |
Response Format |
The format of the response from Vertex |
Results Attribute |
The name of the attribute to write the response to. |
Stop Sequences |
A comma delimited list of strings act as stop sequences. The model will halt after encountering one of the stop sequences. |
System Message |
The system message to send to Vertex. FlowFile attributes may be referenced via Expression Language, and the contents of the FlowFile may be referenced via the flowfile_content variable. E.g., ${flowfile_content} |
Temperature |
The temperature to use for generating the response. Defaults to 1.0. Ranges from 0.0 to 1.0. Use temperature closer to 0.0 for analytical / multiple choice, and closer to 1.0 for creative and generative tasks. |
Top K |
The top K value to use for generating the response. Only sample from the top K options for each subsequent token. Recommended for advanced use cases only. You usually only need to use temperature. |
Top P |
The top P value to use for generating the response. Top P is for nucleus sampling, we compute the cumulative distribution over all the options for each subsequent token in decreasing probability order and cut it off once it reaches a particular probability specified by top_p.Recommended for advanced use cases only. You usually only need to use temperature. |
User Message |
The user message to send to Vertex. FlowFile attributes may be referenced via Expression Language, and the contents of the FlowFile may be referenced via the flowfile_content variable. E.g., ${flowfile_content}. The user message is added first, unless an image is present. |
RELATIONSHIPS¶
NAME |
DESCRIPTION |
|---|---|
failure |
If unable to obtain a valid response from Vertex, the original FlowFile will be routed to this relationship |
success |
The response from Vertex is routed to this relationship |
WRITES ATTRIBUTES¶
NAME |
DESCRIPTION |
|---|---|
vertex.usage.inputTokens |
The number of input tokens read in the request. |
vertex.usage.outputTokens |
The number of output tokens generated in the response. |
vertex.chat.completion.id |
A unique id assigned to the conversation |
mime.type |
The mime type of the response. |
filename |
An updated filename for the response. |