QGPTPromptPipeline
This is a model related to switching between different prompts based on if we are dealing with various tasks or if we are attempting to converse with LLMs via conversation. You will have 2 options- 1) task- This is specifically for 1 off task operations for instance explaning a bit of code 2) conversation- This is specifically for conversing with our LLMs, will provide better results && high fedility responses for instance contextualize code conversations.
Properties
Name | Type |
---|---|
schema | EmbeddedModelSchema |
task | QGPTTaskPipeline |
conversation | QGPTConversationPipeline |
Example
import { QGPTPromptPipeline } from '@pieces.app/pieces-os-client'
// TODO: Update the object below with actual values
const example: QGPTPromptPipeline = {
"schema": null,
"task": null,
"conversation": null,
}
console.log(example)
// Convert the instance to a JSON string
const exampleJSON: string = JSON.stringify(example)
console.log(exampleJSON)
// Parse the JSON string back to an object
const exampleParsed = JSON.parse(exampleJSON) as QGPTPromptPipeline
console.log(exampleParsed)