Skip to content

models bring your own data chat qna

github-actions[bot] edited this page Feb 3, 2024 · 11 revisions

bring-your-own-data-chat-qna

Overview

The "Bring Your Own Data Chat QnA" is a pre-trained chat model, enhanced by GPT3.5, that leverages your personally indexed data and chat history to deliver more concrete and relevant answers. It involves processing the raw query through an embedding procedure, followed by a "Vector Search" to pinpoint the most pertinent context within the user's data. Subsequently, GPT3.5 is employed to generate a comprehensive answer to the question using the sourced documents.

Inference samples

Inference type CLI VS Code Extension
Real time deploy-promptflow-model-cli-example deploy-promptflow-model-vscode-extension-example
Batch N/A N/A

Sample inputs and outputs (for real-time inference)

Sample input

{
    "inputs": {
        "question": "What is Azure compute instance?"
    }
}

Sample output

{
    "outputs": {
        "answer": "An Azure Machine Learning compute instance is a fully managed cloud-based workstation for data scientists. It provides a pre-configured and managed development environment in the cloud for machine learning. Compute instances can also be used as a compute target for training and inferencing for development and testing purposes. They have a job queue, run jobs securely in a virtual network environment, and can run multiple small jobs in parallel. Additionally, compute instances support single-node multi-GPU distributed training jobs."
    }
}

Version: 7

View in Studio: https://ml.azure.com/registries/azureml/models/bring-your-own-data-chat-qna/version/7

Properties

is-promptflow: True

azureml.promptflow.section: gallery

azureml.promptflow.type: chat

azureml.promptflow.name: Multi-Round Q&A on Your Data

azureml.promptflow.description: Create a chatbot that uses LLM and data from your own indexed files to ground multi-round question and answering capabilities in enterprise chat scenarios.

inference-min-sku-spec: 2|0|14|28

inference-recommended-sku: Standard_DS3_v2

Clone this wiki locally