models multi index rerank qna - Azure/azureml-assets GitHub Wiki

multi-index-rerank-qna

Overview

This "Multi-Source Rerank Q&A" demonstrates Q&A application, enabled by reranking data from multiple sources and powered by GPT. It utilizes indexed files and the rerank tool from Azure Machine Learning to provide grounded answers. You can ask a wide range of questions and receive responses based on a wide set of stored data. The process involves taking the query, extracting pre-existing documents across multiple sources, reranking said documents for the most relevant context, and then using GPT to chat with you, given those 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": {
        "query": "How to use SDK V2?"
    }
}

Sample output

{
    "outputs": {
        "answer": "To use the Azure Machine Learning Python SDK v2, you need to have an Azure Machine Learning workspace and the SDK installed. You can either create a compute instance, which automatically installs the SDK and is pre-configured for ML workflows, or use the provided commands to install the SDK. (Source: https://github.com/prakharg-msft/azureml-tutorials/blob/main//how-to-auto-train-image-models.md)"
    }
}

Version: 1

View in Studio: https://ml.azure.com/registries/azureml/models/multi-index-rerank-qna/version/1

Properties

is-promptflow: True

azureml.promptflow.section: gallery

azureml.promptflow.type: standard

azureml.promptflow.name: Multi-Source Rerank Q&A

azureml.promptflow.description: Use LLM and reranked data from multiple indices to ground question and answering capabilities.

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

inference-recommended-sku: Standard_DS3_v2

⚠️ **GitHub.com Fallback** ⚠️