Part of GPT-RAG
Here are the steps to configure cognitive search and deploy ingestion code using the terminal.
First check your environment meets the requirements
- You need AZ CLI to log and run Azure commands in the command line.
- You need Python 3.10 to run the setup script. Miniconda helps you creating and managing your python environments.
- Azure Functions Core Tools will be needeed to deploy the chunking function.
1) Login to Azure
run az login to log into azure. Run az login -i if using a VM with managed identity to run the setup.
2) Clone the repo
If you plan to customize the ingestion logic, create a new repo by clicking on the Use this template button on top of this page.
Clone the repostory locally: git clone https://github.com/azure/gpt-rag-orchestrator
If you created a new repository please update the repository URL before running the command
3) Adjust your prompt
In VSCode you can change the prompt text if you want a custom prompt.
Use the question_answering.prompt located in orc/prompts/ folder, just rembember to keep the Sources: {sources} text in the bottom.
4) Deploy function to Azure
Enter in the cloned repo folder:
cd gpt-rag-orchestrator
Use Azure Functions Core Tools to deploy the function:
func azure functionapp publish FUNCTION_APP_NAME --python
After finishing the deployment run the following command to confirm the function was deployed:
func azure functionapp list-functions FUNCTION_APP_NAME
Replace FUNCTION_APP_NAME with your Orchestrator Function App name before running the command
5) Deploy locally (optional)
With Azure Function extension installed you just need to open orc/orchestrator.py and "Start Debugging" in VSCode.
It will start the server in http://localhost:7071/api/orc.
Since we're now using managed identities you will have to assign the following roles to your user in order to test orchestrator locally:
- Azure CosmosDB 'Cosmos DB Built-in Data Contributor' role.
bash
resourceGroupName='your resource group name'
cosmosDbaccountName='CosmosDB Service name'
roleDefinitionId='00000000-0000-0000-0000-000000000002'
principalId='Object id of your user in Microsoft Entra ID'
az cosmosdb sql role assignment create --account-name $cosmosDbaccountName --resource-group $resourceGroupName --scope "/" --principal-id $principalId --role-definition-id $roleDefinitionId
PowerShell
$resourceGroupName='your resource group name'
$cosmosDbaccountName='CosmosDB Service name'
$roleDefinitionId='00000000-0000-0000-0000-000000000002'
$principalId='Object id of your user in Microsoft Entra ID'
az cosmosdb sql role assignment create --account-name $cosmosDbaccountName --resource-group $resourceGroupName --scope "/" --principal-id $principalId --role-definition-id $roleDefinitionId
- Azure OpenAI resource 'Cognitive Services OpenAI User' role.
bash
subscriptionId='your subscription id'
resourceGroupName='your resource group name'
openAIAccountName='Azure OpenAI service name'
principalId='Object id of your user in Microsoft Entra ID'
az role assignment create --role "Cognitive Services OpenAI User" --assignee $principalId --scope /subscriptions/$subscriptionId/resourceGroups/$resourceGroupName/providers/Microsoft.CognitiveServices/accounts/$openAIAccountName
PowerShell
$subscriptionId='your subscription id'
$resourceGroupName='your resource group name'
$openAIAccountName='Azure OpenAI service name'
$principalId='Object id of your user in Microsoft Entra ID'
az role assignment create --role "Cognitive Services OpenAI User" --assignee $principalId --scope /subscriptions/$subscriptionId/resourceGroups/$resourceGroupName/providers/Microsoft.CognitiveServices/accounts/$openAIAccountName
References
- Cognitive search: Querying a Vector Index, REST API Reference and Querying Samples.
This project welcomes contributions and suggestions. Most contributions require you to agree to a Contributor License Agreement (CLA) declaring that you have the right to, and actually do, grant us the rights to use your contribution. For details, visit https://cla.opensource.microsoft.com.
When you submit a pull request, a CLA bot will automatically determine whether you need to provide a CLA and decorate the PR appropriately (e.g., status check, comment). Simply follow the instructions provided by the bot. You will only need to do this once across all repos using our CLA.
This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact opencode@microsoft.com with any additional questions or comments.
This project may contain trademarks or logos for projects, products, or services. Authorized use of Microsoft trademarks or logos is subject to and must follow Microsoft's Trademark & Brand Guidelines. Use of Microsoft trademarks or logos in modified versions of this project must not cause confusion or imply Microsoft sponsorship. Any use of third-party trademarks or logos are subject to those third-party's policies.