HANDS-ON-LAB

Deploying Bitcoin Search Engine in Azure Project

Problem Statement

Deploy the Bitcoin domain-specific embeddings you have created in the previous project using Word2vec and FastText as a Streamlit app in Azure.

Tasks

  1. Create a resource group in Azure. Add storage account & create folder structure similar to the one shared in the project and upload the bitcoin data file in blob.

  2. Create the preprocessing.py, training_model.py, return_embed.py, top_n.py files similar to the one explained in the guided Azure Text Analytics for medical search engine deployment project. (basically modularizing the code to train embeddings from the previous exercise)

  3. Add DataBricks to the resource group. Create load_data.py (with provisions to load data to Databricks from Azure blob) and main.py file and check the working of the codes in Databricks.

  4. Create a new pipeline in Azure Data Factory and set the trigger.

  5. Create the app.py code similar to the one shared in the project for creating the webApp.

  6. Create a DOCKER file to invoke the relevant files when deployed.

  7. Add Web App Service to the resource group and deploy the docker and run the app.

 

Get started with Azure and Streamlit today

FAQs

Q1. How do I create a resource group in Azure?

To create a resource group in Azure, you can use the Azure portal, Azure CLI, or Azure PowerShell. The following command creates a resource group using Azure CLI:

sql

az group create --name myResourceGroup --location "East US"

 

Q2. How can I deploy a Docker container in Azure Web App Service?

To deploy a Docker container in Azure Web App Service, you can use Azure CLI or Azure portal. With Azure CLI, you can use the following command:

css

 

az webapp create --resource-group myResourceGroup --name myWebApp --plan myAppServicePlan --deployment-container-image-name myDockerImage

 

Q3. What is the purpose of Azure Data Factory in this project?

Azure Data Factory is used to create a pipeline for orchestrating the data processing tasks in this project. It helps automate the process of loading data, transforming it, and triggering the necessary actions for deployment.