YouTube Summariser

This provides sample Blazor apps that summarise a YouTube video transcript to a given language.

Prerequisites

• .NET 7 SDK
• Visual Studio or Visual Studio Code with C# Dev Kit
• Azure Subscription
• Azure OpenAI Service

Known Issues

• If your YouTube video is long enough, you might exceed the maximum number of tokens that Azure OpenAI Service can handle. For more information, see Model summary.
  • gpt-35-turbo: 4K
  • gpt-35-turbo-16k: 16K
  • gpt-4: 4K
  • gpt-4-32K: 32K

Getting Started

Provision Resources to Azure

1. Fork this repository to your GitHub account, {{GITHUB_USERNAME}}.

2. Run the commands below to set up a resource names:

   # PowerShell
   $AZURE_ENV_NAME="summariser$(Get-Random -Min 1000 -Max 9999)"
   $GITHUB_USERNAME="{{GITHUB_USERNAME}}"
   $GITHUB_REPOSITORY_NAME="{{GITHUB_REPOSITORY_NAME}}"
   
   # Bash
   AZURE_ENV_NAME="summariser$RANDOM"
   GITHUB_USERNAME="{{GITHUB_USERNAME}}"
   GITHUB_REPOSITORY_NAME="{{GITHUB_REPOSITORY_NAME}}"

3. Run the commands below to provision Azure resources:

   azd auth login
   azd init -e $AZURE_ENV_NAME
   azd up

   Note: You may be asked to enter your GitHub username/orgname and repository name where your forked repository is located.

4. Get the endpoint, API key and deployment ID.

   # PowerShell
   $AOAI_ENDPOINT_URL = $(az cognitiveservices account show `
                            -g rg-$AZURE_ENV_NAME `
                            -n aoai-$AZURE_ENV_NAME `
                            --query "properties.endpoint" -o tsv)
   
   $AOAI_API_KEY = $(az cognitiveservices account keys list `
                       -g rg-$AZURE_ENV_NAME `
                       -n aoai-$AZURE_ENV_NAME `
                       --query "key1" -o tsv)
   
   $AOAI_DEPLOYMENT_ID = $(az cognitiveservices account deployment list `
                             -g rg-$AZURE_ENV_NAME `
                             -n aoai-$AZURE_ENV_NAME `
                             --query "[0].name" -o tsv)
   
   # Bash
   AOAI_ENDPOINT_URL=$(az cognitiveservices account show \
                         -g rg-$AZURE_ENV_NAME \
                         -n aoai-$AZURE_ENV_NAME \
                         --query "properties.endpoint" -o tsv)
   
   AOAI_API_KEY=$(az cognitiveservices account keys list \
                    -g rg-$AZURE_ENV_NAME \
                    -n aoai-$AZURE_ENV_NAME \
                    --query "key1" -o tsv)
   
   AOAI_DEPLOYMENT_ID=$(az cognitiveservices account deployment list \
                          -g rg-$AZURE_ENV_NAME \
                          -n aoai-$AZURE_ENV_NAME \
                          --query "[0].name" -o tsv)

Deploy Applications to Azure

1. Run the commands below to deploy apps to Azure:

   az login
   gh auth login
   azd pipeline config
   gh workflow run "Azure Dev" --repo $GITHUB_USERNAME/$GITHUB_REPOSITORY_NAME

Deprovision Resources from Azure

1. To avoid unexpected billing shock, run the commands below to deprovision Azure resources:

   azd down --force --purge --no-prompt

Power Platform Custom Connector in Visual Studio

TBD

Teams App with Blazor Server App in Visual Studio

TBD

Blazor WebAssembly App in Visual Studio Code

1. Copy local.settings.sample.json to local.settings.json

2. Update local.settings.json file for backend API.

   "OpenAI__Endpoint": "{{AOAI_ENDPOINT_URL}}",
   "OpenAI__ApiKey": "{{AOAI_API_KEY}}",
   "OpenAI__DeploymentId": "{{AOAI_DEPLOYMENT_ID}}",

3. Build the entire solution.

   dotnet restore && dotnet build

4. Open a new terminal and run the backend API app.

   pushd src/YouTubeSummariser.ApiApp
   func start
   popd

5. Open a new terminal and run the Blazor WebAssembly app.

   pushd /src/YouTubeSummariser.WebApp.Wasm
   dotnet watch run
   popd