This sample demonstrates how to build a text completion sample using Azure Functions and a local build of the Azure OpenAI extension.
The sample is available in the following language stacks:
Please refer to the root level README for prerequisites.
-
Start Azurite for local development storage. See these instructions for more information on how to work with Azurite.
-
Reference the table below for instructions on building and starting the app:
Language Worker Command .NET oo-proc cd samples/textcompletion/csharp-ooproc && dotnet build && cd bin/debug/net6.0 && func start
Javacript cd samples/textcompletion/javascript && npm install && npm run start
TypeScript cd samples/textcompletion/typescript && npm install && npm run build && npm run start
PowerShell cd samples/textcompletion/powershell && dotnet build --output bin && func start
Python cd samples/textcompletion/python && pip install -r requirements.txt && func start
Java cd samples/textcompletion/java && mvn clean package && dotnet build && mvn azure-functions:run
If successful, you should see the following output from the
func
command:Azure Functions Core Tools Core Tools Version: 4.0.5530 Commit hash: N/A (64-bit) Function Runtime Version: 4.28.5.21962 Functions: GenericCompletion: [POST] http://localhost:7071/api/GenericCompletion WhoIs: [GET] http://localhost:7071/api/whois/{name}
- Refer the demo.http file for the format of requests.
- Send a request to the
WhoIs
function. The following is an example request:
GET http://localhost:7071/api/whois/pikachu
The HTTP response should look something like the following example (with newlines added for readability):
Pikachu is a fictional creature from the Pokemon franchise. It is a yellow
mouse-like creature with powerful electrical abilities and a mischievous
personality. Pikachu is one of the most iconic and recognizable characters
from the franchise, and is featured in numerous video games, anime series,
movies, and other media.