From the course: Introduction to Azure AI Foundry: Exploring Services and Capabilities
Unlock this course with a free trial
Join today to access over 24,900 courses taught by industry experts.
Use a standard prompt flow
From the course: Introduction to Azure AI Foundry: Exploring Services and Capabilities
Use a standard prompt flow
- [Instructor] One improvement we can make to our prompt flow is to provide extra information for it to use when generating responses. That information could be from a database, from index documents or from a web search engine. Before we see how to do that, let's talk about this new prompt flow. For this example, I'm using a standard flow. Now the interface for creating the prompt flow looks very similar to what we've seen with the chat flow, but this is simpler than a chat flow for demonstration purposes because we don't have to manage chat history. That makes it easier to focus on how the flow processes inputs and generates responses without handling multi-turn interactions. So this flow is called Hiking-Without-Serp. We're going to start with this one, and then we'll add Serp. That stands for Search Engine Result Page. We start with the inputs. There are two inputs. Hike_location is the name of one and Sun_Info is the name of the other. And I've hard coded the values here. That's…
Contents
-
-
-
-
-
-
-
(Locked)
What is a prompt flow?3m 42s
-
(Locked)
Load and edit system instructions5m 25s
-
(Locked)
Create a simple prompt flow3m 19s
-
(Locked)
Explore the prompt flow editor2m 52s
-
(Locked)
The Python template code4m 40s
-
(Locked)
Use a standard prompt flow3m 37s
-
(Locked)
Use search engine results with Serp API4m 34s
-
(Locked)
Deploy a prompt flow or playground1m 39s
-
(Locked)
-