Unleashing PartyRock: Features and Capabilities

Create AI applications with PartyRock and Amazon Bedrock

Visit PartyRock.aws if you’re prepared to discover about the subject of generative AI while experiencing fun and creating amazing things. Without knowing how to write code or sign up for an AWS account, you can experiment, discover everything there is to know about prompt engineering, create mini-apps, and then share them among your friends. To further improve and customize an application that was originally shared with you, you may also start with it and remix it.

Through PartyRock

1.We go to partyrock.aws click Sign in, and use my Google, Apple, or Amazon account to log in to get started.

2.We have successfully authenticated and now on PartyRock’s home page. We can choose “Build your own app” to get started, or we can go over some sample apps.

3.Alternatively, we may construct the app from scratch, widget by widget, or we can enter a description and use PartyRock’s Generative AI to gain a head start.

4.After entering our prompt, we select Generate App.

5.After a few seconds, AWS software is ready, and we enter some information to check if the output has enough snark for my purposes.

6.That looks fantastic, let’s disassemble it and see how it functions!

7.User Input and Snarky Response are the two widgets available in the program. When we select the Edit icon on the first product, we can see that it has a default value, a title, and loading text. Widgets can relate to one another by name thanks to the title.

8.An InvokeModel function call to Amazon Bedrock is encapsulated in this straightforward widget. The widget links to the User Input widget and provides instructions on how to utilize the Claude v2 model along with a straightforward prompt. We can try adjusting one, saving the modification, and watching for a moment or two to see what happens. Claude Instant, for instance, causes me to receive a somewhat different response when we change the model.

9.We should now like to see a picture of the response. To identify the key nouns in the response, we will use a Text Generation widget; to see the results, we will use an Image Generation widget. we use a straightforward prompt and add the first widget.

10.We click the Retry icon to test it, and the result seems flawless.

11.We adjust the question a little and add the Image Generation widget. It takes us about a minute to get what we want.

After we satisfied with the software, we can Share and Make Public:

But Hold On There’s More

These are some that we didn’t get to:

Empty App: With our needs, we utilized the App Builder, but we can alternatively choose to Start from an Empty App, choose my widgets, and configure them as needed.

Remix: To modify or improve an existing app, either mine or another open source project, we can start with a remix.

Chatbot Widget: We can use a prompt to begin interacting with my app.

 Referencing: When developing my app, we can use the symbol “@” to make name-based references to other widgets.

Advanced parameters: A few widgets include the ability to adjust various parameters. For instance, we can adjust the model’s Heat and Top P parameters using the Text Generation widget.

Behind the scenes at PartyRock we can view my applications and total PartyRock credit consumption in Backstage.

You should be aware of the following two PartyRock facts:

Pricing: To help you start learning the basics without having to worry about spending money, AWS is now offering new PartyRock customers a free trial without requiring them to register for an AWS account or give a credit card. This offer is valid for a short period of time. As seen above, you can monitor how much credit you use in the Backstage. The data you provide in tokens, production tokens, and created photos are used to compute how much credit you use.

Model Access: Over time, AWS intend to grant you access to more models.

In Progress – Keep checking back for updates as AWS continue to develop new features and widgets.