Try it: Test model capabilities using demo playgrounds in Model Garden

This guide shows you how to test a model's capabilities using the free demo playgrounds in Model Garden. It covers the following topics:

  • Supported models: Learn which models offer a demo playground for testing.
  • Try out a model: Follow a step-by-step tutorial to send a prompt to the Gemma 2 model and view the output.

Model Garden hosts public demo playgrounds for supported models. The playgrounds are powered by predeployed Vertex AI online prediction endpoints. When you open the model card for a supported model, an embedded Try out panel lets you quickly test the model's capabilities by sending a text prompt and setting common parameters.

Supported models

The following models have demo playgrounds available.

Provider Models
Google
  • Gemma 2 2B it (instruction tuned)
  • Gemma 2 9B it (instruction tuned)
  • Gemma 2 27B it (instruction tuned)
  • Gemma 2B
  • Gemma 2B it (instruction tuned)
  • Gemma 7B
  • Gemma 7B it (instruction tuned)
Meta
  • Llama 3 8B Instruct
  • Llama 3 70B Instruct
  • Llama 2 7B
  • Llama 2 7B Chat
  • Llama 2 13B Chat
  • Llama 2 70B Chat (Int8)
  • Code Llama 7B Python
TII
  • Falcon 7B
Mistral AI
  • Mixtral 8x7B

Before you begin

Concepts

  • Model Garden: A platform in Vertex AI that helps you discover, test, customize, and deploy Google-proprietary and select open source generative AI models.
  • Gemma: A family of lightweight, state-of-the-art open models from Google.
  • Instruction-tuned models: Models that are fine-tuned to follow user-provided instructions, making them better at conversational and question-answering tasks. The model used in this guide, `Gemma-2b-it`, is instruction-tuned.

Prerequisites

This tutorial requires you to set up a Google Cloud project and enable the Vertex AI API.

  1. Sign in to your Google Cloud account. If you're new to Google Cloud, create an account to evaluate how our products perform in real-world scenarios. New customers also get $300 in free credits to run, test, and deploy workloads.

  2. In the Google Cloud console, on the project selector page, select or create a Google Cloud project.

    Go to project selector

  3. Verify that billing is enabled for your Google Cloud project.

  4. Enable the Vertex AI API.

    Enable the API

  5. In the Google Cloud console, on the project selector page, select or create a Google Cloud project.

    Go to project selector

  6. Verify that billing is enabled for your Google Cloud project.

  7. Enable the Vertex AI API.

    Enable the API

    8.

Try out Gemma 2

In this quickstart, you try out the Gemma-2b-it model. Note that -it stands for instruction-tuned.

  1. In the Google Cloud console, go to the Gemma 2 model card.

    Go to Gemma 2

  2. In the Try out panel:

    1. For Region, accept the default or choose your region.
    2. For Endpoint, select Demo playground (Free)2b-it.
    3. In the Prompt box, enter Why is the sky blue?.

    4. Expand the Advanced options section and view the default parameters.

      The try out panel for Gemma 2b-it

  3. Click Submit. The output appears below the Submit button.

Clean up

To avoid incurring charges to your Google Cloud account for the resources used on this page, follow these steps.

Delete the project

The easiest way to eliminate billing is to delete the project that you created for the tutorial.

To delete the project:

  1. In the Google Cloud console, go to the Manage resources page.

    Go to Manage resources

  2. In the project list, select the project that you want to delete, and then click Delete.
  3. In the dialog, type the project ID, and then click Shut down to delete the project.

What's next

See an overview of Model Garden.