Skip to content

Latest commit

 

History

History
executable file
·
62 lines (37 loc) · 2.73 KB

File metadata and controls

executable file
·
62 lines (37 loc) · 2.73 KB

Claude 3.5 Streamlit Multimodal LLM Playground

Overview

This repository contains a Streamlit-based application for interacting with Claude 3 and 3.5, a powerful language model developed by Anthropic. The application provides a user-friendly interface for exploring the capabilities of Claude 3, supporting both text-only and image+text inputs, custom system prompts, and tool usage (function calling).

Application Versions

We offer two versions of this application, each tailored to different deployment scenarios:

  1. Anthropic API Version
  2. Amazon Bedrock Version

Anthropic API Version

This version interacts directly with the Anthropic API. It's suitable for users who have an Anthropic API key and want to use Claude 3 models without additional cloud infrastructure.

Learn more about the Anthropic API version

Amazon Bedrock Version

This version is designed to work with Claude 3 and 3.5 models through Amazon Bedrock. It's ideal for users who prefer to leverage AWS infrastructure or have existing workflows in the AWS ecosystem.

Learn more about the Amazon Bedrock version

Key Features

Both versions of the application offer:

  • Support for multiple Claude 3 models (Opus, Sonnet, Haiku)
  • Text-only and image+text query capabilities
  • Customizable system prompts
  • Tool definition for function calling
  • Temperature and max tokens adjustment
  • Multiple text input fields for complex prompts
  • Image upload functionality
  • Display of AI responses in markdown format
  • Visualization of tool use results in JSON format
  • Token usage display

Choosing a Version

  • Choose the Anthropic API version if you have direct access to the Anthropic API and prefer a standalone solution.
  • Opt for the Amazon Bedrock version if you're using AWS services or need the additional features and integrations provided by Amazon Bedrock.

Please refer to the README files in each respective folder for detailed setup instructions, usage guidelines, and version-specific information.

Combine Text with Vision to create multimodal inputs for Claude 3.5

Supports Image and text input

Test with your own Tool definitions

Supports Tool Use

Contributing

We welcome contributions to both versions of the application. Please read our CONTRIBUTING.md file for guidelines on how to submit contributions.

License

This project is licensed under the MIT-0 License - see the LICENSE.md file for details.