Conversational interaction with large language model (LLM) based solutions (for example, a chatbot) is quite common. Although production grade LLMs are trained using a huge corpus of data, their knowledge base is inherently limited to the information present in the training data, and they may not possess real-time or the most up-to-date knowledge.
This web application in this repo can be deployed to AWS App Runner using AWS CDK. You can then use it to ask questions based on the content of a link/URL of your choice.
The application is written in Go, but the concepts apply to any other language you might choose. It uses langchaingo
as the framework to interact with the Anthropic Claude (v2) model on Amazon Bedrock. The web app uses the Go embed package to serve the static file for the frontend part (HTML, JavaScript and CSS) from directly within the binary.
See CONTRIBUTING for more information.
This library is licensed under the MIT-0 License. See the LICENSE file.