Skip to main content

Documentation Index

Fetch the complete documentation index at: https://langbear.runbear.io/llms.txt

Use this file to discover all available pages before exploring further.

LangBear introduces a web user interface that enables easy creation and management of prompt templates. Users can conveniently draft, alter, and organize prompt templates on the platform. This practical feature streamlines the prompt development process, ensuring a user-friendly experience. Template management

​
Creating a new prompt template

By clicking the New Prompt button, you can create a new template. There are two primary types of prompt templates: llm and chat.

​
LLM-type blocks

LLM-type prompt templates Supported blocks:
  • Prompt Block - A block for plain text prompt message.
  • FewShotExamples Block - A list of examples for few-shot learning.

​
Chat-type blocks

Chat-type prompt templates Supported blocks:
  • SystemMessage Block - A message for the system.
  • HumanMessage Block - A request message given by a human.
  • AIMessage Block - A response message given by AI assistant.
  • FewShotExamples Block - A list of examples for few-shot learning.

​
Tracking versions of prompt templates

LangBear offers the feature of multiple prompt versions, which greatly enhances flexibility. It encourages users not to shy away from modifying prompts. This functionality provides the advantage of comparing different prompt versions and switching to another if desired.

​
Locallizing prompt templates

You can manage different prompts for each language. While Language Learning Models (LLMs) support content generation in various languages, supplying prompts in different languages enhances the user experience.