Skip to content

Build LLM apps in Typescript/Javascript. πŸ§‘β€πŸ’» πŸ§‘β€πŸ’» πŸ§‘β€πŸ’» πŸš€ πŸš€ πŸš€

License

Notifications You must be signed in to change notification settings

anupom/promptable

Β 
Β 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 

Repository files navigation

Promptable

Build fullstack AI apps in Typescript/Javascript.

What is this?

Promptable is library that enables you to build powerful AI applications with LLMs and Embeddings providers such as OpenAI, Hugging Face, Cohere and Anthropic. It provides a flexible and extensible API that makes it easy to compose LLMs with data and tools to build complex applications quickly and easily.

With Promptable, you can combine LLMs with other powerful tools and data sources, such as databases and APIs, to create a wide range of AI applications.

What is LLM?

An LLM is a Large Language Model. It is a model that can generate text given a prompt. It is a type of AI that can be used to generate text, answer questions, and more.

What is an Embedding?

An Embedding is a vector representation of a piece of text. It is a type of AI that can be used to find similar pieces of text, search for text, and more.

! Right now the library is in early development and is very much experimental. Don't use this in production yet! The API is subject to change as we get feedback.

Github Repo. Discord. Twitter.

Use Cases:

  • πŸ’¬ Chatbots & Conversational AI
  • ❓ Question Answering Bots
  • ✍️ Writing Apps
  • πŸ§‘β€βœˆοΈ Copilot apps built with Chrome Extensions, VSCode Extensions, and more!
  • πŸ” AI Semantic Search apps
  • πŸ› οΈ AI first Automations, Workflows and Tools
  • πŸ€– Autonomous Agents & Personal Assistants

Features

  • StarterKits ex. NextJS <-> Promptable Starter kit with Streaming Chat Bot!
  • Prompts for templating and formatting
  • Model Providers for Text Generation and Embedding Generation
  • Embeddings for creating Embeddings, Indexing and Search
  • Tracing for debugging your applications!
  • Chains for composing LLMs and Embeddings with data and other tools.
  • Utilities for working with text and data.
  • More Coming Soon :)

Install Library

npm i promptable

Usage

See our docs for more info on how to use the library. Documentation

Examples

To run an example, clone the repo and follow these steps

  1. Copy contents of /apps/web/.env.example into a new file, /apps/web/.env
  2. Copy contents of /examples/.env.example into a new file, /examples/.env and add your OpenAI API Key to the OPENAI_API_KEY field
  3. Run the following commands:
pnpm i
pnpm dev
  1. Run the following command in a different terminal window/tab: *NOTE: example names found in /examples/src
pnpm run example <example name>

What's inside?

This a Turborepo monorepo of tooling for Typescript developers building LLM apps.

It uses pnpm as a package manager and includes the following packages/apps:

  • packages/promptable: The Promptable Library for building LLM apps in Typescript / Javascript!
  • examples: Examples using the Promptable.js library!
  • apps/docs: The Promptable.js Docs
  • apps/web: A nextjs app for visualizing Promptable.js steps.
  • apps/nextjs-promptable: A nextjs app starter kit with prebuilt features (ChatBot with Streaming, QA Bot etc.)

Contributing

Install and Run

Then to install run (at the root)

pnpm i

To install package in a single workspace


pnpm i <package> --filter <workspace>

First copy the apps/web/.env.example file to apps/web/.env.

Then, To develop all apps and packages, run the following command:


cd my-turborepo
pnpm run dev

This will start watching the files for changes.

UI

See our Docs.

About

Build LLM apps in Typescript/Javascript. πŸ§‘β€πŸ’» πŸ§‘β€πŸ’» πŸ§‘β€πŸ’» πŸš€ πŸš€ πŸš€

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • TypeScript 86.8%
  • JavaScript 9.3%
  • CSS 3.1%
  • Shell 0.8%