🚀 Big News: Socket Acquires Coana to Bring Reachability Analysis to Every Appsec Team.Learn more
Socket
DemoInstallSign in
Socket

promptcompression

Package Overview
Dependencies
Maintainers
1
Alerts
File Explorer

Advanced tools

Socket logo

Install Socket

Detect and block malicious and high-risk dependencies

Install

promptcompression

Taking text input and compressing it into a smaller representation for cost savings.

0.1.4
PyPI
Maintainers
1

LLM Prompt Compression

Are LLM's costing you too much in your application? This library will allow your prompts to have the minimum amount of text needed to convey what you want the LLM to do. This is perfect for prompts given by users, where they are constantly eating at your costs.

What is this suitable for?

  • Chat bots
  • Local LLMS
  • Input that does not need to have a strict structure
  • Reducing text which reduces tokens which reduces $$$

What is this not suitable for?

  • Code
  • Math
  • Inputs that require structure, like reviewing a writing sample

Install

This package is available on PyPI:

$ pip install promptcompression

Usage

This is very simple to use, because it only requires a library import and a function wrapper.

You use the compress() function, which has the argument text for the input prompt, and the optional argument savings which, when True, returns a tuple of the number of tokens saved, and the ratio of tokens saved by the original tokens.

Example:

from promptcompression import compress

prompt, tokens, ratio = compress("Summarize this text while sounding more natural. Please do not use contractions or an em dash to be more human-like.", savings = True)

print("Compressed text:", prompt)
print("Tokens saved:", tokens)
print(f"Compression ratio: {ratio:.2f}")

>>> Compressed text: summar text sound natur pleas dont use contract em dash humanlik
>>> Tokens saved: 13
>>> Compression ratio: 0.50

If you want to learn more...

... please check out the markdown file WHATWASITHINKING.md, where I explain my process throught the project. This is located in the library's GitHub repo.

Keywords

compression

FAQs

Did you know?

Socket

Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.

Install

Related posts