Welcome to Lexer!

We build tools designed to help machine learning engineers code models with confidence. By using our library, you can:

• Compare machine learning execution engines
• Understand model performance before deploying to production
• And save yourself and colleagues countless hours of headaches ensuring models perform as expected!

Product Overview

Lexer's primary product is a Python package featuring a decorator library. The Lexer class provides an easy interface for benchmarking your model outputs. This enables an easy comparison of the model when its executed using different runtimes.

Documentation

• The Lexer Decorator
• Configuring Lexer

Tutorials

• Using Lexer for Model Benchmarking

Installing Lexer

The Lexer Python library can be install via pip:

pip install lexer-sdk

For additional details on the library's version history, see the PyPi project page.

Future Plans

The Lexer product suite will be consistently evolving, with planned support for additional:

• Machine learning frameworks
• Model operations
• Hardware backends

Most importantly we will build what people need, so we welcome anyone to reach out to us!

Get in Touch

We love to hear from our users. If you have any questions, feature requests, or just want to talk to the team, don't hesitate to join our Discord server or email us at [email protected]!