PygmalionAI's large-scale inference engine
pygmalion.chat

It is designed to serve as the inference endpoint for the PygmalionAI website, and to allow serving the Pygmalion models to a large number of users with blazing fast speeds (thanks to vLLM's Paged Attention).

AlpinDale 28866137ea feat: add swiglu activation il y a 1 an
aphrodite b48fe85378 chore: utilities for modeling il y a 1 an
assets fefbf029c9 revert previous commit il y a 1 an
kernels 28866137ea feat: add swiglu activation il y a 1 an
.gitignore 3c3944153c feat: add generic attention and FP32 dtype kernels il y a 1 an
LICENSE fefbf029c9 revert previous commit il y a 1 an
README.md fefbf029c9 revert previous commit il y a 1 an
requirements.txt fefbf029c9 revert previous commit il y a 1 an

README.md

Aphrodite - The Pygmalion Backend

Work in Progress

aphrodite

Aphrodite is the backend service for PygmalionAI, built on top of FastChat, vLLM, SkyPilot, and more.

Currently a work in progress, not remotely functional.