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 d40a8d6bb0 chore: bind single_query_cached_kv_attention to python před 1 rokem
aphrodite b48fe85378 chore: utilities for modeling před 1 rokem
assets fefbf029c9 revert previous commit před 1 rokem
kernels d40a8d6bb0 chore: bind single_query_cached_kv_attention to python před 1 rokem
.gitignore 3c3944153c feat: add generic attention and FP32 dtype kernels před 1 rokem
LICENSE fefbf029c9 revert previous commit před 1 rokem
README.md fefbf029c9 revert previous commit před 1 rokem
requirements.txt fefbf029c9 revert previous commit před 1 rokem

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.