Metadata-Version: 2.4
Name: tempo9
Version: 0.0.1
Summary: Reserved for Tempo9 — server-grade LLM inference engine (Apple Silicon and CUDA). Real releases will appear here.
Author: Jiejing Zhang
License: Apache-2.0
Project-URL: Homepage, https://github.com/thinkspread
Classifier: Development Status :: 1 - Planning
Requires-Python: >=3.8
Description-Content-Type: text/markdown

# tempo9

This name is reserved for **Tempo9**, a server-grade LLM inference engine
for Apple Silicon and CUDA — continuous batching, paged KV cache, GGUF
direct load, speculative decoding.

Real releases will appear here. Until then, nothing to install.
