Python package for LLM compression

Serverless LLM Serving for Everyone.

A framework for performing live pre-recorded command line demos in the wild 📼

My experiments with LLMs and MCP

Kepler (Kubernetes-based Efficient Power Level Exporter) is a Prometheus exporter that measures energy consumption metrics at the container, pod, and node levels in Kubernetes clusters.

Carbon Limiting Auto Tuning for Kubernetes

Model Server for Kepler

Kepler Estimator

Slim: OS Kernel Support for a Low-Overhead Container Overlay Network