Skip to content
View R3hankhan123's full-sized avatar
7 followers · 4 following

Achievements

Achievement: Pair ExtraordinaireAchievement: Pull Sharkx2Achievement: Quickdraw

Achievements

Achievement: Pair ExtraordinaireAchievement: Pull Sharkx2Achievement: Quickdraw

Highlights

  • Pro

Block or report R3hankhan123

Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
R3hankhan123/README.md

Hey 👋, I'm Rehan Khan

🔧 I build systems that run models not demo notebooks that break outside Jupyter

🧠 What I Actually Do

  • 🚀 Make LLMs run on spyre accelerator and CPUs (yes CPUs)
  • 🔩 Bend vLLM into places it wasn't designed for
  • 🏗️ Fix build systems across architectures
  • 📦 Make multi-arch containers actually behave

🔭 Currently Building

  • ⚡ CPU-only LLM inference pipelines
  • 🏛️ s390x(cpu) and spyre support for modern ML stacks
  • ☸️ Infra that works beyond a single machine

⚙️ Current Obsessions

🔥 squeezing every drop of performance.
⚠️  making PyTorch do questionable things
☸️  running clean infra on Kubernetes

🛠️ Languages & Tools

📊 GitHub Stats

r3hankhan123

r3hankhan123

🤝 Connect With Me

linkedin

Pinned Loading

  1. containerd containerd Public

    Forked from containerd/containerd

    An open and reliable container runtime

    Go

  2. torch-spyre torch-spyre Public

    Forked from torch-spyre/torch-spyre

    C++

  3. pytorch/pytorch pytorch/pytorch Public

    Tensors and Dynamic neural networks in Python with strong GPU acceleration

    Python 100k 27.9k

  4. vllm-project/vllm vllm-project/vllm Public

    A high-throughput and memory-efficient inference and serving engine for LLMs

    Python 80.7k 17.1k