helm repo add daocloud https://daocloud.github.io/dce-charts-repackage/

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

OpenAIOS vGPU scheduler for Kubernetes is originated from the OpenAIOS project to virtualize GPU device memory.

SGLang is a fast serving framework for large language models and vision language models.

Intelligent Mixture-of-Models Router for Efficient LLM Inference

Minimalist ML framework for Rust

Redis for LLMs

A Datacenter Scale Distributed Inference Serving Framework

ModelScope: bring the notion of Model-as-a-Service to life.

Dynamic Resource Allocation (DRA) for NVIDIA GPUs in Kubernetes

Community maintained hardware plugin for vLLM on Ascend

HAMi-core compiles libvgpu.so, which ensures hard limit on GPU in container

contaiNERD CTL - Docker-compatible CLI for containerd, with support for Compose, Rootless, eStargz, OCIcrypt, IPFS, ...

vLLM’s reference system for K8S-native cluster-wide deployment with community-driven performance optimization

An open and reliable container runtime

Production-Grade Container Scheduling and Management

A Friendly Kubernetes CloudShell (Web Terminal) !

A shim driver allows in-docker nvidia-smi showing correct process list without modify anything

A bidirectional pipeline parallelism algorithm for computation-communication overlap in V3/R1 training.

SCUDA is a GPU over IP bridge allowing GPUs on remote machines to be attached to CPU-only machines.