Stars
GPUd automates monitoring, diagnostics, and issue identification for GPUs
Curated collection of papers in machine learning systems
Writing an OS in 1,000 lines.
Python Fire is a library for automatically generating command line interfaces (CLIs) from absolutely any Python object.
High performance container overlay networks on Linux. Enabling RDMA (on both InfiniBand and RoCE) and accelerating TCP to bare metal performance. Freeflow requires zero modification on application …
Hackable and optimized Transformers building blocks, supporting a composable construction.
A scalable generative AI framework built for researchers and developers working on Large Language Models, Multimodal, and Speech AI (Automatic Speech Recognition and Text-to-Speech)
A Python-embedded modeling language for convex optimization problems.
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Optimized primitives for collective multi-GPU communication
OpenMMLab Pre-training Toolbox and Benchmark
Pytorch Implementation of DQN / DDQN / Prioritized replay/ noisy networks/ distributional values/ Rainbow/ hierarchical RL
Code for "Heterogenity-Aware Cluster Scheduling Policies for Deep Learning Workloads", which appeared at OSDI 2020
A high performance and generic framework for distributed DNN training
NUMA support for Linux
Example models using DeepSpeed
Multi-Instance-GPU profiling tool
Running large language models on a single GPU for throughput-oriented scenarios.
Distributed ML Training and Fine-Tuning on Kubernetes
DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.