jianzi123's repositories
adaptdl
Resource-adaptive cluster scheduler for deep learning training.
AFFiNE
There can be more than Notion and Miro. AFFiNE(pronounced [ə‘fain]) is a next-gen knowledge base that brings planning, sorting and creating all together. Privacy first, open-source, customizable and ready to use.
algorithm_design
Use several algorithm design methods to solve several common problems with C++11.
cricket
cricket is a virtualization solution for GPUs
cuda-api-wrappers
Thin, unified, C++-flavored wrappers for the CUDA APIs
cuda_hook
Hooked CUDA-related dynamic libraries by using automated code generation tools.
cuda_scheduling_examiner_mirror
A tool for examining GPU scheduling behavior.
DeepLearningSystem
Deep Learning System core principles introduction.
DeepSpeed
DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.
FleetX
Paddle Distributed Training Extended. 飞桨分布式训练扩展包
godel-scheduler
an unified scheduler for online and offline tasks
h2ogpt
Private Q&A and summarization of documents+images or chat with local GPT, 100% private, Apache 2.0. Supports LLaMa2, llama.cpp, and more. Demo: https://gpt.h2o.ai/ https://codellama.h2o.ai/
HiCCL
A hierarchical collective communications library with portable optimizations
kluster-capacity
Cluster capacity analysis tool for capacity estimation、scheduler simulation、cluster compression、fragmentation etc.
kubernetes-1
Apuntes sobre k8s
langchain
⚡ Building applications with LLMs through composability ⚡
learn-vgpu-the-hard-way
qemu, cuda, virtio and kernel driver etc, none of which I understand, I just in awe.
ML-Papers-of-the-Week
🔥Highlighting the top ML papers every week.
nvidia-patch
This patch removes restriction on maximum number of simultaneous NVENC video encoding sessions imposed by Nvidia to consumer-grade GPUs.
pipedream_experiment
private repo of msr-fiddle/pipedream
ray
Ray is a unified framework for scaling AI and Python applications. Ray consists of a core distributed runtime and a toolkit of libraries (Ray AIR) for accelerating ML workloads.
unsloth
Finetune Llama 3.1, Mistral, Phi & Gemma LLMs 2-5x faster with 80% less memory
vgpu_unlock
Unlock vGPU functionality for consumer grade GPUs.