Yineng Zhang

Inference Lead at SGLang; Research Scientist at Together AI

Starred Projects (123)

iree by iree-org

MLIR-based compiler and runtime toolkit for machine learning models

ogabrielluiz:

mfuntowicz:

parano:

infwinston:

Created 6 years ago

Updated 17 hours ago

nanobot by HKUDS

Personal AI assistant for task automation and knowledge management

transitive-bullshit:

jmorganca:

dguido:

Created 2 months ago

Updated 16 hours ago

Kimi-K2.5 by MoonshotAI

Multimodal agentic AI for vision-grounded reasoning and task execution

Created 2 months ago

Updated 2 months ago

hpc-ops by Tencent

Boosts LLM inference speed with production-grade operators

Created 3 months ago

Updated 1 week ago

SkyRL by NovaSky-AI

RL training pipeline for multi-turn tool use LLMs, optimized for real-world tasks

infwinston:

lewtun:

hiyouga:

WoosukKwon:

Created 1 year ago

Updated 20 hours ago

Engram by deepseek-ai

Scalable conditional memory for large language models

pgarbacki:

Created 3 months ago

Updated 3 months ago

codex by openai

Coding agent CLI tool for terminal-based chat-driven development

codekansas:

hsbt:

didierrlopes:

ValentaTomas:

Created 1 year ago

Updated 16 hours ago

TileGym by NVIDIA

CUDA Tile kernel library for efficient GPU programming

zhuohan123:

Created 5 months ago

Updated 3 days ago

xllm by jd-opensource

LLM inference engine optimized for diverse AI accelerators

Created 8 months ago

Updated 19 hours ago

cocoon by TelegramMessenger

Confidential AI inference on a decentralized network

Created 4 months ago

Updated 6 days ago

DeepSeek-Math-V2 by deepseek-ai

LLM for self-verifiable mathematical reasoning

chiphuyen:

vincentweisser:

omarsar:

simonw:

Created 4 months ago

Updated 4 months ago

open-instruct by allenai

Training codebase for instruction-following language models

hammer:

vincentweisser:

RJT1990:

pgarbacki:

Created 2 years ago

Updated 19 hours ago

pplx-garden by perplexityai

High-performance LLM inference engine

Created 5 months ago

Updated 3 months ago

KernelAgent by meta-pytorch

Autonomous GPU kernel generation and optimization via AI agents

winglian:

Created 9 months ago

Updated 6 days ago

tunix by google

JAX-native library for efficient LLM post-training

ekzhang:

hiyouga:

ebursztein:

hammer:

Created 1 year ago

Updated 1 day ago

kimi-cli by MoonshotAI

AI agent enhancing terminal operations and software development workflows

joewalnes:

parano:

omarsar:

Created 6 months ago

Updated 16 hours ago

DeepSeek-OCR by deepseek-ai

Context-aware OCR model for visual-text compression

pgarbacki:

omarsar:

MagMueller:

shizhediao:

Created 6 months ago

Updated 2 months ago

modular by modular

AI toolchain unifying fragmented AI deployment workflows

lattner:

tobi:

parano:

flaque:

Created 3 years ago

Updated 17 hours ago

FastVideo by hao-ai-lab

Framework for accelerated video generation

zhuohan123:

hiyouga:

luiscape:

Created 1 year ago

Updated 2 days ago

ThunderKittens by HazyResearch

CUDA kernel framework for fast deep learning primitives

karpathy:

codekansas:

vincentweisser:

gakonst:

Created 2 years ago

Updated 1 week ago

tinker-cookbook by thinking-machines-lab

Advanced LLM fine-tuning SDK and example cookbook

pcmoritz:

Jiayi-Pan:

hammer:

ekzhu:

Created 9 months ago

Updated 1 day ago

DeepSeek-V3.2-Exp by deepseek-ai

Experimental LLM boosting long-context efficiency

Jiayi-Pan:

merrymercy:

Created 6 months ago

Updated 5 months ago

K2-Vendor-Verifier by MoonshotAI

Verifying Kimi K2 API vendor toolcall precision and reliability

hiyouga:

t3dotgg:

Created 7 months ago

Updated 2 months ago

sglang-jax by sgl-project

High-performance LLM inference engine for JAX/TPU serving

mattjj:

froystig:

hammer:

merrymercy:

Created 9 months ago

Updated 1 day ago

checkpoint-engine by MoonshotAI

Middleware for efficient LLM weight updates during inference

pgarbacki:

zhuohan123:

luiscape:

didierrlopes:

Created 7 months ago

Updated 1 month ago

batch_invariant_ops by thinking-machines-lab

Enhance LLM inference determinism

Edward-Sun:

ekzhang:

willccbb:

zhuohan123:

Created 7 months ago

Updated 5 months ago

xtuner by InternLM

LLM fine-tuning toolkit for research

casper-hansen:

winglian:

lysandrejik:

JustinLin610:

Created 2 years ago

Updated 17 hours ago

agents.md by agentsmd

A format for guiding coding agents

dror-weiss:

gregpr07:

transitive-bullshit:

romainhuet:

Created 8 months ago

Updated 1 month ago

aiconfigurator by ai-dynamo

LLM serving configuration optimization

pgarbacki:

Created 9 months ago

Updated 16 hours ago

axolotl by axolotl-ai-cloud

CLI tool for streamlined post-training of AI models

tobi:

jn2clark:

beyang:

patrickvonplaten:

Created 3 years ago

Updated 17 hours ago

DiffSynth-Studio by modelscope

Open-source project for diffusion model exploration

rodrigosnader:

doriandarko:

Created 2 years ago

Updated 17 hours ago

coze-studio by coze-dev

AI agent development platform with visual tools

Created 9 months ago

Updated 16 hours ago

Step3 by stepfun-ai

Multimodal reasoning model for efficient vision-language tasks

Created 8 months ago

Updated 8 months ago

SpecForge by sgl-project

Train speculative decoding models for faster inference

merrymercy:

xiezhq-hermann:

Ying1123:

Created 10 months ago

Updated 2 weeks ago

LightX2V by ModelTC

Video generation inference framework for efficient synthesis

jiamings:

Created 1 year ago

Updated 18 hours ago

chatterbox by resemble-ai

Open-source TTS model

transitive-bullshit:

ebursztein:

gakonst:

dguido:

Created 1 year ago

Updated 3 weeks ago

Kimi-K2 by MoonshotAI

State-of-the-art MoE language model

krisr:

lucidrains:

merrymercy:

shizhediao:

Created 9 months ago

Updated 2 months ago

gemini-cli by google-gemini

AI agent for terminal workflows

ValentaTomas:

MagMueller:

lucidrains:

khou22:

Created 1 year ago

Updated 17 hours ago

trae-agent by bytedance

LLM-powered CLI for software engineering tasks

chiphuyen:

hiyouga:

yiranwu0:

syrusakbary:

Created 10 months ago

Updated 2 months ago

Math-Verify by huggingface

Math evaluator for LLM outputs in mathematical tasks

shizhediao:

thomwolf:

hiyouga:

lewtun:

Created 1 year ago

Updated 3 months ago

Skills by NVIDIA-NeMo

LLM skill-improvement pipelines for synthetic data generation, training, and evaluation

winglian:

Created 2 years ago

Updated 1 day ago

RL-Factory by Simple-Efficient

RL post-training framework for agentic learning

Created 11 months ago

Updated 4 months ago

genai-bench by sgl-project

LLM serving performance benchmarking

merrymercy:

Ying1123:

Created 10 months ago

Updated 3 days ago

minions by HazyResearch

Communication protocol for cost-efficient LLM collaboration

JohannesHa:

w4nderlust:

jmorganca:

mchiang0610:

Created 1 year ago

Updated 1 month ago

slime by THUDM

LLM post-training framework for RL scaling

luiscape:

jiamings:

hammer:

pgarbacki:

Created 10 months ago

Updated 19 hours ago

Hunyuan3D-2.1 by Tencent-Hunyuan

Image to 3D asset generation with PBR materials

zhuohan123:

Created 10 months ago

Updated 6 months ago

dots.llm1 by rednote-hilab

MoE model for research

Created 11 months ago

Updated 8 months ago

ROLL by alibaba

RL library for large language models

chiphuyen:

winglian:

yiranwu0:

pcmoritz:

Created 10 months ago

Updated 23 hours ago

cosmos-reason1 by nvidia-cosmos

Multimodal LLMs for physical common sense and embodied decisions

Created 1 year ago

Updated 3 months ago

torchtune by meta-pytorch

PyTorch library for LLM post-training and experimentation

lewtun:

patrickvonplaten:

JustinLin610:

shizhediao:

Created 2 years ago

Updated 1 day ago

ome by sgl-project

Kubernetes operator for LLM serving

merrymercy:

Ying1123:

Created 11 months ago

Updated 2 days ago

unsloth by unslothai

Finetuning tool for LLMs, targeting speed and memory efficiency

tobi:

karpathy:

alexchen4ai:

danielhanchen:

Created 2 years ago

Updated 18 hours ago

RL by NVIDIA-NeMo

Scalable toolkit for efficient model reinforcement learning

vincentweisser:

hammer:

wangshangsam:

pcmoritz:

Created 1 year ago

Updated 17 hours ago

VeOmni by ByteDance-Seed

Framework for scaling multimodal model training across accelerators

hiyouga:

JustinLin610:

Created 1 year ago

Updated 18 hours ago

MiMo by XiaomiMiMo

LLM for reasoning, pre-trained and post-trained for math/code tasks

alexchen4ai:

WoosukKwon:

omarsar:

Created 11 months ago

Updated 10 months ago

dia by nari-labs

TTS model for ultra-realistic dialogue generation

shimmyshimmer:

alexchen4ai:

omarsar:

codekansas:

Created 1 year ago

Updated 5 months ago

SpargeAttn by thu-ml

Training-free sparse attention for model inference acceleration

Created 1 year ago

Updated 1 month ago

profile-data by deepseek-ai

Profiling data for computation-communication overlap analysis

Created 1 year ago

Updated 1 year ago

multi-swe-bench by multi-swe-bench

Multilingual benchmark for LLM-powered code issue resolution

Created 1 year ago

Updated 4 months ago

aider by Aider-AI

AI pair programming in your terminal

bcherny:

nat:

hammer:

ekzhu:

Created 2 years ago

Updated 1 week ago

Triton-distributed by ByteDance-Seed

Distributed compiler for computation-communication overlapping, based on Triton

antiagainst:

Created 1 year ago

Updated 3 days ago

AReaL by inclusionAI

Distributed RL system for LLM reasoning

lewtun:

yiranwu0:

casper-hansen:

merrymercy:

Created 1 year ago

Updated 16 hours ago

dynamo by ai-dynamo

Inference framework for distributed generative AI model serving

vincentweisser:

willingc:

chiphuyen:

luiscape:

Created 1 year ago

Updated 17 hours ago

QwQ by QwenLM

Reasoning model for complex problem-solving, based on Qwen2.5

shimmyshimmer:

huybery:

JustinLin610:

Created 1 year ago

Updated 1 year ago

3FS by deepseek-ai

Distributed file system for AI training/inference workloads

chiphuyen:

joewalnes:

wesm:

alexey-milovidov:

Created 1 year ago

Updated 3 weeks ago

EPLB by deepseek-ai

Load balancer for expert-parallel MoE models

parano:

comaniac:

Created 1 year ago

Updated 1 year ago

DualPipe by deepseek-ai

Pipeline parallelism algorithm for training large models

luiscape:

JohannesHa:

Created 1 year ago

Updated 3 months ago

DeepGEMM by deepseek-ai

CUDA library for efficient FP8 GEMM kernels with fine-grained scaling

chiphuyen:

ekzhang:

natolambert:

patrickvonplaten:

Created 1 year ago

Updated 3 days ago

DeepEP by deepseek-ai

Expert-parallel communication library for MoE, targeting high-throughput and low-latency

hammer:

parano:

JustinLin610:

CodeCreator:

Created 1 year ago

Updated 4 days ago

FlashMLA by deepseek-ai

Efficient CUDA kernels for MLA decoding

chiphuyen:

hammer:

jmorganca:

shizhediao:

Created 1 year ago

Updated 1 week ago

Moonlight by MoonshotAI

MoE model and optimizer tech report

jph00:

lucidrains:

Created 1 year ago

Updated 8 months ago

open-infra-index by deepseek-ai

AI infrastructure tools for efficient AGI development

lattner:

vincentweisser:

hammer:

pankajroark:

Created 1 year ago

Updated 11 months ago

ktransformers by kvcache-ai

Framework for LLM inference optimization experimentation

parano:

luiscape:

patrickvonplaten:

JustinLin610:

Created 1 year ago

Updated 19 hours ago

skypilot by skypilot-org

Framework for cloud AI/batch jobs, unifying execution across diverse infrastructure

karpathy:

tobi:

amin3141:

luiscape:

Created 4 years ago

Updated 16 hours ago

ZhiLight by zhihu

LLM inference engine for Llama and variants, optimized for PCIe GPUs

hiyouga:

Created 1 year ago

Updated 1 month ago

Janus by deepseek-ai

Unified multimodal model research paper for understanding and generation

jiamings:

omarsar:

taranjeet:

xiaofan-luan:

Created 1 year ago

Updated 1 year ago

UI-TARS by bytedance

Multimodal agent for GUI interaction in virtual worlds (research paper)

jiamings:

hugs:

jmorganca:

omarsar:

Created 1 year ago

Updated 2 months ago

rtp-llm by alibaba

LLM inference engine for diverse applications

chiphuyen:

Created 2 years ago

Updated 16 hours ago

tilelang by tile-ai

DSL for high-performance GPU/CPU kernel development (GEMM, attention, etc.)

parano:

hiyouga:

JustinLin610:

luiscape:

Created 1 year ago

Updated 16 hours ago

DeepSeek-R1 by deepseek-ai

Reasoning models research paper

shimmyshimmer:

rasbt:

transitive-bullshit:

omarsar:

Created 1 year ago

Updated 9 months ago

MInference by microsoft

Framework for long-context LLM inference speedup via sparse attention

pgarbacki:

Created 1 year ago

Updated 1 week ago

flash-linear-attention by fla-org

Efficient Torch/Triton implementations for linear attention models

hiyouga:

xiezhq-hermann:

yang-song:

winglian:

Created 2 years ago

Updated 1 day ago

MiniMax-01 by MiniMax-AI

Large language & vision-language models based on linear attention

sxyu:

transitive-bullshit:

omarsar:

hiyouga:

Created 1 year ago

Updated 9 months ago

dash-infer by modelscope

LLM inference engine for optimized performance across diverse hardware

Created 2 years ago

Updated 8 months ago

smoothquant by mit-han-lab

Post-training quantization research paper for large language models

zhiyuan8:

gakonst:

forresti:

Edward-Sun:

Created 3 years ago

Updated 1 year ago

picotron by huggingface

Minimalist distributed training framework for educational use

karpathy:

transitive-bullshit:

srush:

pgarbacki:

Created 1 year ago

Updated 7 months ago

hqq by dropbox

Model quantizer for fast, accurate post-training quantization, skipping calibration

danielhanchen:

winglian:

hiyouga:

osanseviero:

Created 2 years ago

Updated 1 month ago

torchtitan by pytorch

PyTorch platform for generative AI model training research

karpathy:

WoosukKwon:

pgarbacki:

lewtun:

Created 2 years ago

Updated 22 hours ago

LightLLM by ModelTC

Python framework for LLM inference and serving

chiphuyen:

pgarbacki:

JustinLin610:

winglian:

Created 2 years ago

Updated 16 hours ago

cline by cline

VS Code extension for autonomous coding agent

didierrlopes:

swyxio:

snowzurfer:

ishaan-jaff:

Created 1 year ago

Updated 22 hours ago

DeepSeek-V3 by deepseek-ai

MoE language model research paper with 671B total parameters

tobi:

shimmyshimmer:

jiamings:

syrusakbary:

Created 1 year ago

Updated 7 months ago

infinity by michaelfeil

REST API for high-throughput, low-latency embedding and reranking

luiscape:

ishaan-jaff:

ankane:

jn2clark:

Created 2 years ago

Updated 3 weeks ago

Star-Attention by NVIDIA

PyTorch code for efficient LLM inference on long sequences

Created 1 year ago

Updated 9 months ago

cudnn-frontend by NVIDIA

C++ and Python interface for NVIDIA cuDNN and high-performance kernels

hammer:

antiagainst:

Created 5 years ago

Updated 3 days ago

Qwen3-Coder by QwenLM

Code LLM for code completion, generation, and assistant use cases

hugs:

victortaelin:

luiscape:

cournape:

Created 2 years ago

Updated 3 weeks ago

xgrammar by mlc-ai

Library for efficient structured generation

merrymercy:

ogabrielluiz:

youkaichao:

simonw:

Created 1 year ago

Updated 3 days ago

prompt-lookup-decoding by apoorvumang

Decoding method for faster LLM generation

stas00:

jph00:

casper-hansen:

ggerganov:

Created 2 years ago

Updated 1 year ago

verl by verl-project

RL training library for LLMs

WoosukKwon:

hammer:

yiranwu0:

luiscape:

Created 1 year ago

Updated 16 hours ago

duo-attention by mit-han-lab

Framework for efficient long-context LLM inference

Created 1 year ago

Updated 1 year ago

corenet by apple

DNN toolkit for training standard and novel models

transitive-bullshit:

omarsar:

t3dotgg:

hammer:

Created 2 years ago

Updated 6 months ago

tensorrtllm_backend by triton-inference-server

Triton backend for serving TensorRT-LLM models

NikolaBorisov:

guberti:

Ying1123:

tuhins:

Created 2 years ago

Updated 4 days ago

Pai-Megatron-Patch by alibaba

Training toolkit for LLMs & VLMs using Megatron

Jiayi-Pan:

Created 2 years ago

Updated 4 months ago

whisper by openai

Speech recognition model for multilingual transcription/translation

bcherny:

karpathy:

ekzhu:

gregpr07:

Created 3 years ago

Updated 5 days ago

TensorRT-LLM by NVIDIA

LLM inference optimization SDK for NVIDIA GPUs

beyang:

hammer:

shizhediao:

luiscape:

Created 2 years ago

Updated 16 hours ago

EAGLE by SafeAILab

Speculative decoding research paper for faster LLM inference

shizhediao:

philschmid:

pgarbacki:

JustinLin610:

Created 2 years ago

Updated 1 month ago

Nanoflow by efeslab

LLM serving framework for high throughput

WoosukKwon:

Ying1123:

zhuohan123:

Created 1 year ago

Updated 3 weeks ago

truss by basetenlabs

Model deployment tool for productionizing AI/ML models

apsdehal:

ogabrielluiz:

hammer:

pankajroark:

Created 3 years ago

Updated 3 days ago

mscclpp by microsoft

GPU-driven communication stack for scalable AI applications

jrk:

WoosukKwon:

Created 3 years ago

Updated 1 day ago

ao by pytorch

PyTorch library for quantization and sparsity in training/inference

danielhanchen:

shimmyshimmer:

parano:

willccbb:

Created 2 years ago

Updated 1 day ago

simple-evals by openai

Lightweight library for evaluating language models

winglian:

hammer:

patrickvonplaten:

simonw:

Created 2 years ago

Updated 8 months ago

sglang by sgl-project

Fast serving framework for LLMs and vision language models

amin3141:

shimmyshimmer:

beyang:

samlambert:

Created 2 years ago

Updated 16 hours ago

QuaRot by spcl

Code for a NeurIPS 2024 research paper on LLM quantization

jph00:

Created 2 years ago

Updated 1 year ago

FlagGems by flagos-ai

Operator library for LLM training/inference, implemented in Triton

Created 2 years ago

Updated 17 hours ago

ring-attention-pytorch by lucidrains

Pytorch impl of Ring Attention for near-infinite context

sxyu:

vincentweisser:

Created 2 years ago

Updated 11 months ago

Mooncake by kvcache-ai

Research paper on a disaggregated architecture for LLM serving

jiamings:

luiscape:

merrymercy:

WoosukKwon:

Created 1 year ago

Updated 17 hours ago

DeepSeek-V2 by deepseek-ai

MoE language model for research/API use

gakonst:

jiamings:

Jiayi-Pan:

Created 2 years ago

Updated 1 year ago

llama3 by meta-llama

*Deprecated* minimal example for loading and running Llama 3 models

tobi:

mckaywrigley:

osanseviero:

simonw:

Created 2 years ago

Updated 1 year ago

mlx by ml-explore

Array framework for machine learning on Apple silicon

geohot:

karpathy:

ekzhang:

lysandrejik:

Created 2 years ago

Updated 16 hours ago

flashinfer by flashinfer-ai

Kernel library for LLM serving

joker-eph:

chiphuyen:

hammer:

JustinLin610:

Created 2 years ago

Updated 21 hours ago

cutlass by NVIDIA

CUDA C++ and Python DSLs for high-performance linear algebra

pgarbacki:

pankajroark:

tridao:

chiphuyen:

Created 8 years ago

Updated 4 days ago

flash-attention by Dao-AILab

Fast, memory-efficient attention implementation

karpathy:

Jiayi-Pan:

zhiyuan8:

alexchen4ai:

Created 3 years ago

Updated 1 day ago

Medusa by FasterDecoding

Framework for accelerating LLM generation using multiple decoding heads

osanseviero:

parano:

luiscape:

chiphuyen:

Created 2 years ago

Updated 1 year ago

lmdeploy by InternLM

Toolkit for LLM compression, deployment, and serving

shimmyshimmer:

wsxiaoys:

luiscape:

jn2clark:

Created 2 years ago

Updated 20 hours ago

Feedback? Help us improve.