About Me Learning in LLMs and MLsys, recently focused on RL training <!-- My primary work and research focus on RL: Design better architectures and algorithms to conduct efficient/effective RL…
About Me
Learning in LLMs and MLsys, recently focused on RL training
<!--
My primary work and research focus on
- RL: Design better architectures and algorithms to conduct efficient/effective RL training
- Framework: Support pre-/post-training workflows, scale up training, and improve performance
- GPU: Optimize GPU memory utilization
<!--
Working on open-source Projects:
- RL or Framework: slime-ROCm version, verl-ROCm version, SGLang
- GPU: torch_memory_saver
- (Public archive): AgentLaboratory, AgentVerse, ChatDev
<!-- Framework: Prompt-Transferability -->
Maintain and work on:
Contact:
- 💬 Personal Website: https://yushengsu-thu.github.io/
- Google Scholar: https://scholar.google.com/citations?user=xwy6Va4AAAAJ
- 📫 E-mail: [email protected]
<!--
Github Stats:
-->
<!--
-->
<!--
-->
<!--
-->
<!--
-->
<!--| | |
|---------|-------|-->
<!--
yushengsu-thu/yushengsu-thu is a ✨ _special_ ✨ repository because its README.md (this file) appears on your GitHub profile.
Here are some ideas to get you started:
- 🔭 I’m currently working on ...
- 🌱 I’m currently learning ...
- 👯 I’m looking to collaborate on ...
- 🤔 I’m looking for help with ...
- 💬 Ask me about ...
- 📫 How to reach me: ...
- 😄 Pronouns: ...
- ⚡ Fun fact: ...
-
sglang-miles-hand-on
No description.
Jupyter Notebook ★ 4 1mo agoExplain → -
torch_memory_saver ⑂
Allow torch tensor memory to be released and resumed later
Python ★ 2 6mo agoExplain → -
PET_Scaling
Exploring the Impact of Model Scaling on Parameter-efficient Tuning Methods
Python ★ 2 1y agoExplain → -
miles ⑂
No description.
Python ★ 1 17h agoExplain → -
lora_perf_lora_profile
LoRA vs no-LoRA perf benchmarks + torch profiles (GB300 dev runs)
Python ★ 1 4d agoExplain → -
tune-lora-perf
No description.
Shell ★ 1 6d agoExplain → -
sglang ⑂
SGLang is a fast serving framework for large language models and vision language models.
Python ★ 1 2d agoExplain → -
lora-dev-script
No description.
Shell ★ 1 27d agoExplain → -
yushengsu-thu.github.io ⑂
Github Pages template for academic personal websites, forked from mmistakes/minimal-mistakes
JavaScript ★ 1 1mo agoExplain → -
miles-dev-script
No description.
Python ★ 1 2mo agoExplain → -
claude-vim-IDE
No description.
Lua ★ 1 2mo agoExplain → -
hack-vimrc
my vim configure
Vim Script ★ 1 2mo agoExplain → -
Megatron-Bridge ⑂
HuggingFace conversion and training library for Megatron-based models
Python ★ 1 2mo agoExplain → -
yushengsu-thu
No description.
★ 1 4mo agoExplain → -
verl ⑂
verl: Volcano Engine Reinforcement Learning for LLMs
Python ★ 1 11mo agoExplain → -
Pai-Megatron-Patch-amd_version ⑂
The official repo of Pai-Megatron-Patch for LLM & VLM large scale training developed by Alibaba Cloud.
Python ★ 1 11mo agoExplain → -
Megatron-LM-amd_version ⑂
Ongoing research training transformer models at scale
Python ★ 1 11mo agoExplain → -
Awesome-ML-SYS-Tutorial ⑂
My learning notes/codes for ML SYS.
Python ★ 1 1y agoExplain → -
FacebookChatBot
No description.
JavaScript ★ 1 3y agoExplain → -
miles_dev
No description.
★ 0 18h agoExplain → -
merge_ci
No description.
Shell ★ 0 25d agoExplain → -
LeetCUDA ⑂
📚LeetCUDA: Modern CUDA Learn Notes with PyTorch for Beginners🐑, 200+ CUDA Kernels, Tensor Cores, HGEMM, FA-2 MMA.🎉
★ 0 4mo agoExplain → -
lm-evaluation-harness ⑂
A framework for few-shot evaluation of language models.
★ 0 4mo agoExplain → -
tilelang ⑂
Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels
★ 0 4mo agoExplain → -
slime ⑂
slime is an LLM post-training framework for RL Scaling.
Python ★ 0 5mo agoExplain → -
nano-sglang ⑂
No description.
★ 0 8mo agoExplain → -
mbridge ⑂
Bridge Megatron-Core to Hugging Face/Reinforcement Learning
Python ★ 0 8mo agoExplain → -
batch_invariant_ops ⑂
No description.
★ 0 9mo agoExplain → -
llm.c ⑂
LLM training in simple, raw C/CUDA
Cuda ★ 0 1y agoExplain → -
Megatron-LLM ⑂
distributed trainer for LLMs
Python ★ 0 1y agoExplain → -
verl_training_log
verl_training_log
★ 0 1y agoExplain → -
datasciencecoursera ⑂
for Data Science class on Coursera
★ 0 6y agoExplain → -
LunarVim ⑂
An IDE layer for Neovim with sane defaults. Completely free and community driven.
★ 0 3y agoExplain → -
lua-nvim-config ⑂
No description.
★ 0 3y agoExplain → -
ProKil ⑂
No description.
★ 0 3y agoExplain → -
nvimdots ⑂
A well configured and structured Neovim.
★ 0 3y agoExplain → -
ModelCenter ⑂
Efficient, Low-Resource, Distributed transformer implementation based on BMTrain
Python ★ 0 4y agoExplain → -
PromptPapers ⑂
Must-read papers on prompt-based tuning for pre-trained language models.
★ 0 4y agoExplain → -
House-Renting ⑂
No description.
HTML ★ 0 8y agoExplain → -
nccucs-se2016-dapp-example ⑂
105學年度 國立政治大學 資訊科學系 軟體工程 DApp 範例
JavaScript ★ 0 9y agoExplain → -
Click_Through_Rate_Prediction
Predict (User pattern): Predict every user's Commercial Click-through-rate with random forest model and kdd-naggle cvs data
Jupyter Notebook ★ 0 9y agoExplain → -
OOP_assignment4_RPG_Game
RPG_Game
C++ ★ 0 8y agoExplain →
No repos match these filters.