Skip to content
View vvvdwbvvv's full-sized avatar
💭
I like Machine Learning
💭
I like Machine Learning
14 followers · 45 following
  • National Chengchi University
  • Taipei, Taiwan
  • 19:19 (UTC +08:00)

Achievements

Achievement: Pull Sharkx2Achievement: Pair ExtraordinaireAchievement: YOLOAchievement: Quickdraw

Achievements

Achievement: Pull Sharkx2Achievement: Pair ExtraordinaireAchievement: YOLOAchievement: Quickdraw

Highlights

  • Pro

Block or report vvvdwbvvv

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
vvvdwbvvv/README.md

Hi 👋

I build and study generative models, computer vision, and high-performance ML systems.

  • 🔬 Exploring diffusion models, flow matching, inverse problems, and image restoration
  • ⚡ Learning how modern models run efficiently through CUDA kernels and ML systems
  • 🧠 Reproducing papers and systems to understand them from first principles
  • 🛠️ Also interested in practical tools across Python, Go, and the web
  • 📫 The best way to reach me is through GitHub

Tech I work with

Languages

Python C++ Go JavaScript   ML & Data

PyTorch NumPy Pandas OpenCV   Systems & Tools

CUDA Docker Git Linux Neovim

GitHub contribution snake

Selected work

Project Focus What it explores
BlurDM Diffusion · Image restoration A blur diffusion model for image deblurring, published at NeurIPS 2025
FlowDPS Flow matching · Inverse problems Flow-driven posterior sampling for inverse problems, published at ICCV 2025
ID-Blau Diffusion · Deblurring Implicit diffusion-based reblurring augmentation, published at CVPR 2024
DeepSeekV3 MoE CUDA CUDA · ML systems CUDA work targeting DeepSeek-V3 mixture-of-experts workloads
forge-cute-py CuTe DSL · GPU kernels A workflow for implementing, validating, benchmarking, and profiling GPU kernels
AudioPilot LLMs · Computer vision An automatic flight-control system built around language and vision models

Pinned Loading

  1. linkedin/Liger-Kernel linkedin/Liger-Kernel Public

    Efficient Triton Kernels for LLM Training

    Python 6.5k 567

  2. pytorch pytorch Public

    Forked from pytorch/pytorch

    Tensors and Dynamic neural networks in Python with strong GPU acceleration

    Python

  3. flow_matching flow_matching Public

    Forked from facebookresearch/flow_matching

    A PyTorch library for implementing flow matching algorithms, featuring continuous and discrete flow matching implementations. It includes practical examples for both text and image modalities.

    Python

  4. stochastic-interpolants stochastic-interpolants Public

    Forked from malbergo/stochastic-interpolants

    Jupyter Notebook