Researcher at @Nvidia on efficient large models. Previously @MIT, @SFU, @ZJU.
pytorch-OpCounter. Count the MACs / FLOPs of your PyTorch model.
5.1kEfficient-PyTorch. My best practice of training large dataset using PyTorch.
1.1kpytorch-memonger. Sublinear memory optimization for deep learning. https://arxiv.org/abs/1604.06174
603SparseNet. [ECCV 2018] Sparsely Aggreagated Convolutional Networks https://arxiv.org/abs/1801.05895
124arXiv-stats. Python
67hf-torrent. Python
44mxbox. Simple, efficient and flexible vision toolbox for mxnet framework.
32Bayesian-Compression-for-Deep-Learning. Remplementation of paper https://arxiv.org/abs/1705.08665
29PyTorch-Template. A template for PyTorch projects.
23Colorize-Your-World. Let there be color!
22Machine-Learning-for-Image-Colorization. (Torch + Tensorflow) A deep magic brings color to your monochrome image!
13GroupNorm.pytorch. PyTorch implementation of Group Normalization https://arxiv.org/abs/1803.08494
13Colorizing-Color-Images. [HVEI 2018] Colorizing Color Images
12Project-Page-Render. HTML
11Echoo. Let your program echo to you.
10arch-viz. Python
6hf-torrent-store.
6MNasNet-TensorFlow. Implementation of MnasNet: Platform-Aware Neural Architecture Search for Mobile
5PyTorch-via-PyTorch. C++
5Deep-Learning-Live. From linear regression to multi-layer perceptron, an introductive tutorial for deep learning beginners.
5tvm-notes. Python
5ffmpeg-cuda-docker. A docker container to launch GPU accelerated FFmpeg
4edge-cloud-train. Python
4HW-for-COMP. HTML
4TGFont. Simple hook to change Telegram desktop's font in Windows.
4FlashATM.
4pi-tools. A repo includes some useful tools for raspberry pi farm setup. https://hub.docker.com/repository/docker/lyken/pi-tools
3Colorize.PyTorch. Python
3tvm-hack. Python
3EIE-pytorch. PyTorch implementation for EIE https://arxiv.org/abs/1602.01528
3awesome-multimodal-reasoning.
3dlg. [NeurIPS 2019] Deep Leakage From Gradients
2wandb-example. Python
2sample-video.
2pythonLearn. Python
2Surge-Proxy-Setting.
2ubuntun-research. Common scripts I've used for setting up my ubuntu server
2R1-V. Witness the aha moment of VLM with less than $3.
1ink-epoch. CSS
1the-importance-of-color.
1T2C. Expand TOML configuration to commands
1k8s-mpi.
1Boostup-ImageNet. Python
1Test-Repo2.
1