Xiuying Wei

NextLat. Official codebase for "Next-Latent Prediction Transformers Learn Compact World Models"

beyond-next-token-prediction. Curated collection of research on the limitations of next-token prediction and methods that go beyond it.

FLARE.py. Fast Low-rank Attention Routing Engine

rat-plus. RAT+: Train Dense, Infer Sparse - Recurrence Augmented Attention for Dilated Inference (ICML2026)

2027-AI-College-Jobs. 2027 AI/ML internship & new graduate job list updated daily

pgm. [ICLR 2026 Oral] Partition Generative Modeling (PGM): Masked Modeling Without Masks

OLMo. Modeling, training, eval, and inference code for OLMo

quantile-reward-policy-optimization. Official codebase for "Quantile Reward Policy Optimization: Alignment with Pointwise Regression and Exact Partition Functions" (Matrenok et al. 2025).

RAT. Official code for the NeurIPS25 paper "RAT: Bridging RNN Efficiencyand Attention Accuracy in Language Modeling" (https://arxiv.org/abs/2507.04416))

torchtitan. A PyTorch native platform for training generative AI models

mad-lab. A MAD laboratory to improve AI architecture designs 🧪

xlstm. Official repository of the xLSTM.

awesome-o1. A bibliography and survey of the papers surrounding o1

accelerated-scan. Accelerated First Order Parallel Associative Scan

flash-linear-attention. 🚀 Efficient implementations for emerging model architectures

lectures. Material for gpu-mode lectures

StructuredFFN. The official code of "Building on Efficient Foundations: Effectively Training LLMs with Structured Feedforward Layers"

nanotron. Minimalistic large language model 3D-parallelism training

Pai-Megatron-Patch. The official repo of Pai-Megatron-Patch for LLM & VLM large scale training developed by Alibaba Cloud.

nanoGPT. The simplest, fastest repository for training/finetuning medium-sized GPTs.

LeetCUDA. 📚LeetCUDA: Modern CUDA Learn Notes with PyTorch for Beginners🐑, 200+ CUDA Kernels, Tensor Cores, HGEMM, FA-2 MMA.🎉

CUDA_Freshman. Cuda

cutlass. CUDA Templates and Python DSLs for High-Performance Linear Algebra

python-ml-research-template. A template for starting reproducible Python machine-learning projects with hardware acceleration. Find an example at https://github.com/CLAIRE-Labo/no-representation-no-trust

cuda-samples. Samples for CUDA Developers which demonstrates features in CUDA Toolkit

RepDistiller. [ICLR 2020] Contrastive Representation Distillation (CRD), and benchmark of recent knowledge distillation methods

ZOOpt. A python package of Zeroth-Order Optimization (ZOOpt)

conv_arithmetic. A technical report on convolution arithmetic in the context of deep learning

bolei_awesome_posters. CVPR and NeurIPS poster examples and templates

FasterTransformer. Transformer related optimization, including BERT, GPT

effective_transformer. Running BERT without Padding

MT-Reading-List. A machine translation reading list maintained by Tsinghua Natural Language Processing Group

Awesome-Knowledge-Distillation. Awesome Knowledge-Distillation. 分类整理的知识蒸馏paper(2014-2021)。

OpenPrompt. An Open-Source Framework for Prompt-Learning.

transformer-quantization. Python

PyHessian. PyHessian is a Pytorch library for second-order based analysis and training of Neural Networks

nlp-architect. A model library for exploring state-of-the-art deep learning topologies and techniques for optimizing Natural Language Processing neural networks

aimet. AIMET is a library that provides advanced quantization and compression techniques for trained neural network models.

pytorch-cifar100. Practice on cifar100(ResNet, DenseNet, VGG, GoogleNet, InceptionV3, InceptionV4, Inception-ResNetv2, Xception, Resnet In Resnet, ResNext,ShuffleNet, ShuffleNetv2, MobileNet, MobileNetv2, SqueezeNet, NasNet, Residual Attention Network, SENet, WideResNet)

AMP-Regularizer. Code for our paper "Regularizing Neural Networks via Adversarial Model Perturbation", CVPR2021

sam.pytorch. A PyTorch implementation of Sharpness-Aware Minimization for Efficiently Improving Generalization

TensorRT. NVIDIA® TensorRT™ is an SDK for high-performance deep learning inference on NVIDIA GPUs. This repository contains the open source components of TensorRT.