thu-ml

An elegant PyTorch deep reinforcement learning library.

[ICLR2025, ICML2025, NeurIPS2025 Spotlight] Quantized Attention achieves speedup of 2-5x compared to FlashAttention, without losing end-to-end metrics across language, image, and video models.

TurboDiffusion: 100–200× Acceleration for Video Diffusion Models

A probabilistic programming library for Bayesian deep learning, generative models, based on Tensorflow

RDT-1B: a Diffusion Foundation Model for Bimanual Manipulation

ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with Variational Score Distillation (NeurIPS 2023 Spotlight)

Code and models for the paper "One Transformer Fits All Distributions in Multi-Modal Diffusion"

Official code of Motus: A Unified Latent Action World Model

[ICML2025] SpargeAttention: A training-free sparse attention that accelerates any model inference.

[ICML 2026] Official codebase for "Causal Forcing: Autoregressive Diffusion Distillation Done Right for High-Quality Real-Time Interactive Video Generation" & Causal Forcing++

Official implementation for "RIFLEx: A Free Lunch for Length Extrapolation in Video Diffusion Transformers" (ICML 2025) , UltraViCo (ICLR 2026) and UltraImage

Official code of RDT 2

[ECCV 2024] Single Image to 3D Textured Mesh in 10 seconds with Convolutional Reconstruction Model.

SLA: Beyond Sparsity in Diffusion Transformers via Fine-Tunable Sparse–Linear Attention

A Python library for adversarial machine learning focusing on benchmarking adversarial robustness.

Official implementation for "Identifying and Solving Conditional Image Leakage in Image-to-Video Diffusion Model" (NeurIPS 2024)

Official implementation for "ControlVideo: Adding Conditional Control for One Shot Text-to-Video Editing"

A toolbox for benchmarking trustworthiness of multimodal large language models (MultiTrust, NeurIPS 2024 Track Datasets and Benchmarks)

Benchmarking Robustness of 3D Object Detection to Common Corruptions in Autonomous Driving, CVPR 2023

Cache efficient implementation for Latent Dirichlet Allocation

Low-bit optimizers for PyTorch

Official code for "DPM-Solver-v3: Improved Diffusion ODE Solver with Empirical Model Statistics" (NeurIPS 2023)

Official codebase for "Schrodinger Bridges Beat Diffusion Models on Text-to-Speech Synthesis" (https://arxiv.org/abs/2312.03491).

Official codebase for "Diffusion Bridge Implicit Models" (ICLR 2025) and "Consistency Diffusion Bridge Models" (NeurIPS 2024)

[ICLR2025] Codebase for "ReMoE: Fully Differentiable Mixture-of-Experts with ReLU Routing", built on Megatron-LM.

天授中文文档

Hierarchical Decomposition of Prompt-Based Continual Learning: Rethinking Obscured Sub-optimality (NeurIPS 2023, Spotlight)

Official codebase for "STAIR: Improving Safety Alignment with Introspective Reasoning"

Stochastic training of graph convolutional networks

Versatile Cardiovascular Signal Generation with a Unified Diffusion Transformer

A toolbox for benchmarking Multimodal LLM Agents trustworthiness across truthfulness, controllability, safety and privacy dimensions through 34 interactive tasks

Code accompanying the paper "Noise Contrastive Alignment of Language Models with Explicit Rewards" (NeurIPS 2024)