Owner · thu-ml

thu-ml

thu-ml's tracked open-source repos, sorted by stars.

10 repos

#1
tianshou ↗@thu-mlPython
An elegant PyTorch deep reinforcement learning library.
★ 10,955+10Star change over the last 7 days
#2
SageAttention ↗@thu-mlCuda
[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.
★ 3,699+14Star change over the last 7 days
#3
TurboDiffusion ↗@thu-mlPython
TurboDiffusion: 100–200× Acceleration for Video Diffusion Models
★ 3,634+6Star change over the last 7 days
#4
RoboticsDiffusionTransformer ↗@thu-mlPython
RDT-1B: a Diffusion Foundation Model for Bimanual Manipulation
★ 1,784+3Star change over the last 7 days
#5
Motus ↗@thu-mlPython
Official code of Motus: A Unified Latent Action World Model
★ 1,260+11Star change over the last 7 days
#6
SpargeAttn ↗@thu-mlCuda
[ICML2025] SpargeAttention: A training-free sparse attention that accelerates any model inference.
★ 1,043+1Star change over the last 7 days
#7
Causal-Forcing ↗@thu-mlPython
[ICML 2026] Official codebase for "Causal Forcing: Autoregressive Diffusion Distillation Done Right for High-Quality Real-Time Interactive Video Generation" & Causal Forcing++
★ 948+10Star change over the last 7 days
#8
DiT-Extrapolation ↗@thu-mlPython
Official implementation for "RIFLEx: A Free Lunch for Length Extrapolation in Video Diffusion Transformers" (ICML 2025) , UltraViCo (ICLR 2026) and UltraImage
★ 827+2Star change over the last 7 days
#9
RDT2 ↗@thu-mlPython
Official code of RDT 2
★ 807+0Star change over the last 7 days
#10
CRM ↗@thu-mlPython
[ECCV 2024] Single Image to 3D Textured Mesh in 10 seconds with Convolutional Reconstruction Model.
★ 696+0Star change over the last 7 days

← Back to owner ranking