MACKO: Sparse matrix vector multiplication for low sparsity

PaCoRe: Learning to Scale Test-Time Compute with Parallel Coordinated Reasoning

Tests for PyTorch api.

Use pytorch profile api to further analysis the training detailed information, like heaps and stacks, time consuming.

Official implementation of MATPO: Multi-Agent Tool-Integrated Policy Optimization.

SGLang is a fast serving framework for large language models and vision language models.

A Quirky Assortment of CuTe Kernels

Fast and memory-efficient exact attention

校招、秋招、春招、实习好项目！带你从零实现一个高性能的深度学习推理库，支持大模型 llama2 、Unet、Yolov5、Resnet等模型的推理。Implement a high-performance deep learning inference library step by step

Dexbotic: Open-Source Vision-Language-Action Toolbox

An MCP-based chatbot | 一个基于MCP的聊天机器人

MMR1: Enhancing Multimodal Reasoning with Variance-Aware Sampling and Open Resources

Official implementation of MATPO: Multi-Agent Tool-Integrated Policy Optimization.

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

Long-Tailed Visual Recognition via Self-Heterogeneous Integration with Knowledge Excavation (CVPR 2023)

MiroRL is an MCP-first reinforcement learning framework for deep research agent.

MiroThinker is a series of open-source agentic models trained for deep research and complex tool use scenarios.

MiroMind-M1 is a fully open-source series of reasoning language models built on Qwen-2.5, focused on advancing mathematical reasoning.

End to end LLM agent RL training with verl

System 2 Reasoning Link Collection

Minimal reproduction of DeepSeek R1-Zero

Official implementation of SFPK-pruner in ICLR 2025 "Probabilistic Neural Pruning via Sparsity Evolutionary Fokker-Planck-Kolmogorov Equation"

Official implementation of ICLR 2025 'LORO: Parameter and Memory Efficient Pretraining via Low-rank Riemannian Optimization'

Accessible large language models via k-bit quantization for PyTorch.

介绍docker、docker compose的使用。

A curated list of neural network pruning resources.