Analyze computation-communication overlap in V3/R1.

DeepGEMM: clean and efficient FP8 GEMM kernels with fine-grained scaling

flash attention tutorial written in python, triton, cuda, cutlass

FlashMLA: Efficient MLA decoding kernels

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

Production-tested AI infrastructure tools for efficient AGI development and community-driven innovation

Mooncake is the serving platform for Kimi, a leading LLM service provided by Moonshot AI.

A Flexible Framework for Experiencing Cutting-edge LLM Inference Optimizations

An autoregressive character-level language model for making more things

Clean, minimal, accessible reproduction of DeepSeek R1-Zero

Repository of Jupyter notebook tutorials for teaching the Deep Learning Course at the University of Amsterdam (MSc AI), Fall 2023

🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.

🌸 Run LLMs at home, BitTorrent-style. Fine-tuning and inference up to 10x faster than offloading

LangGPT: Empowering everyone to become a prompt expert!🚀 Structured Prompt，Language of GPT, 结构化提示词，结构化Prompt

LLM&VLM Tutorial

Auto-Editor: Efficient media analysis and rendering

C, C++ and Python Code for Exercises and Solutions

PPL Quantization Tool (PPQ) is a powerful offline neural network quantization tool.

Neural Networks: Zero to Hero

Tile primitives for speedy kernels

Tool for visual profiling Core ML models, compatible with both package and compiled versions, including reasons for unsupported operations on the Neural Engine

High-efficiency floating-point neural network inference operators for mobile, server, and Web

On-device AI across mobile, embedded and edge for PyTorch

本项目是一个通过文字生成图片的项目，基于开源模型Stable Diffusion V1.5生成可以在手机的CPU和NPU上运行的模型，包括其配套的模型运行框架。

PainterEngine is a application/game engine with software renderer,PainterEngine can be transplanted to any platform that supports C

Code release for book "Efficient Training in PyTorch"

This repository contains the official implementation of the research paper, "MobileCLIP: Fast Image-Text Models through Multi-Modal Reinforced Training" CVPR 2024

Object Detection toolkit based on PaddlePaddle. It supports object detection, instance segmentation, multiple object tracking and real-time multi-person keypoint detection.

📚200+ Tensor/CUDA Cores Kernels, ⚡️flash-attn-mma, ⚡️hgemm with WMMA, MMA and CuTe (98%~100% TFLOPS of cuBLAS/FA2 🎉🎉).