240 发简信
IP属地:河北
  • PyTorch中"诡异"的动量SGD实现

    pytorch中SGD的动量实现是如下这个样子的 将pytorch的实现翻译为公式,就是如下这个样子:为什么说它诡异呢?因为它与Polyak, ...

  • Resize,w 360,h 240
    如何用朴素贝叶斯对多维高斯分布进行分类

    There is a classification problem having two classes, with equal prior pr...

  • Resize,w 360,h 240
    从0实现高斯混合模型(EM-GMM)

    Problem: Please build a Gaussian mixture model (GMM) to model the data i...

  • 如何在单机上搭建分布式深度学习环境?以Pytorch+cuda+mpi on LXC为例

    环境:Manjaro - arch linux框架:pytorch 1.0-cuda-10.0Nividia driver:410cuda:10...

  • 白话梯度下降

    Homework2 1. Generate n = 2,000 points uniformly at random in the two-di...

  • Resize,w 360,h 240
    EASGD-简单介绍

    EASGDfrom:http://joerihermans.com/ramblings/distributed-deep-learning-pa...

  • 白话多项式拟合

    数学推导 过程参考于: 李航《统计学习方法》多项式函数拟合问题V2 https://blog.csdn.net/xiaolewennofollo...

  • Resize,w 360,h 240
    Talk - Scaling Distributed Machine Learning with System and Algorithm Co-design 笔记

    来源于李沐大神的talk:基于系统和算法的协同设计的大规模分布式机器学习(Scaling Distributed Machine Learnin...