IP屬地：貴州
-
PyTorch中"詭異"的動(dòng)量SGD實(shí)現(xiàn)
pytorch中SGD的動(dòng)量實(shí)現(xiàn)是如下這個(gè)樣子的 將pytorch的實(shí)現(xiàn)翻譯為公式，就是如下這個(gè)樣子：為什么說它詭異呢？因?yàn)樗cPolyak, ...
-
如何用樸素貝葉斯對多維高斯分布進(jìn)行分類There is a classi?cation problem having two classes, with equal prior pr...
-
從0實(shí)現(xiàn)高斯混合模型（EM-GMM）Problem: Please build a Gaussian mixture model (GMM) to model the data i...
-
如何在單機(jī)上搭建分布式深度學(xué)習(xí)環(huán)境？以Pytorch+cuda+mpi on LXC為例
環(huán)境：Manjaro - arch linux框架：pytorch 1.0-cuda-10.0Nividia driver：410cuda：10...
-
白話梯度下降
Homework2 1. Generate n = 2,000 points uniformly at random in the two-di...
-
EASGD-簡單介紹EASGDfrom:http://joerihermans.com/ramblings/distributed-deep-learning-pa...
-
白話多項(xiàng)式擬合
數(shù)學(xué)推導(dǎo) 過程參考于: 李航《統(tǒng)計(jì)學(xué)習(xí)方法》多項(xiàng)式函數(shù)擬合問題V2 https://blog.csdn.net/xiaolewennofollo...
-
Talk - Scaling Distributed Machine Learning with System and Algorithm Co-design 筆記來源于李沐大神的talk：基于系統(tǒng)和算法的協(xié)同設(shè)計(jì)的大規(guī)模分布式機(jī)器學(xué)習(xí)(Scaling Distributed Machine Learnin...