作者celestialgod (天)
看板R_Language
标题[分享] MxNet CSVIter
时间Sun Apr 16 00:42:31 2017
[关键字]: MxNet, deep learning, large dataset, mx.io.CSVIter
[出处]:
http://chingchuan-chen.github.io/posts/2017/04/16/mxnet-io-csviter
[重点摘要]:
本篇的重点在於下面这段R，用MxNet提供的mx.io.CSVIter去batch的训练Net模型
而这里的`train-64x64-data.csv`，每一行都是经过resized的30`张图片
所以`data.shape`是`64 x 64 x 30`
而label则每一行是长度600的binary vector，其`shape`设定成`600`
然後给好`batch.size`，MxNet就可以批次的从csv抓资料出来train模型了
不用一股脑地把资料全部汇入到R/gpu里面跑，不然再多的记忆体也用不完Orz
``` R
data_train <- mx.io.CSVIter(
data.csv = "train-64x64-data.csv", data.shape = c(64, 64, 30),
label.csv = "train-systole.csv", label.shape = 600,
batch.size = batch_size
)
```
--
R资料整理套件系列文：
magrittr #1LhSWhpH (R_Language) https://goo.gl/72l1m9
data.table #1LhW7Tvj (R_Language) https://goo.gl/PZa6Ue
dplyr(上.下) #1LhpJCfB,#1Lhw8b-s (R_Language) https://goo.gl/I5xX9b
tidyr #1Liqls1R (R_Language) https://goo.gl/i7yzAz
pipeR #1NXESRm5 (R_Language) https://goo.gl/zRUISx
--
※ 发信站: 批踢踢实业坊(ptt.cc), 来自: 111.246.26.70
※ 文章网址: https://webptt.com/cn.aspx?n=bbs/R_Language/M.1492274557.A.348.html
※ 编辑: celestialgod (111.246.26.70), 04/16/2017 00:45:38