Caffe需要数据洗牌吗？ [英] Does Caffe need data to be shuffled?

问题描述

我使用C ++将我的图像数据转换为caffe db格式（leveldb，lmdb）作为示例我使用此代码 imagenet 。

I convert my image data to caffe db format (leveldb, lmdb) using C++ as example I use this code for imagenet.

数据需要改写，我可以写db所有的正数，像00000000111111111，或者数据需要改组，标签应该看起来像010101010110101011010？

Is data need to be shuffled, can I write to db all my positives and then all my negatives like 00000000111111111, or data need to be shuffled and labels should look like 010101010110101011010?

如何caffe从DB数据，它是真的，它使用所有数据的随机子集size = batch_size ？

How caffe sample data from DB, is it true that it use random subset of all data with size = batch_size?

推荐答案

？如果你不洗牌，想想学习过程; caffe只看到 0 samples - 你期望算法推导什么？简单地预测 0 所有的时间，一切都很酷。如果你在第一个 1 之前有足够的 0 caffe将非常有信心预测总是 0 。

另一方面，如果它经常看到 0 和<$ c的混合$ c> 1 它从头开始学习了有意义的功能，用于分隔示例。

底线：非常有利于洗牌训练样本，特别是当使用基于SGD的方法时。

Should you shuffle the samples? Think about the learning process if you don't shuffle; caffe sees only 0 samples - what do you expect the algorithm to deduce? simply predict 0 all the time and everything is cool. If you have plenty of 0 before you hit the first 1 caffe will be very confident in predicting always 0. It will be very difficult to move the model from this point.
On the other hand, if it constantly sees a mix of 0 and 1 it learns from the beginning meaningful features for separating the examples.
Bottom line: it is very advantageous to shuffle the training samples, especially when using SGD-based approaches.

AFAIK，caffe不会随机抽样 batch_size 样本，而是在输入数据库<$ c $

AFAIK, caffe does not randomly sample batch_size samples, but rather goes sequentially over the input DB batch_size after batch_size samples.

TL; DR strong>

shuffle。

TL;DR
shuffle.

这篇关于Caffe需要数据洗牌吗？的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！