# Why the reshape in Flux mnist convolution example

**URL:** <https://discourse.julialang.org/t/why-the-reshape-in-flux-mnist-convolution-example/13973>\
**Category:** Machine Learning\
**Tags:** question\
**Created:** [August 24, 2018, 1:27am UTC](https://discourse.julialang.org/t/why-the-reshape-in-flux-mnist-convolution-example/13973 "2018-08-24T01:27:38Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![hskramer](https://avatars.discourse-cdn.com/v4/letter/h/58f4c7/32.png) [@hskramer](https://discourse.julialang.org/u/hskramer)\
**Post date:** [August 24, 2018, 1:27am UTC](https://discourse.julialang.org/t/why-the-reshape-in-flux-mnist-convolution-example/13973/1 "2018-08-24T01:27:38Z")

</div>

I have learned to use most of the ML packages and I can compute  
the outcome of convolutions given kernel, stride pad…, but I don’t understand why or how to do a reshape when using Flux with convolutions and probably other layers so I need to understand this.  
Please help.  
Thank you  
H.Kramer

---

<div class="post-metadata">

**Author:** ![hskramer](https://avatars.discourse-cdn.com/v4/letter/h/58f4c7/32.png) [@hskramer](https://discourse.julialang.org/u/hskramer)\
**Post date:** [August 25, 2018, 12:57am UTC](https://discourse.julialang.org/t/why-the-reshape-in-flux-mnist-convolution-example/13973/2 "2018-08-25T00:57:02Z")

</div>

Thanks working on this problem myself help me gain a lot insight into  
how flux and julia work. The reshape is a Flatten layer taking the 6x6x8xN and turning it into a 288xN for the Dense layer or what many call a fully connected layer.