puregit/cuda-convnet2

Element wise sum not working as expected

Opened this issue · 1 comments

What steps will reproduce the problem?
1. Add an element-wise sum layer to your config file.
2. Specify exactly same inputs to the 'inputs' parameter
3. Specify coeffs=1, -1

What is the expected output? What do you see instead?
With exactly same inputs and the coeffs specified as 1 and -1, it is expected 
that the output of the layer produce 0.0
Instead we see a non-zero output. Also changing the coeffs values does not seem 
to have any effect.

What version of the product are you using? On what operating system?
cuda-convnet2 on ubuntu-linux

Please provide any additional information below.


Original issue reported on code.google.com by aniket.v...@gmail.com on 29 Oct 2014 at 11:17

Hey,

As mentioned in one of the comment boards (cannot remember where) the problem 
is quite simple actually:
the line
coeffs= 1,-1 

should be written to the params.cgf file, not the layers.cfg file

So to solve the problem for all users, one should either correct the wiki 
(simpler) or add the coefficient read-in somewhere where the layer params are 
read in.

Best wishes

Ardi Tampuu
Comp.Neuro Team, University of Tartu

Original comment by arditam...@gmail.com on 18 Feb 2015 at 12:25

  • Added labels: ****
  • Removed labels: ****