This repository has been archived by the owner on Oct 30, 2019. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 666
running time of ResNets #189
Comments
This is the running time of a forward pass for each of these models on a CPU and a TITAN-X GPU
|
@aabobakr I have similar running time on ResNet-50/101/152. However, I run VGG16 using caffe(with cudnn 5.1, TITAN X GPU maxwell), it only costs me 12ms. Best! |
Hi,does anyone test the speed of resnet-18? Thanks! |
This is the running time of a forward pass of different models a Core i-7 CPU and a TITAN-X GPU: the time is average over 1000 iterations.
|
Thanks! |
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
In resnet's paper, it claims that Resnet50/101/152 have lower FLOPS compared to VGG16/19. However, I get much longer forward running time in Resnets than VGG-nets, e.g., Resnet 152 is 4~5 times longer than VGG16). Does anyone encounter similar problem?
The text was updated successfully, but these errors were encountered: