dear author， can you tell me what is your device ,what is you used time with the default config;

[reaIws]

作者您好，我想纵向比较下显卡性能，您能告诉我，在默认配置下，您得训练所用时间是多少么，我现在正在用tesla显卡训练，如果训练结束，我会告诉您，我得用时：》

[naturomics]

                      The following table is the approximate training time

num x GPU device	other configurations	speed(min/epoch)
1 x 1080ti	default	3
1 x K80	default	11

I also ran on a 750ti GPU, but that cards is gone and the record is lost, remember it's 8-10 min/epoch

[reaIws]

谢谢，我是两个显卡不知道会用到几个，但是速度是 6.5min/epoch :>

[naturomics]

You can check it with the command nvidia-smi to see how many GPU the program is using under Linux. For windows user, find it out in Task Manager(任务管理器). But by default it should use one GPU.

[naturomics]

BTW, which tesla GPU and CPU u'r using?

[reaIws]

ma computer has two tesla k80
cpu is E5-2630v4
it computer capability is 3.7
1080Ti's capability is 6.1
then:
when i run pyhon main.py
i get:
createing tensorflow device {/device:GPU:0}
createing tensorflow device {/device:GPU:1}

when i run nvidia-smi
it show:

GPU	PID	Type	GPU memory usage	volatile GPU-Util	Temp
0	4853	C	9367MiB	83%	59C
1	4853	C	62MiB	0%	35C

[jackalwoods]

So this code will only work on GPU and not on CPU?

[naturomics]

Both should work. Training for 30min on a 1080ti is enough to have a good result, but on CPU, it might take for hours. Prashant Desai <[email protected]> 于2019年2月27日周三 下午9:06写道：

…

So this code will only work on GPU and not on CPU? — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#53 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ASh0sSYxX9EnJx2VBu7gFTrDfQcuKIZ_ks5vRoLEgaJpZM4RaOTC> .

[jackalwoods]

Thanks for reply. Not sure why but I am getting below error on my CPU. tensorflow.python.framework.errors_impl.NotFoundError: C:\Users\prashant.desai\AppData\Local\Continuum\Anaconda3\lib\site-packages\tensorflow\contrib\rnn\python\ops\_gru_ops.so not found Regards, Prashant

…

On Thu, Feb 28, 2019 at 11:51 AM Huadong Liao ***@***.***> wrote: Both should work. Training for 30min on a 1080ti is enough to have a good result, but on CPU, it might take for hours. Prashant Desai ***@***.***> 于2019年2月27日周三 下午9:06写道： > So this code will only work on GPU and not on CPU? > > — > You are receiving this because you commented. > Reply to this email directly, view it on GitHub > < #53 (comment) >, > or mute the thread > < https://github.com/notifications/unsubscribe-auth/ASh0sSYxX9EnJx2VBu7gFTrDfQcuKIZ_ks5vRoLEgaJpZM4RaOTC > > . > — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#53 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AEJXwDp9SOrM7C917WdaE71BYHD60bnHks5vR3V1gaJpZM4RaOTC> .

[naturomics]

@jackalwoods Looks like it's your tf env problem. please make sure your tf is installed correctly or a right version (the code was developed with tf 1.3, a newer one but older than 2.0 should work).

[jackalwoods]

Thanks, yes it was the problem with tf indeed. Got solved. I am trying to modify this for image similarity problem. Which layer should i pull before final layer to match with other image? appreciate any guidance on this.

…

On Sun, Mar 10, 2019 at 3:06 PM Huadong Liao ***@***.***> wrote: @jackalwoods <https://github.com/jackalwoods> Looks like it's your tf env problem. please make sure your tf is installed correctly or a right version (the code was developed with tf 1.3, a newer one but older than 2.0 should work). — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#53 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AEJXwBD49WDlssELRrbz1Pk85fQ4Gqx8ks5vVNIwgaJpZM4RaOTC> .

[jackalwoods]

Hello there, I just signed the petition "Ministry of Labour and Employment India: Stop Hiring in 15 Days and 90 Days Notice Period - Labour Law Attention Required" and wanted to see if you could help by adding your name. Our goal is to reach 25,000 signatures and we need more support. You can read more and sign the petition here: http://chng.it/zBNrW4fjT6 Thanks! Prashant