Re: [theano-users] Re: Is PyCUDA needed?

2016-10-18 Thread Frédéric Bastien
Check Theano tutoriel

deeplearning.com/software/theano/tutorial

Le 18 oct. 2016 05:40, "senthil Nathan"  a
écrit :

> Dear Pascal , Thanks a lot for answers. I'm a new-bee and trying to learn
> on my own. Any tips on where to start after the installation.
> Regards
> Senthil
>
> On Tuesday, October 18, 2016 at 1:14:16 PM UTC+8, Pascal Lamblin wrote:
>>
>> On Mon, Oct 17, 2016, senthil Nathan wrote:
>> > btw Im also having messages like , CVM does not support memory profile
>> ,
>> > using stack VM.
>>
>> This is normal.
>>
>> > > Dear Members , if I use CPU to run theano , is the pycuda need to be
>> > > setup. Thanks
>>
>> No, PyCUDA is not needed at all. We have some integration for adding
>> code written in PyCUDA into Theano, and this is what the test is trying
>> to do, but you are not missing on any feature if you do not have PyCUDA.
>>
>> > > When i'm running the test im getting All-NaN slice encountered ,
>> All-NaN
>> > > axis  encounteredas runtime warnings as well. Is it normal ?
>>
>> Yes, these are warnings emitted by NumPy when we test certain things.
>>
>>
>> --
>> Pascal
>>
> --
>
> ---
> You received this message because you are subscribed to the Google Groups
> "theano-users" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to theano-users+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"theano-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to theano-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Re: [theano-users] Re: Is PyCUDA needed?

2016-10-18 Thread senthil Nathan
Dear Pascal , Thanks a lot for answers. I'm a new-bee and trying to learn 
on my own. Any tips on where to start after the installation. 
Regards
Senthil 

On Tuesday, October 18, 2016 at 1:14:16 PM UTC+8, Pascal Lamblin wrote:
>
> On Mon, Oct 17, 2016, senthil Nathan wrote: 
> > btw Im also having messages like , CVM does not support memory profile , 
> > using stack VM. 
>
> This is normal. 
>
> > > Dear Members , if I use CPU to run theano , is the pycuda need to be 
> > > setup. Thanks 
>
> No, PyCUDA is not needed at all. We have some integration for adding 
> code written in PyCUDA into Theano, and this is what the test is trying 
> to do, but you are not missing on any feature if you do not have PyCUDA. 
>
> > > When i'm running the test im getting All-NaN slice encountered , 
> All-NaN 
> > > axis  encounteredas runtime warnings as well. Is it normal ? 
>
> Yes, these are warnings emitted by NumPy when we test certain things. 
>
>
> -- 
> Pascal 
>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"theano-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to theano-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Re: [theano-users] Re: Is PyCUDA needed?

2016-10-17 Thread Pascal Lamblin
On Mon, Oct 17, 2016, senthil Nathan wrote:
> btw Im also having messages like , CVM does not support memory profile , 
> using stack VM. 

This is normal.

> > Dear Members , if I use CPU to run theano , is the pycuda need to be 
> > setup. Thanks

No, PyCUDA is not needed at all. We have some integration for adding
code written in PyCUDA into Theano, and this is what the test is trying
to do, but you are not missing on any feature if you do not have PyCUDA.

> > When i'm running the test im getting All-NaN slice encountered , All-NaN 
> > axis  encounteredas runtime warnings as well. Is it normal ? 

Yes, these are warnings emitted by NumPy when we test certain things.


-- 
Pascal

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"theano-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to theano-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.