Theano: Not able to use GPU

Created on 11 Nov 2015 · 13Comments · Source: Theano/Theano

Hi,

After I successfully installed Theano and LibGPUArray and pyGPU on a Ubuntu 14.04 machine with a 4 NVIDIA Titan X cards with CUDA 7.5 and NVIDIA 352 drivers

I wanted to run the check1.py file with the following command:

THEANO_FLAGS=mode=FAST_RUN,device=gpu,floatX=float32 python check1.py

but this gave the following error

WARNING (theano.sandbox.cuda): CUDA is installed, but device gpu is not available  (error: Unable to get the number of gpus available: unknown error)
[Elemwise{exp,no_inplace}(<TensorType(float32, vector)>)]
Looping 1000 times took 2.364756 seconds
Result is [ 1.23178029  1.61879337  1.52278066 ...,  2.20771813  2.29967761
  1.62323284]
Used the cpu

So I tried running it like this with the assumption that I was perhaps not using the libgpuarry or pygpu

THEANO_FLAGS=mode=FAST_RUN,device=cuda,floatX=float32 python check1.py

However the outcome was this:

ERROR (theano.sandbox.gpuarray): Could not initialize pygpu, support disabled
Traceback (most recent call last):
  File "/usr/local/lib/python2.7/dist-packages/Theano-0.7.0-py2.7.egg/theano/sandbox/gpuarray/__init__.py", line 65, in <module>
    init_dev(config.device)
  File "/usr/local/lib/python2.7/dist-packages/Theano-0.7.0-py2.7.egg/theano/sandbox/gpuarray/__init__.py", line 49, in init_dev
    init_dev.devmap[dev] = pygpu.init(dev)
  File "pygpu/gpuarray.pyx", line 605, in pygpu.gpuarray.init (pygpu/gpuarray.c:8352)
  File "pygpu/gpuarray.pyx", line 576, in pygpu.gpuarray.pygpu_init (pygpu/gpuarray.c:8281)
  File "pygpu/gpuarray.pyx", line 981, in pygpu.gpuarray.GpuContext.__cinit__ (pygpu/gpuarray.c:12083)
GpuArrayException: API not initialized
[Elemwise{exp,no_inplace}(<TensorType(float32, vector)>)]

Running it with cuda0 instead of cuda gave me the following error message

ERROR (theano.sandbox.gpuarray): Could not initialize pygpu, support disabled
Traceback (most recent call last):
  File "/usr/local/lib/python2.7/dist-packages/Theano-0.7.0-py2.7.egg/theano/sandbox/gpuarray/__init__.py", line 65, in <module>
    init_dev(config.device)
  File "/usr/local/lib/python2.7/dist-packages/Theano-0.7.0-py2.7.egg/theano/sandbox/gpuarray/__init__.py", line 49, in init_dev
    init_dev.devmap[dev] = pygpu.init(dev)
  File "pygpu/gpuarray.pyx", line 605, in pygpu.gpuarray.init (pygpu/gpuarray.c:8352)
  File "pygpu/gpuarray.pyx", line 576, in pygpu.gpuarray.pygpu_init (pygpu/gpuarray.c:8281)
  File "pygpu/gpuarray.pyx", line 981, in pygpu.gpuarray.GpuContext.__cinit__ (pygpu/gpuarray.c:12083)
GpuArrayException: Unknown internal error

Can you please help me

Source

TeunKrikke

Most helpful comment

Use device=gpu0

If that work, try device=cuda0

To be sure you need to reboot after installing the driver.

Fred
Le 11 nov. 2015 10:31, "TeunKrikke" [email protected] a écrit :

Hi,

After I successfully installed Theano and LibGPUArray and pyGPU on a
Ubuntu 14.04 machine with a 4 NVIDIA Titan X cards with CUDA 7.5 and NVIDIA
352 drivers

I wanted to run the check1.py file with the following command:

THEANO_FLAGS=mode=FAST_RUN,device=gpu,floatX=float32 python check1.py

but this gave the following error

WARNING (theano.sandbox.cuda): CUDA is installed, but device gpu is not available (error: Unable to get the number of gpus available: unknown error)
[Elemwise{exp,no_inplace}( Looping 1000 times took 2.364756 seconds
Result is [ 1.23178029 1.61879337 1.52278066 ..., 2.20771813 2.29967761
1.62323284]
Used the cpu

So I tried running it like this with the assumption that I was perhaps not
using the libgpuarry or pygpu

THEANO_FLAGS=mode=FAST_RUN,device=cuda,floatX=float32 python check1.py

However the outcome was this:

ERROR (theano.sandbox.gpuarray): Could not initialize pygpu, support disabled
Traceback (most recent call last):
File "/usr/local/lib/python2.7/dist-packages/Theano-0.7.0-py2.7.egg/theano/sandbox/gpuarray/init.py", line 65, in
init_dev(config.device)
File "/usr/local/lib/python2.7/dist-packages/Theano-0.7.0-py2.7.egg/theano/sandbox/gpuarray/init.py", line 49, in init_dev
init_dev.devmap[dev] = pygpu.init(dev)
File "pygpu/gpuarray.pyx", line 605, in pygpu.gpuarray.init (pygpu/gpuarray.c:8352)
File "pygpu/gpuarray.pyx", line 576, in pygpu.gpuarray.pygpu_init (pygpu/gpuarray.c:8281)
File "pygpu/gpuarray.pyx", line 981, in pygpu.gpuarray.GpuContext.cinit (pygpu/gpuarray.c:12083)
GpuArrayException: API not initialized
[Elemwise{exp,no_inplace}(
Running it with cuda0 instead of cuda gave me the following error message

ERROR (theano.sandbox.gpuarray): Could not initialize pygpu, support disabled
Traceback (most recent call last):
File "/usr/local/lib/python2.7/dist-packages/Theano-0.7.0-py2.7.egg/theano/sandbox/gpuarray/init.py", line 65, in
init_dev(config.device)
File "/usr/local/lib/python2.7/dist-packages/Theano-0.7.0-py2.7.egg/theano/sandbox/gpuarray/init.py", line 49, in init_dev
init_dev.devmap[dev] = pygpu.init(dev)
File "pygpu/gpuarray.pyx", line 605, in pygpu.gpuarray.init (pygpu/gpuarray.c:8352)
File "pygpu/gpuarray.pyx", line 576, in pygpu.gpuarray.pygpu_init (pygpu/gpuarray.c:8281)
File "pygpu/gpuarray.pyx", line 981, in pygpu.gpuarray.GpuContext.cinit (pygpu/gpuarray.c:12083)
GpuArrayException: Unknown internal error

Can you please help me

—
Reply to this email directly or view it on GitHub
https://github.com/Theano/Theano/issues/3629.

nouiz on 12 Nov 2015

👍2

All 13 comments

Use device=gpu0

If that work, try device=cuda0

To be sure you need to reboot after installing the driver.

Fred
Le 11 nov. 2015 10:31, "TeunKrikke" [email protected] a écrit :

Hi,

After I successfully installed Theano and LibGPUArray and pyGPU on a
Ubuntu 14.04 machine with a 4 NVIDIA Titan X cards with CUDA 7.5 and NVIDIA
352 drivers

I wanted to run the check1.py file with the following command:

THEANO_FLAGS=mode=FAST_RUN,device=gpu,floatX=float32 python check1.py

but this gave the following error

WARNING (theano.sandbox.cuda): CUDA is installed, but device gpu is not available (error: Unable to get the number of gpus available: unknown error)
[Elemwise{exp,no_inplace}( Looping 1000 times took 2.364756 seconds
Result is [ 1.23178029 1.61879337 1.52278066 ..., 2.20771813 2.29967761
1.62323284]
Used the cpu

So I tried running it like this with the assumption that I was perhaps not
using the libgpuarry or pygpu

THEANO_FLAGS=mode=FAST_RUN,device=cuda,floatX=float32 python check1.py

However the outcome was this:

ERROR (theano.sandbox.gpuarray): Could not initialize pygpu, support disabled
Traceback (most recent call last):
File "/usr/local/lib/python2.7/dist-packages/Theano-0.7.0-py2.7.egg/theano/sandbox/gpuarray/init.py", line 65, in
init_dev(config.device)
File "/usr/local/lib/python2.7/dist-packages/Theano-0.7.0-py2.7.egg/theano/sandbox/gpuarray/init.py", line 49, in init_dev
init_dev.devmap[dev] = pygpu.init(dev)
File "pygpu/gpuarray.pyx", line 605, in pygpu.gpuarray.init (pygpu/gpuarray.c:8352)
File "pygpu/gpuarray.pyx", line 576, in pygpu.gpuarray.pygpu_init (pygpu/gpuarray.c:8281)
File "pygpu/gpuarray.pyx", line 981, in pygpu.gpuarray.GpuContext.cinit (pygpu/gpuarray.c:12083)
GpuArrayException: API not initialized
[Elemwise{exp,no_inplace}(
Running it with cuda0 instead of cuda gave me the following error message

ERROR (theano.sandbox.gpuarray): Could not initialize pygpu, support disabled
Traceback (most recent call last):
File "/usr/local/lib/python2.7/dist-packages/Theano-0.7.0-py2.7.egg/theano/sandbox/gpuarray/init.py", line 65, in
init_dev(config.device)
File "/usr/local/lib/python2.7/dist-packages/Theano-0.7.0-py2.7.egg/theano/sandbox/gpuarray/init.py", line 49, in init_dev
init_dev.devmap[dev] = pygpu.init(dev)
File "pygpu/gpuarray.pyx", line 605, in pygpu.gpuarray.init (pygpu/gpuarray.c:8352)
File "pygpu/gpuarray.pyx", line 576, in pygpu.gpuarray.pygpu_init (pygpu/gpuarray.c:8281)
File "pygpu/gpuarray.pyx", line 981, in pygpu.gpuarray.GpuContext.cinit (pygpu/gpuarray.c:12083)
GpuArrayException: Unknown internal error

Can you please help me

—
Reply to this email directly or view it on GitHub
https://github.com/Theano/Theano/issues/3629.

nouiz on 12 Nov 2015

👍2

Thanks for your quick response

Rebooted the machine. After that tried

THEANO_FLAGS=device=gpu0 python check1.py

Which gives :

WARNING (theano.sandbox.cuda): CUDA is installed, but device gpu0 is not available  (error: Unable to get the number of gpus available: unknown error)

and tried

THEANO_FLAGS=device=cuda0 python check1.py

with this result:

ERROR (theano.sandbox.gpuarray): Could not initialize pygpu, support disabled
Traceback (most recent call last):
  File "/usr/local/lib/python2.7/dist-packages/Theano-0.7.0-py2.7.egg/theano/sandbox/gpuarray/__init__.py", line 65, in <module>
    init_dev(config.device)
  File "/usr/local/lib/python2.7/dist-packages/Theano-0.7.0-py2.7.egg/theano/sandbox/gpuarray/__init__.py", line 49, in init_dev
    init_dev.devmap[dev] = pygpu.init(dev)
  File "pygpu/gpuarray.pyx", line 605, in pygpu.gpuarray.init (pygpu/gpuarray.c:8352)
  File "pygpu/gpuarray.pyx", line 576, in pygpu.gpuarray.pygpu_init (pygpu/gpuarray.c:8281)
  File "pygpu/gpuarray.pyx", line 981, in pygpu.gpuarray.GpuContext.__cinit__ (pygpu/gpuarray.c:12083)
GpuArrayException: unknown error

Running

python -c "import pygpu;pygpu.test()"

Gives me six times the same error:

ERROR: Failure: GpuArrayException (Unknown error)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/usr/lib/python2.7/dist-packages/nose/loader.py", line 411, in loadTestsFromName
    addr.filename, addr.module)
  File "/usr/lib/python2.7/dist-packages/nose/importer.py", line 47, in importFromPath
    return self.importFromDir(dir_path, fqname)
  File "/usr/lib/python2.7/dist-packages/nose/importer.py", line 94, in importFromDir
    mod = load_module(part_fqname, fh, filename, desc)
  File "/usr/local/lib/python2.7/dist-packages/pygpu-0.2.1-py2.7-linux-x86_64.egg/pygpu/tests/test_tools.py", line 5, in <module>
    from .support import (guard_devsup, rand, check_flags, check_meta, check_all,
  File "/usr/local/lib/python2.7/dist-packages/pygpu-0.2.1-py2.7-linux-x86_64.egg/pygpu/tests/support.py", line 32, in <module>
    context = gpuarray.init(get_env_dev())
  File "pygpu/gpuarray.pyx", line 605, in pygpu.gpuarray.init (pygpu/gpuarray.c:8352)
  File "pygpu/gpuarray.pyx", line 576, in pygpu.gpuarray.pygpu_init (pygpu/gpuarray.c:8281)
  File "pygpu/gpuarray.pyx", line 981, in pygpu.gpuarray.GpuContext.__cinit__ (pygpu/gpuarray.c:12083)
GpuArrayException: Unknown error

TeunKrikke on 12 Nov 2015

There is a problem with the way your computer is configured.

Is the nvidia driver running? Maybe not. Can you execute nvidia-smi on the
command line?

Maybe reinstalling nvidia driver will help.

On Thu, Nov 12, 2015 at 7:14 AM, TeunKrikke [email protected]
wrote:

Thanks for your quick response

Rebooted the machine. After that tried

THEANO_FLAGS=device=gpu0 python check1.py

Which gives :

WARNING (theano.sandbox.cuda): CUDA is installed, but device gpu0 is not available (error: Unable to get the number of gpus available: unknown error)

and tried

THEANO_FLAGS=device=cuda0 python check1.py

with this result:

ERROR (theano.sandbox.gpuarray): Could not initialize pygpu, support disabled
Traceback (most recent call last):
File "/usr/local/lib/python2.7/dist-packages/Theano-0.7.0-py2.7.egg/theano/sandbox/gpuarray/init.py", line 65, in
init_dev(config.device)
File "/usr/local/lib/python2.7/dist-packages/Theano-0.7.0-py2.7.egg/theano/sandbox/gpuarray/init.py", line 49, in init_dev
init_dev.devmap[dev] = pygpu.init(dev)
File "pygpu/gpuarray.pyx", line 605, in pygpu.gpuarray.init (pygpu/gpuarray.c:8352)
File "pygpu/gpuarray.pyx", line 576, in pygpu.gpuarray.pygpu_init (pygpu/gpuarray.c:8281)
File "pygpu/gpuarray.pyx", line 981, in pygpu.gpuarray.GpuContext.cinit (pygpu/gpuarray.c:12083)
GpuArrayException: unknown error

Running

python -c "import pygpu;pygpu.test()"

Gives me six times the same error:

ERROR: Failure: GpuArrayException (Unknown error)

Traceback (most recent call last):
File "/usr/lib/python2.7/dist-packages/nose/loader.py", line 411, in loadTestsFromName
addr.filename, addr.module)
File "/usr/lib/python2.7/dist-packages/nose/importer.py", line 47, in importFromPath
return self.importFromDir(dir_path, fqname)
File "/usr/lib/python2.7/dist-packages/nose/importer.py", line 94, in importFromDir
mod = load_module(part_fqname, fh, filename, desc)
File "/usr/local/lib/python2.7/dist-packages/pygpu-0.2.1-py2.7-linux-x86_64.egg/pygpu/tests/test_tools.py", line 5, in
from .support import (guard_devsup, rand, check_flags, check_meta, check_all,
File "/usr/local/lib/python2.7/dist-packages/pygpu-0.2.1-py2.7-linux-x86_64.egg/pygpu/tests/support.py", line 32, in
context = gpuarray.init(get_env_dev())
File "pygpu/gpuarray.pyx", line 605, in pygpu.gpuarray.init (pygpu/gpuarray.c:8352)
File "pygpu/gpuarray.pyx", line 576, in pygpu.gpuarray.pygpu_init (pygpu/gpuarray.c:8281)
File "pygpu/gpuarray.pyx", line 981, in pygpu.gpuarray.GpuContext.cinit (pygpu/gpuarray.c:12083)
GpuArrayException: Unknown error

—
Reply to this email directly or view it on GitHub
https://github.com/Theano/Theano/issues/3629#issuecomment-156088000.

nouiz on 12 Nov 2015

👍1

Yes this is the outcome:

| NVIDIA-SMI 352.39     Driver Version: 352.39         |                       
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|===============================+======================+======================|
|   0  GeForce GTX TIT...  Off  | 0000:02:00.0      On |                  N/A |
| 22%   39C    P8    16W / 250W |    216MiB / 12285MiB |      0%      Default |
+-------------------------------+----------------------+----------------------+
|   1  GeForce GTX TIT...  Off  | 0000:03:00.0     Off |                  N/A |
| 22%   43C    P8    16W / 250W |     23MiB / 12287MiB |      0%      Default |
+-------------------------------+----------------------+----------------------+
|   2  GeForce GTX TIT...  Off  | 0000:82:00.0     Off |                  N/A |
| 22%   43C    P8    17W / 250W |     23MiB / 12287MiB |      0%      Default |
+-------------------------------+----------------------+----------------------+
|   3  GeForce GTX TIT...  Off  | 0000:83:00.0     Off |                  N/A |
| 22%   41C    P8    16W / 250W |     23MiB / 12287MiB |      0%      Default |
+-------------------------------+----------------------+----------------------+

+-----------------------------------------------------------------------------+
| Processes:                                                       GPU Memory |
|  GPU       PID  Type  Process name                               Usage      |
|=============================================================================|
|    0      1997    G   /usr/bin/X                                     192MiB |
+-----------------------------------------------------------------------------+

TeunKrikke on 12 Nov 2015

Reinstalled the gpu driver but it gave the same result

TeunKrikke on 12 Nov 2015

Try to run nvidia samples in the sdk. If they don't work, Theano can't work.

Did you install the cuda sdk that have nvcc? I suppose so.

On Thu, Nov 12, 2015 at 9:24 AM, TeunKrikke [email protected]
wrote:

Reinstalled the gpu driver but it gave the same result

—
Reply to this email directly or view it on GitHub
https://github.com/Theano/Theano/issues/3629#issuecomment-156114989.

nouiz on 12 Nov 2015

Hi,

I have reinstalled the cuda 7.5 toolkit and used the included display driver. Before I used the apt-get display driver but that seemed to be the source of all the trouble. It works now with cuda0 and gpu 0 as parameters.

Thanks for all you help!!!

TeunKrikke on 12 Nov 2015

Hello @TeunKrikke,

Can you please elaborate the work which you did to work gpu 0.

I am getting same error on EC2 p2.2 large instance.

I have cuda-8 installed.

nvidia-smi gives me ouptut.

+-----------------------------------------------------------------------------+
| Processes: GPU Memory |
| GPU PID Type Process name Usage |
|=============================================================================|
| No running processes found |
+-----------------------------------------------------------------------------+

nvcc --version give me cuda version 8.0 installed.

But I am getting below warning:

modprobe: ERROR: could not insert 'nvidia_375_uvm': Invalid argument
WARNING (theano.sandbox.cuda): The cuda backend is deprecated and will be red in the next release (v0.10). Please switch to the gpuarray backend. You cet more information about how to switch at this URL:
https://github.com/Theano/Theano/wiki/Converting-to-the-new-gpu-back-end%28rray%29

WARNING (theano.sandbox.cuda): CUDA is installed, but device gpu 0 is not avble (error: Unable to get the number of gpus available: unknown error)
Using Theano backend.

Thank you for your help.

Regards,
Bhushan

muthiyanbhushan on 28 Jun 2017

this is a problem with the driver. Did you started the VM from a VM with
cuda and the driver installed? If not, I would recommand to start from such
working VM.

On Wed, Jun 28, 2017 at 2:53 PM BHUSHAN MUTHIYAN notifications@github.com
wrote:

Hello @TeunKrikke https://github.com/teunkrikke,

Can you please elaborate the work which you did to work gpu 0.

I am getting same error on EC2 p2.2 large instance.

I have cuda-8 installed.

nvidia-smi gives me ouptut.

nvidia-smi
Wed Jun 28 18:45:13 2017

+-----------------------------------------------------------------------------+
| NVIDIA-SMI 367.57 Driver Version: 367.57 |

|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |

|===============================+======================+======================|

| 0 Tesla K80 Off | 0000:00:0F.0 Off | 0 |
| N/A 48C P8 28W / 149W | 0MiB / 11439MiB | 0% Default |

+-------------------------------+----------------------+----------------------+
| 1 Tesla K80 Off | 0000:00:10.0 Off | 0 |
| N/A 40C P8 31W / 149W | 0MiB / 11439MiB | 0% Default |

+-------------------------------+----------------------+----------------------+
| 2 Tesla K80 Off | 0000:00:11.0 Off | 0 |
| N/A 58C P8 30W / 149W | 0MiB / 11439MiB | 0% Default |

+-------------------------------+----------------------+----------------------+
| 3 Tesla K80 Off | 0000:00:12.0 Off | 0 |
| N/A 48C P8 31W / 149W | 0MiB / 11439MiB | 0% Default |

+-------------------------------+----------------------+----------------------+
| 4 Tesla K80 Off | 0000:00:13.0 Off | 0 |
| N/A 52C P8 57W / 149W | 0MiB / 11439MiB | 0% Default |

+-------------------------------+----------------------+----------------------+
| 5 Tesla K80 Off | 0000:00:14.0 Off | 0 |
| N/A 43C P0 69W / 149W | 0MiB / 11439MiB | 0% Default |

+-------------------------------+----------------------+----------------------+
| 6 Tesla K80 Off | 0000:00:15.0 Off | 0 |
| N/A 52C P0 56W / 149W | 0MiB / 11439MiB | 0% Default |

+-------------------------------+----------------------+----------------------+
| 7 Tesla K80 Off | 0000:00:16.0 Off | 0 |
| N/A 45C P0 69W / 149W | 0MiB / 11439MiB | 0% Default |

+-------------------------------+----------------------+----------------------+
| 8 Tesla K80 Off | 0000:00:17.0 Off | 0 |
| N/A 48C P0 55W / 149W | 0MiB / 11439MiB | 0% Default |

+-------------------------------+----------------------+----------------------+
| 9 Tesla K80 Off | 0000:00:18.0 Off | 0 |
| N/A 41C P0 69W / 149W | 0MiB / 11439MiB | 0% Default |

+-------------------------------+----------------------+----------------------+
| 10 Tesla K80 Off | 0000:00:19.0 Off | 0 |
| N/A 49C P0 61W / 149W | 0MiB / 11439MiB | 0% Default |

+-------------------------------+----------------------+----------------------+
| 11 Tesla K80 Off | 0000:00:1A.0 Off | 0 |
| N/A 44C P0 71W / 149W | 0MiB / 11439MiB | 0% Default |

+-------------------------------+----------------------+----------------------+
| 12 Tesla K80 Off | 0000:00:1B.0 Off | 0 |
| N/A 46C P0 57W / 149W | 0MiB / 11439MiB | 0% Default |

+-------------------------------+----------------------+----------------------+
| 13 Tesla K80 Off | 0000:00:1C.0 Off | 0 |
| N/A 43C P0 69W / 149W | 0MiB / 11439MiB | 0% Default |

+-------------------------------+----------------------+----------------------+
| 14 Tesla K80 Off | 0000:00:1D.0 Off | 0 |
| N/A 49C P0 57W / 149W | 0MiB / 11439MiB | 0% Default |

+-------------------------------+----------------------+----------------------+
| 15 Tesla K80 Off | 0000:00:1E.0 Off | 0 |
| N/A 44C P0 68W / 149W | 0MiB / 11439MiB | 0% Default |

+-------------------------------+----------------------+----------------------+

+-----------------------------------------------------------------------------+
| Processes: GPU Memory |
| GPU PID Type Process name Usage |

|=============================================================================|

| No running processes found |

+-----------------------------------------------------------------------------+

nvcc --version give me cuda version 8.0 installed.

But I am getting below warning:

modprobe: ERROR: could not insert 'nvidia_375_uvm': Invalid argument
WARNING (theano.sandbox.cuda): The cuda backend is deprecated and will be
red in the next release (v0.10). Please switch to the gpuarray backend. You
cet more information about how to switch at this URL:

https://github.com/Theano/Theano/wiki/Converting-to-the-new-gpu-back-end%28rray%29

WARNING (theano.sandbox.cuda): CUDA is installed, but device gpu 0 is not
avble (error: Unable to get the number of gpus available: unknown error)
Using Theano backend.

Thank you for your help.

Regards,
Bhushan

—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
https://github.com/Theano/Theano/issues/3629#issuecomment-311753747, or mute
the thread
https://github.com/notifications/unsubscribe-auth/AALC-01iEImjtjg06t5ZpmcvHJl05LzJks5sIqE4gaJpZM4GgT6v
.

nouiz on 28 Jun 2017

@muthiyanbhushan Given it is a vm I don't know if you have rights to do anything with the driver. In my case it was reinstalling the driver and removing apt-get version at the time. Since then I have not updated it regulary and have adopted the if it isn't broken than don't fix it style :).

The easiest option in your case would be to follow @nouiz's advice.

The more difficult option would involve installing cuda and using the display driver that it provides but I am not sure if that is possible because I have not worked with that VM.

TeunKrikke on 28 Jun 2017

hi
i have 1060 gtx gpu and cuda v8 with cudnn 5.1 is installed.
i have theano 0.8.2
i got this error.

EnvironmentError: ('The following error happened while compiling the node', GpuFromHost(Elemwise{exp,no_inplace}.0), '\n', 'You forced the use of gpu device gpu, but CUDA initialization failed with error:\ncuda unavailable')

MahdiHeybati on 23 Apr 2018

how can i fix this issue??

MahdiHeybati on 23 Apr 2018

Do not re-activate old issue for non-related problems. We do not support that version of Theano anymore. I would suggest that you try the latest version or find another framework (see https://groups.google.com/forum/#!msg/theano-users/7Poq8BZutbY/rNCIfvAEAwAJ)

nouiz on 23 Apr 2018

Was this page helpful?

0 / 5 - 0 ratings