error: /usr/local/cuda-10.1/lib64/libcudnn.so: ELF section name out of range\ncollect2: error: ld returned 1 exit status\n'
Light-- opened this issue · comments
Light-- commented
my code
$ python3.6
Python 3.6.9 (default, Jul 3 2019, 15:36:16)
[GCC 5.4.0 20160609] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import theano
ERROR (theano.gpuarray): Could not initialize pygpu, support disabled
Traceback (most recent call last):
File "/usr/local/lib/python3.6/dist-packages/Theano-1.0.4-py3.6.egg/theano/gpuarray/__init__.py", line 227, in <module>
use(config.device)
File "/usr/local/lib/python3.6/dist-packages/Theano-1.0.4-py3.6.egg/theano/gpuarray/__init__.py", line 214, in use
init_dev(device, preallocate=preallocate)
File "/usr/local/lib/python3.6/dist-packages/Theano-1.0.4-py3.6.egg/theano/gpuarray/__init__.py", line 121, in init_dev
dnn.dnn_available.msg)
RuntimeError: You enabled cuDNN, but we aren't able to use it: cannot compile with cuDNN. We got this error:
b'/usr/bin/ld: error: /usr/local/cuda-10.1/lib64/libcudnn.so: ELF section name out of range\ncollect2: error: ld returned 1 exit status\n'
>>>
my theano configuration
[global]
device = cuda
floatX = float32
[dnn]
enabled = True
include_path=/usr/local/cuda-10.1/include
library_path=/usr/local/cuda-10.1/lib64
my theano package
$ pip3.6 show theano
Name: Theano
Version: 1.0.4
Summary: Optimizing compiler for evaluating mathematical expressions on CPUs and GPUs.
Home-page: http://deeplearning.net/software/theano/
Author: LISA laboratory, University of Montreal
Author-email: theano-dev@googlegroups.com
License: BSD
Location: /usr/local/lib/python3.6/dist-packages/Theano-1.0.4-py3.6.egg
Requires: numpy, scipy, six
Required-by:
Frédéric Bastien commented
Maybe this page will help you:
http://www.deeplearning.net/software/theano/library/gpuarray/dnn.html
You must install cudnn separately of Theano. Also make sure to install a version supported by Theano.
Theano isn’t maintained and probably doesn’t support the most recent version of cudnn.
From: Light-- <notifications@github.com>
Sent: Saturday, February 15, 2020 3:32 AM
To: Theano/Theano <Theano@noreply.github.com>
Cc: Subscribed <subscribed@noreply.github.com>
Subject: [Theano/Theano] error: /usr/local/cuda-10.1/lib64/libcudnn.so: ELF section name out of range\ncollect2: error: ld returned 1 exit status\n' (#6742)
my code
$ python3.6
Python 3.6.9 (default, Jul 3 2019, 15:36:16)
[GCC 5.4.0 20160609] on linux
Type "help", "copyright", "credits" or "license" for more information.
>> import theano
ERROR (theano.gpuarray): Could not initialize pygpu, support disabled
Traceback (most recent call last):
File "/usr/local/lib/python3.6/dist-packages/Theano-1.0.4-py3.6.egg/theano/gpuarray/__init__.py", line 227, in <module>
use(config.device)
File "/usr/local/lib/python3.6/dist-packages/Theano-1.0.4-py3.6.egg/theano/gpuarray/__init__.py", line 214, in use
init_dev(device, preallocate=preallocate)
File "/usr/local/lib/python3.6/dist-packages/Theano-1.0.4-py3.6.egg/theano/gpuarray/__init__.py", line 121, in init_dev
dnn.dnn_available.msg)
RuntimeError: You enabled cuDNN, but we aren't able to use it: cannot compile with cuDNN. We got this error:
b'/usr/bin/ld: error: /usr/local/cuda-10.1/lib64/libcudnn.so: ELF section name out of range\ncollect2: error: ld returned 1 exit status\n'
>>
my theano configuration
[global]
device = cuda
floatX = float32
[dnn]
enabled = True
include_path=/usr/local/cuda-10.1/include
library_path=/usr/local/cuda-10.1/lib64
my theano package
$ pip3.6 show theano
Name: Theano
Version: 1.0.4
Summary: Optimizing compiler for evaluating mathematical expressions on CPUs and GPUs.
Home-page: http://deeplearning.net/software/theano/
Author: LISA laboratory, University of Montreal
Author-email: theano-dev@googlegroups.com<mailto:theano-dev@googlegroups.com>
License: BSD
Location: /usr/local/lib/python3.6/dist-packages/Theano-1.0.4-py3.6.egg
Requires: numpy, scipy, six
Required-by:
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub<#6742?email_source=notifications&email_token=AABMF62NSGBECQ2JQECGYZTRC6R7FA5CNFSM4KVWSKG2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4INYFDZA>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AABMF6ZP6PYKB65ZPO27AEDRC6R7FANCNFSM4KVWSKGQ>.
…-----------------------------------------------------------------------------------
This email message is for the sole use of the intended recipient(s) and may contain
confidential information. Any unauthorized review, use, disclosure or distribution
is prohibited. If you are not the intended recipient, please contact the sender by
reply email and destroy all copies of the original message.
-----------------------------------------------------------------------------------