Navigation Menu

Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Cannot remove entries from nonexistent #622

Closed
vinhqdang opened this issue Dec 26, 2015 · 26 comments
Closed

Cannot remove entries from nonexistent #622

vinhqdang opened this issue Dec 26, 2015 · 26 comments

Comments

@vinhqdang
Copy link

Hello everyone

I am using Anaconda on Mac OS 10.11

I created a new environment with anaconda

conda create -n tensorflow python

Then tried to install tensorflow

pip install --upgrade https://storage.googleapis.com/tensorflow/mac/tensorflow-0.6.0-py2-none-any.whl

but I faced an error

Installing collected packages: six, setuptools, protobuf, numpy, tensorflow
  Found existing installation: setuptools 19.1.1
**Cannot remove entries from nonexistent file /Users/qdang/anaconda/envs/tensorflow/lib/python2.7/site-packages/easy-install.pth**

How could I install tensorflow with anaconda python 2.7?

Btw, I have CUDA 3.0 (Nvidia 750M) with my Macbook? How could I utilize its power other than running only on CPU mode?

Thanks,

@vinhqdang
Copy link
Author

The solution is posted here

#135

It is for tensorflow 0.5.0 but still works with version 0.6.0

@sunformoon
Copy link

Hi, I followed the solution in the provided linking by downloading .whl file manually ( https://storage.googleapis.com/tensorflow/mac/tensorflow-0.6.0-py2-none-any.whl ) and run $pip install tensorflow-0.6.0-py2-none-any.whl.

And install the latest tensorflow by $ pip install --upgrade https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.6.0-cp27-none-linux_x86_64.whl

However, the issue still comes out:
Found existing installation: setuptools 19.1.1
Cannot remove entries from nonexistent file...

Could you please give more detailed solution when you solved the 'nonexistent' issue? Thanks a lot!

@vinhqdang
Copy link
Author

Hi

It works for me (in Mac OS 10.11). Hope it helps

http://vinhdq.blogspot.com/2015/12/installing-tensorflow-on-mac-os-1011.html http://vinhdq.blogspot.com/2015/12/installing-tensorflow-on-mac-os-1011.html

Best Regards

Vinh DANG

On 30 Dec 2015, at 07:14, Zhuotun Zhu notifications@github.com wrote:

Hi, I followed the solution in the provided linking by downloading .whl file manually ( https://storage.googleapis.com/tensorflow/mac/tensorflow-0.6.0-py2-none-any.whl https://storage.googleapis.com/tensorflow/mac/tensorflow-0.6.0-py2-none-any.whl ) and run $pip install tensorflow-0.6.0-py2-none-any.whl.

And install the latest tensorflow by $ pip install --upgrade https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.6.0-cp27-none-linux_x86_64.whl https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.6.0-cp27-none-linux_x86_64.whl
However, the issue still comes out:
Found existing installation: setuptools 19.1.1
Cannot remove entries from nonexistent file...

Could you please give more detailed solution when you solved the 'nonexistent' issue? Thanks a lot!


Reply to this email directly or view it on GitHub #622 (comment).

@sunformoon
Copy link

I actually have installed Tensorflow 0.6.0 successfully on Ubuntu 14.04.
I am wondering how to solve the "Found existing installation: setuptools 19.1.1 Cannot remove entries from nonexistent file" issue. Maybe it worked for you but not me due to the different OS?

@vinhqdang
Copy link
Author

Hi

How did you install tensor flow?

I faced a same problem when I installed with pip, but after I created a totally new environment there is no issue.

Best Regards

Vinh DANG

On 30 Dec 2015, at 07:31, Zhuotun Zhu notifications@github.com wrote:

I actually have installed Tensorflow 0.6.0 successfully on Ubuntu 14.04.
I am wondering how to solve the "Found existing installation: setuptools 19.1.1 Cannot remove entries from nonexistent file" issue. Maybe it worked for you but not me due to the different OS?


Reply to this email directly or view it on GitHub #622 (comment).

@sunformoon
Copy link

Hi, I installed the tensorflow on Ubuntu 14.04 as following:

$ conda install virtualenv
$ conda create --name=tensorflow_env python=2.7
$ source activate tensorflow_env
$ pip install --upgrade https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.5.0-cp27-none-linux_x86_64.whl

Then I updated to the 0.6.0 version. Although there is still the nonexistent issue, I tested the tensorflow successfully on MNIST demo. I think the issue is just somehow a warning that may not affect the use of tensorflow at present. However, it's better to solve it.

@vinhqdang
Copy link
Author

Why do you need to install version 0.5.0 then upgrade. Maybe it’s better to install 0.6.0 directly?


Best Regards

Vinh DANG

On 30 Dec 2015, at 08:06, Zhuotun Zhu notifications@github.com wrote:

Hi, I installed the tensorflow on Ubuntu 14.04 as following:

$ conda install virtualenv
$ conda create --name=tensorflow_env python=2.7
$ source activate tensorflow_env
$ pip install --upgrade https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.5.0-cp27-none-linux_x86_64.whl https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.5.0-cp27-none-linux_x86_64.whl
Then I updated to the 0.6.0 version. Although there is still the nonexistent issue, I tested the tensorflow successfully on MNIST demo. I think the issue is just somehow a warning that may not affect the use of tensorflow at present. However, it's better to solve it.


Reply to this email directly or view it on GitHub #622 (comment).

@sunformoon
Copy link

Hi,
Cause the instruction on homepage of Virtualenv installation is the 0.5.0 version. I didn't notice the latest 0.6.0 version. After I installed, I was informed from other source that maybe I could update to the 0.6.0 version. However, the nonexistent issue comes out both on my installation of 0.5.0 and 0.6.0.

@vinhqdang
Copy link
Author

Well

I have no idea about Ubuntu environment.

However, I could suggest you try (nothing to lose LOL):

  • create a new environment with conda (using sklearn environment instead of pure python)
  • install tensor flow 0.6.0 offline.

    Best Regards

Vinh DANG

On 30 Dec 2015, at 08:31, Zhuotun Zhu notifications@github.com wrote:

Hi,
Cause the instruction on homepage of Virtualenv installation is the 0.5.0 version. I didn't notice the latest 0.6.0 version. After I installed, I was informed from other source that maybe I could update to the 0.6.0 version. However, the nonexistent issue comes out both on my installation of 0.5.0 and 0.6.0.


Reply to this email directly or view it on GitHub #622 (comment).

@osdf
Copy link

osdf commented Jan 6, 2016

See here: pypa/pip#2751.
Possible work-around: pip install --upgrade --I setuptools

@frankcarey
Copy link

@osdf - yeah I hit this too. Used the following to solve it when using conda (after I already successfully installed using the ubuntu pip directly) so that I had iPython notebook (Jupyter) installed.

List conda environments
$ conda env list

Enable a conda environment..
$ source activate [your env or root]

check python..
$ which python
.. this should be something like ~/anaconda2/bin/python

$ pip install --upgrade https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.6.0-cp27-none-linux_x86_64.whl

.. got the error above, so added the --ignore-installed flag per pypa/pip#2751

$ pip install --upgrade https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.6.0-cp27-none-linux_x86_64.whl --ignore-installed

.. everything was successful..

$ python
>>> import tensorflow

.. success and back in business!

@naim
Copy link

naim commented Jan 14, 2016

Thanks @frankcarey for posting your workaround.

@inoryy
Copy link

inoryy commented Jan 21, 2016

Had same error with Python 3.4 version, solution by @frankcarey worked, thanks!

@kli-casia
Copy link

@frankcarey
I use your workaround and successfully installed tensorflow
But when I import it, there is something wrong.

---------------------------------------------------------------------------
ImportError                               Traceback (most recent call last)
<ipython-input-1-a649b509054f> in <module>()
----> 1 import tensorflow

/public/home/kli/anaconda/envs/tensorflow/lib/python2.7/site-packages/tensorflow/__init__.py in <module>()
     21 from __future__ import print_function
     22
---> 23 from tensorflow.python import *

/public/home/kli/anaconda/envs/tensorflow/lib/python2.7/site-packages/tensorflow/python/__init__.py in <module>()
     39 please exit the tensorflow source tree, and relaunch your python interpreter
     40 from there.""" % traceback.format_exc()
---> 41   raise ImportError(msg)
     42
     43 from tensorflow.core.framework.summary_pb2 import *

ImportError: Traceback (most recent call last):
  File "/public/home/kli/anaconda/envs/tensorflow/lib/python2.7/site-packages/tensorflow/python/__init__.py", line 35, in <module>
    from tensorflow.core.framework.graph_pb2 import *
  File "/public/home/kli/anaconda/envs/tensorflow/lib/python2.7/site-packages/tensorflow/core/framework/graph_pb2.py", line 6, in <module>
    from google.protobuf import descriptor as _descriptor
  File "/public/home/kli/anaconda/envs/tensorflow/lib/python2.7/site-packages/google/protobuf/descriptor.py", line 46, in <module>
    from google.protobuf.pyext import _message
ImportError: /lib64/libc.so.6: version `GLIBC_2.14' not found (required by /public/home/kli/anaconda/envs/tensorflow/lib/python2.7/site-packages/google/protobuf/pyext/_message.so)


Error importing tensorflow.  Unless you are using bazel,
you should not try to import tensorflow from its source directory;
please exit the tensorflow source tree, and relaunch your python interpreter
from there.

@yhuanghamu
Copy link

@kli-nlpr Did you figure out the GLIBC_2.14 problem?

@kli-casia
Copy link

@StevenHuang4321
I use centos6 when encounter this problem.
When I use Ubuntu14.04, the problem goes away.

@yhuanghamu
Copy link

Since our server use Centos 6, it is not possible for me to change to
ubuntu. Thanks.

On Tue, Mar 8, 2016 at 5:28 PM, Kai Li notifications@github.com wrote:

@StevenHuang4321 https://github.com/StevenHuang4321

I use centos6 when encounter this problem.
When I use Ubuntu14.04, the problem goes away.


Reply to this email directly or view it on GitHub
#622 (comment)
.

@kli-casia
Copy link

@StevenHuang4321
The possible solution is to update your glibc version.
But I think this is a complicated task and dangerous...

In fact I encounter a lot of problem when I use centos6 while doing deep learning research.
After change to Ubuntu, life becomes much easier.

@yhuanghamu
Copy link

Our system is CentOS 6.7, but it seems that only CentOS 7 supprot glib_2.14
or higher version.

Apparently, CentOS is not a good linux distribution to do develop and
research.

I already suggest my advisor to migrate our server from CentOS6.7 to Ubuntu
server.

On Tue, Mar 8, 2016 at 5:41 PM, Kai Li notifications@github.com wrote:

@StevenHuang4321 https://github.com/StevenHuang4321
The possible solution is to update your glibc version.
But I think this is a complicated task and dangerous...

In fact I encounter a lot of problem when I use centos6 while doing deep
learning research.


Reply to this email directly or view it on GitHub
#622 (comment)
.

@fengmei
Copy link

fengmei commented Mar 18, 2016

solution works, thanks!

@ppuliu
Copy link

ppuliu commented Apr 1, 2016

Another solution

pip install --ignore-installed --upgrade https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.7.1-cp27-none-linux_x86_64.whl

@bzamecnik
Copy link
Contributor

Observed on Ubuntu 15.10 Server in Anaconda 4.0.0 with python 3.4, tensorflow-0.7.1-cp34-none-linux_x86_64.whl GPU. Downloading and installing from local file works!

@wuyanghere
Copy link

@osdf
thanks for your solution!
it works
pip install --upgrade -I setuptools

@geyang
Copy link

geyang commented Aug 8, 2016

Both solutions worked for me:
1.

pip install --ignore-installed --upgrade https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.7.1-cp27-none-linux_x86_64.whl
pip install --upgrade -I setuptools

export TF_BINARY_URL=https://storage.googleapis.com/tensorflow/mac/gpu/tensorflow-0.10.0rc0-py3-none-any.whl
sudo pip3 install --upgrade $TF_BINARY_URL

sudo python3 -m pip install --upgrade https://storage.googleapis.com/tensorflow/mac/protobuf-3.0.0b2.post2-cp35-none-any.whl 

@brando90
Copy link

just as a comment...I accidentally downloaded the MAC version in linux because the --ignore-installed allowed me. So if it still doesn't work that might be it.

@ghost
Copy link

ghost commented Oct 13, 2016

1.download ez_setup.py
2.run script >python ez_setup.py
It works for me (in windows 7)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests