Not able to change conda environment

[PBhar]

When a user creates a new conda environment, he expects the environment to be be completely isolated from the global path. But in this case, the path remains fixed (/glob/intel-python/python3/bin//python)  and I'm not able to change.

I've attached the screenshot below.

[AthiraM_Intel]

Hi,

We are unable to recreate your issue from our end.

Could you please try creating a fresh environment and see if the issue still persists.

If yes, please share the following:

1.  Steps you used to create the environment

2. Result of "echo $PATH"

3. List of folders in your environment bin folder using ls command.

[PBhar]

I want isolated environments. I want two conda environments which are completely seperate from each other. For some reason, all the packages are loaded from /glob/intel-python/python3/bin//python. I want different environments.

(base) u37216@login-2:~$ which python
/glob/intel-python/python3/bin//python
(base) u37216@login-2:~$ conda activate torch
(torch) u37216@login-2:~$ which python
/glob/intel-python/python3/bin//python

In torch environment, the python path should be different.

[AthiraM_Intel]

Hi,

Thanks for the reply.

Could you please share the following details

1. Result of "echo $PATH"   after exporting the path using below command.    

    export PATH="/home/uxxxxx/.conda/envs/env_name/bin":$PATH

2. List of folders in your environment bin folder using below commands .

    cd /home/uxxxxx/.conda/envs/env_name/bin

    ls

[PBhar]

(base) u37216@login-2:~$ echo $PATH
/glob/development-tools/versions/intel-parallel-studio/compilers_and_libraries_2019.3.199/linux/bin/intel64:/glob/development-tools/versions/intel-parallel-studio/compilers_and_libraries_2019.3.199/linux/mpi/intel64/libfabric/bin:/glob/development-tools/versions/intel-parallel-studio/compilers_and_libraries_2019.3.199/linux/mpi/intel64/bin:/glob/development-tools/versions/intel-parallel-studio/debugger_2019/gdb/intel64/bin:/glob/intel-python/python3/bin/:/glob/intel-python/python2/bin/:/home/u37216/miniconda3/bin:/home/u37216/miniconda3/bin:/home/u37216/.conda/envs/torch/bin:/glob/intel-python/versions/2018u2/intelpython3/bin:/glob/intel-python/versions/2018u2/intelpython3/condabin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/usr/local/games:/home/u37216/.local/bin:/home/u37216/bin:/bin
(base) u37216@login-2:~$ conda activate torch
(torch) u37216@login-2:~$ echo $PATH
/glob/development-tools/versions/intel-parallel-studio/compilers_and_libraries_2019.3.199/linux/bin/intel64:/glob/development-tools/versions/intel-parallel-studio/compilers_and_libraries_2019.3.199/linux/mpi/intel64/libfabric/bin:/glob/development-tools/versions/intel-parallel-studio/compilers_and_libraries_2019.3.199/linux/mpi/intel64/bin:/glob/development-tools/versions/intel-parallel-studio/debugger_2019/gdb/intel64/bin:/glob/intel-python/python3/bin/:/glob/intel-python/python2/bin/:/home/u37216/miniconda3/bin:/home/u37216/miniconda3/bin:/home/u37216/.conda/envs/torch/bin:/home/u37216/.conda/envs/torch/bin:/glob/intel-python/versions/2018u2/intelpython3/condabin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/usr/local/games:/home/u37216/.local/bin:/home/u37216/bin:/bin
(torch) u37216@login-2:~$ cd /home/u37216/.conda/envs/torch/bin
(torch) u37216@login-2:~/.conda/envs/torch/bin$ ls
2to3                      jupyter-nbconvert        openssl            sqlite3
2to3-3.7                  jupyter-nbextension      pandoc             tclsh8.6
c_rehash                  jupyter-notebook         pandoc-citeproc    unlzma
curve_keygen              jupyter-run              pip                unxz
easy_install              jupyter-serverextension  protoc             wheel
easy_install-3.7          jupyter-troubleshoot     pydoc              wish8.6
idle3                     jupyter-trust            pydoc3             xz
idle3.7                   lzcat                    pydoc3.7           xzcat
iptest                    lzcmp                    pygmentize         xzcmp
iptest3                   lzdiff                   python             xzdec
ipython                   lzegrep                  python3            xzdiff
ipython3                  lzfgrep                  python3.7          xzegrep
jsonschema                lzgrep                   python3.7-config   xzfgrep
jupyter                   lzless                   python3.7m         xzgrep
jupyter-bundlerextension  lzma                     python3.7m-config  xzless
jupyter-kernel            lzmadec                  python3-config     xzmore
jupyter-kernelspec        lzmainfo                 pyvenv
jupyter-migrate           lzmore                   pyvenv-3.7

[AthiraM_Intel]

Hi,

We have observed this issue in some DevCloud accounts.However we were able to resolve the issue on exporting the PATH of the created environment. 

PFA screenshot, which shows that after exporting the PATH it took the conda environment path.

We have informed the concerned admin team regarding your issue and will get back to you soon.

[AthiraM_Intel]

Hi,

Thank you for your patience.

Looks like the issue is not persistent.Could you please follow the below steps in a fresh terminal and share the results if the issue still persists.

1. qsub -I (login to compute node)

2. conda activate <env_name1>

3   which python

4  export PATH="/home/uxxxx/.conda/envs/<env_name1>/bin":$PATH

5. which python

6. conda deactivate

7. conda activate <env_name2>

8. which python

9. export PATH="/home/uxxxx/.conda/envs/<env_name2>/bin":$PATH

10. which python

Please find the attached screenshot , in which we were successfully able to change the environments and PATH variables using the above mentioned steps.

Thanks

[PBhar]

This issue is persistent.

Please read this:

1. Firstly when I conda activate a new environment, I shouldn't have to export PATH, activating an environment already means I require a new isolated environment

2. Secondly exporting PATH doesn't work. It still points to the base intel environment which is a faulty implementation.

3. I require isolated environments. When I conda create a new environment, I shouldn't be able to import any packages.

Here is the output as you asked

(base) u37216@s001-n116:~$ conda info --envs
# conda environments:
#
t                        /home/u37216/.conda/envs/t
torch                    /home/u37216/.conda/envs/torch
                         /home/u37216/miniconda3
                         /home/u37216/miniconda3/envs/ee
base                  *  /opt/intel/inteloneapi/intelpython/latest
pytorch                  /opt/intel/inteloneapi/intelpython/latest/envs/pytorch
pytorch-1.3.1            /opt/intel/inteloneapi/intelpython/latest/envs/pytorch-1.3.1
tensorflow               /opt/intel/inteloneapi/intelpython/latest/envs/tensorflow
tensorflow-1.15.0        /opt/intel/inteloneapi/intelpython/latest/envs/tensorflow-1.15.0

(base) u37216@s001-n116:~$ which python
/glob/intel-python/python3/bin/python
(base) u37216@s001-n116:~$ conda activate t
(t) u37216@s001-n116:~$ which python
/glob/intel-python/python3/bin/python
(t) u37216@s001-n116:~$ export PATH="/home/u37216/.conda/envs/t/bin":$PATH
(t) u37216@s001-n116:~$ which python
/glob/intel-python/python3/bin/python
(t) u37216@s001-n116:~$ python3
Python 3.6.3 |Intel Corporation| (default, Feb 12 2018, 06:37:09)
[GCC 4.8.2 20140120 (Red Hat 4.8.2-15)] on linux
Type "help", "copyright", "credits" or "license" for more information.
Intel(R) Distribution for Python is brought to you by Intel Corporation.
Please check out: https://software.intel.com/en-us/python-distribution
>>> import torch

>>> >>>
(t) u37216@s001-n116:~$

[PBhar]

Even after explicilty specifying the PYTHONPATH, python distribution and inside a conda environment, I'm able to import torch which shouldn't happen because I create a new environment. It means even after changing the path, all of the things still points to /intel/glob.

(pytorch) u37216@login-2:~$ PYTHONPATH=/home/u37216/miniconda3/lib/python3.7/ python3.7
Python 3.7.4 (default, Aug 13 2019, 20:35:49)
[GCC 7.3.0] :: Anaconda, Inc. on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import torch
>>>

(/glob/intel-python/versions/2018u2/intelpython3) u37216@login-2:~$ conda create --name pytorch

(pytorch) u37216@login-2:~$ export PYTHONPATH=/home/u37216/miniconda3/lib/python3.6/
(pytorch) u37216@login-2:~$ export PYTHONPATH=/home/u37216/miniconda3/lib/python3.7/
(pytorch) u37216@login-2:~$ unset PYTHONPATH
(pytorch) u37216@login-2:~$ PYTHONPATH

[AthiraM_Intel]

Hi Prajjwal,

We are forwarding this case to DevCloud admin team.

[AthiraM_Intel]

Hi Prajjwal.

Could you please check if the issue exists. If it still persists, we will create a new account for you. 

Thanks

[PBhar]

Yes it persists. 

[AthiraM_Intel]

Hi Prajjwal,

We have created a 120 days new account for you. This would fix your isolated conda environment problem( /glob/ issue). 

Could you please check and confirm?

Thanks

[PBhar]

This will do for now. Thanks!

[AthiraM_Intel]

Hi Prajjwal,

Hope your issue is resolved. Could you please confirm so that we can close this case.

Thanks

[PBhar]

Things have gotten worse now. Jupyterlab server takes a long time to load. Installing conda packages have become more challenging. Don't know why no one takes a proper look.For instance:

(base) u39403@s001-n152:~$ conda info --envs
# conda environments:
#
base                  *  /glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest
2021.1-beta05            /glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/envs/2021.1-beta05
pytorch                  /glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/envs/pytorch
pytorch-1.4.0            /glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/envs/pytorch-1.4.0
tensorflow               /glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/envs/tensorflow
tensorflow-1.15.0        /glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/envs/tensorflow-1.15.0
torch                    /home/u39403/.conda/envs/torch

(base) u39403@s001-n152:~$ conda activate pytorch-1.4.0
(pytorch-1.4.0) u39403@s001-n152:~$ python3
Python 3.6.3 |Intel Corporation| (default, Feb 12 2018, 06:37:09)
[GCC 4.8.2 20140120 (Red Hat 4.8.2-15)] on linux
Type "help", "copyright", "credits" or "license" for more information.
Intel(R) Distribution for Python is brought to you by Intel Corporation.
Please check out: https://software.intel.com/en-us/python-distribution
>>> import torch
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ModuleNotFoundError: No module named 'torch'
>>>
(pytorch-1.4.0) u39403@s001-n152:~$ conda activate torch
-bash: /home/u39403/.conda/envs/torch/etc/conda/activate.d/mpivars.activate.sh: No such file or directory
(torch) u39403@s001-n152:~$ python3
Python 3.6.3 |Intel Corporation| (default, Feb 12 2018, 06:37:09)
[GCC 4.8.2 20140120 (Red Hat 4.8.2-15)] on linux
Type "help", "copyright", "credits" or "license" for more information.
Intel(R) Distribution for Python is brought to you by Intel Corporation.
Please check out: https://software.intel.com/en-us/python-distribution
>>> import torch
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ModuleNotFoundError: No module named 'torch'
 

Even after installing packages, there are nowhere to be found. conda is less responsive. When I install, it says the requirements are satisfied but it's always looking to some other path.

(torch) u39403@s001-n152:~$ conda install pytorch torchvision cpuonly -c pytorch

Collecting package metadata (current_repodata.json): done
Solving environment: /
Warning: 8 possible package resolutions (only showing differing packages):
  - file:///opt/intel/inteloneapi/conda_channel::libgcc-ng-9.1.0-hdf63c60_0, file:///opt/intel/inteloneapi/conda_channel::libstdcxx-ng-9.1.0-hdf63c60_0, file:///opt/intel/inteloneapi/conda_channel::ninja-1.9.0-py37hfd86e86_0
  - file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::ninja-1.9.0-py37hfd86e86_0, file:///opt/intel/inteloneapi/conda_channel::libgcc-ng-9.1.0-hdf63c60_0, file:///opt/intel/inteloneapi/conda_channel::libstdcxx-ng-9.1.0-hdf63c60_0
  - file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::libgcc-ng-9.1.0-hdf63c60_0, file:///opt/intel/inteloneapi/conda_channel::libstdcxx-ng-9.1.0-hdf63c60_0, file:///opt/intel/inteloneapi/conda_channel::ninja-1.9.0-py37hfd86e86_0
  - file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::libgcc-ng-9.1.0-hdf63c60_0, file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::ninja-1.9.0-py37hfd86e86_0, file:///opt/intel/inteloneapi/conda_channel::libstdcxx-ng-9.1.0-hdf63c60_0
  - file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::libgcc-ng-9.1.0-hdf63c60_0, file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::libstdcxx-ng-9.1.0-hdf63c60_0, file:///opt/intel/inteloneapi/conda_channel::ninja-1.9.0-py37hfd86e86_0
  - file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::libgcc-ng-9.1.0-hdf63c60_0, file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::libstdcxx-ng-9.1.0-hdf63c60_0, file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::ninja-1.9.0-py37hfd86e86_0
  - file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::libstdcxx-ng-9.1.0-hdf63c60_0, file:///opt/intel/inteloneapi/conda_channel::libgcc-ng-9.1.0-hdf63c60_0, file:///opt/intel/inteloneapi/conda_channel::ninja-1.9.0-py37hfd86e86_0
  - file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::libstdcxx-ng-9.1.0-hdf63c60_0, file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::ninja-1.9.0-py37hfd86e86_0, file:///opt/intel/inteloneapi/conda_channel::libgcc-ng-9.1.0-hdf63c60done

# All requested packages already installed.
 

Even after using devcloud for 2 years, my experience so far is not good. 

[Dan_P_Intel]

Hi Prajjwal,

I've just tried it for myself and it seems to work.
Perhaps I followed a different set of steps than you. Could you please tell me if you ran the commands at the terminal provided by Jupyter Lab or at the SSH terminal? Perhaps that makes the difference. This could be a Jupyter config issue.

In my case, I favor SSH over Jupyter so that's what I used.

Here's my list of steps. The full output at the bottom of the message.
It appears that the environments are separate as shown by the python location under conda which appears to be indeed different than the system python. Also, both pytorch and ternsorflow load just fine.

login-2:~
$ qsub -l nodes=1:gpu:ppn=2 -d . -I

s001-n158:~
$ which python
/glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/bin/python

s001-n158:~
$ conda activate pytorch

(pytorch)
s001-n158:~
$ which python
/glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/envs/pytorch/bin/python

(pytorch)
s001-n158:~
$ python
>>> import torch
>>> exit()

(pytorch)
s001-n158:~
$ conda deactivate
(base)
s001-n158:~
$ conda deactivate

s001-n158:~
$ conda activate tensorflow

(tensorflow)
s001-n158:~
$ which python
/glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/envs/tensorflow/bin/python

(tensorflow)
s001-n158:~
$ python
>>> import tensorflow
>>> exit()
(tensorflow)
s001-n158:~
$ conda deactivate

s001-n158:~
$

login-2:~
$ qsub -l nodes=1:gpu:ppn=2 -d . -I
qsub: waiting for job 557510.v-qsvr-1.aidevcloud to start
qsub: job 557510.v-qsvr-1.aidevcloud ready

########################################################################
#      Date:           Thu Mar 26 02:47:06 PDT 2020
#    Job ID:           557510.v-qsvr-1.aidevcloud
#      User:           u25802
# Resources:           neednodes=1:gpu:ppn=2,nodes=1:gpu:ppn=2,walltime=06:00:00
########################################################################

s001-n158:~
$ which python
/glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/bin/python

s001-n158:~
$ conda info --envs
# conda environments:
#
base                  *  /glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest
2021.1-beta05            /glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/envs/2021.1-beta05
pytorch                  /glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/envs/pytorch
pytorch-1.4.0            /glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/envs/pytorch-1.4.0
tensorflow               /glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/envs/tensorflow
tensorflow-1.15.0        /glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/envs/tensorflow-1.15.0

s001-n158:~
$ conda activate pytorch
(pytorch)
s001-n158:~
$ which python
/glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/envs/pytorch/bin/python
(pytorch)
s001-n158:~
$ python
Python 3.7.5 (default, Mar  6 2020, 05:03:49)
[GCC 7.3.0] :: Intel(R) Corporation on linux
Type "help", "copyright", "credits" or "license" for more information.
Intel(R) Distribution for Python is brought to you by Intel Corporation.
Please check out: https://software.intel.com/en-us/python-distibution
>>> import torch
>>> exit()
(pytorch)
s001-n158:~
$ conda deactivate
(base)
s001-n158:~
$ conda deactivate

s001-n158:~
$ conda activate tensorflow
(tensorflow)
s001-n158:~
$ which python
/glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/envs/tensorflow/bin/python
(tensorflow)
s001-n158:~
$ python
Python 3.7.5 (default, Mar  6 2020, 05:03:49)
[GCC 7.3.0] :: Intel(R) Corporation on linux
Type "help", "copyright", "credits" or "license" for more information.
Intel(R) Distribution for Python is brought to you by Intel Corporation.
Please check out: https://software.intel.com/en-us/python-distibution
>>> import tensorflow
>>> exit()
(tensorflow)
s001-n158:~
$ conda deactivate

s001-n158:~
$

[PBhar]

This problem still persists. I'm using only SSH. I tried it just now.

(base) u39403@s001-n101:~$ which python
/glob/intel-python/python3/bin/python
(base) u39403@s001-n101:~$ conda activate torch
-bash: /home/u39403/.conda/envs/torch/etc/conda/activate.d/mpivars.activate.sh: No such file or directory
(torch) u39403@s001-n101:~$ which python
/glob/intel-python/python3/bin/python
 

The path still remains the same. It doesn't change. Every path leads to the same default python path,

[Dan_P_Intel]

The torch environment doesn't exist.

Please use instead:

conda activate pytorch.

[PBhar]

It exists.

base) u39403@s001-n152:~$ conda info --envs
# conda environments:
#
base                  *  /glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest
2021.1-beta05            /glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/envs/2021.1-beta05
pytorch                  /glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/envs/pytorch
pytorch-1.4.0            /glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/envs/pytorch-1.4.0
tensorflow               /glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/envs/tensorflow
tensorflow-1.15.0        /glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/envs/tensorflow-1.15.0
torch                    /home/u39403/.conda/envs/torch

(base) u39403@s001-n152:~$ conda activate pytorch-1.4.0
(pytorch-1.4.0) u39403@s001-n152:~$ python3
Python 3.6.3 |Intel Corporation| (default, Feb 12 2018, 06:37:09)
[GCC 4.8.2 20140120 (Red Hat 4.8.2-15)] on linux
Type "help", "copyright", "credits" or "license" for more information.
Intel(R) Distribution for Python is brought to you by Intel Corporation.
Please check out: https://software.intel.com/en-us/python-distribution
>>> import torch
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ModuleNotFoundError: No module named 'torch'
>>>
(pytorch-1.4.0) u39403@s001-n152:~$ conda activate torch
-bash: /home/u39403/.conda/envs/torch/etc/conda/activate.d/mpivars.activate.sh: No such file or directory
(torch) u39403@s001-n152:~$ python3
Python 3.6.3 |Intel Corporation| (default, Feb 12 2018, 06:37:09)
[GCC 4.8.2 20140120 (Red Hat 4.8.2-15)] on linux
Type "help", "copyright", "credits" or "license" for more information.
Intel(R) Distribution for Python is brought to you by Intel Corporation.
Please check out: https://software.intel.com/en-us/python-distribution
>>> import torch
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ModuleNotFoundError: No module named 'torch'
 

Even after installing packages, there are nowhere to be found. conda is less responsive. When I install, it says the requirements are satisfied but it's always looking to some other path.

(torch) u39403@s001-n152:~$ conda install pytorch torchvision cpuonly -c pytorch

Collecting package metadata (current_repodata.json): done
Solving environment: /
Warning: 8 possible package resolutions (only showing differing packages):
  - file:///opt/intel/inteloneapi/conda_channel::libgcc-ng-9.1.0-hdf63c60_0, file:///opt/intel/inteloneapi/conda_channel::libstdcxx-ng-9.1.0-hdf63c60_0, file:///opt/intel/inteloneapi/conda_channel::ninja-1.9.0-py37hfd86e86_0
  - file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::ninja-1.9.0-py37hfd86e86_0, file:///opt/intel/inteloneapi/conda_channel::libgcc-ng-9.1.0-hdf63c60_0, file:///opt/intel/inteloneapi/conda_channel::libstdcxx-ng-9.1.0-hdf63c60_0
  - file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::libgcc-ng-9.1.0-hdf63c60_0, file:///opt/intel/inteloneapi/conda_channel::libstdcxx-ng-9.1.0-hdf63c60_0, file:///opt/intel/inteloneapi/conda_channel::ninja-1.9.0-py37hfd86e86_0
  - file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::libgcc-ng-9.1.0-hdf63c60_0, file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::ninja-1.9.0-py37hfd86e86_0, file:///opt/intel/inteloneapi/conda_channel::libstdcxx-ng-9.1.0-hdf63c60_0
  - file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::libgcc-ng-9.1.0-hdf63c60_0, file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::libstdcxx-ng-9.1.0-hdf63c60_0, file:///opt/intel/inteloneapi/conda_channel::ninja-1.9.0-py37hfd86e86_0
  - file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::libgcc-ng-9.1.0-hdf63c60_0, file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::libstdcxx-ng-9.1.0-hdf63c60_0, file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::ninja-1.9.0-py37hfd86e86_0
  - file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::libstdcxx-ng-9.1.0-hdf63c60_0, file:///opt/intel/inteloneapi/conda_channel::libgcc-ng-9.1.0-hdf63c60_0, file:///opt/intel/inteloneapi/conda_channel::ninja-1.9.0-py37hfd86e86_0
  - file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::libstdcxx-ng-9.1.0-hdf63c60_0, file:///glob/development-tools/versions/oneapi/beta05/inteloneapi/conda_channel::ninja-1.9.0-py37hfd86e86_0, file:///opt/intel/inteloneapi/conda_channel::libgcc-ng-9.1.0-hdf63c60done

# All requested packages already installed.

[Dan_P_Intel]

Prajjwal, I appreciate your patience. Regarding the torch environment, I can't test that on my side because it is a custom environment, I assume created by you.

I just tried the commands on the same node you tried some time ago (s001-n152) and it appears to work. We just pushed some fixes on the DevCloud and that might have fixed your issue as well. Please try again.

s001-n152:~
$ which python
/glob/intel-python/python3/bin/python

s001-n152:~
$ conda activate pytorch-1.4.0

(pytorch-1.4.0)
s001-n152:~
$ which python
/glob/development-tools/versions/oneapi/beta05/inteloneapi/intelpython/latest/envs/pytorch-1.4.0/bin/python
 

(pytorch-1.4.0)
s001-n152:~
$ python
Python 3.7.5 (default, Mar  6 2020, 05:03:49)
[GCC 7.3.0] :: Intel(R) Corporation on linux
Type "help", "copyright", "credits" or "license" for more information.
Intel(R) Distribution for Python is brought to you by Intel Corporation.
Please check out: https://software.intel.com/en-us/python-distibution
>>> import torch
>>> exit()