I'm trying to set up tensorflow-gpu on my local machine to train neural networks on my RTX 2070 super. Unfortunately, I get the issue of the environment failing to solve, even in a brand new environment.
I originally tried to install tensorflow-gpu directly through Anaconda, but even with cudnn and cudatoolkit installed, it wouldn't recognize the GPU. I think it may be time to jump ship for pytorch...
Error codes;
(base) C:\Users\Albert>conda install -c conda-forge cudnn
Collecting package metadata (current_repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Solving environment: failed with repodata from current_repodata.json, will retry with next repodata source.
Collecting package metadata (repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
The packages you are trying to install may be conflicting with your already installed packages. Try to create a new environment for tensorflow with a minimal set of packages:
conda create --name tf python cudnn -c conda-forge
conda activate tf
You can then install other packages with either conda or pip in this new environment.
In general, I find that creating separate conda environments for different tasks or projects is a better way to use conda
The conda documentation has details about managing environments
It's might occour because of the version conflict, you can just uninstall the python version in the existing conda environment or simply create new Conda environment.
To Change the python version-latest on existing environment:
conda uninstall python
conda install python -y
To create a new conda env
conda create -n <env-name> python -y
conda activate <env-name>
To install conda use any of the following commands:
conda install -c conda-forge tensorflow
conda install -c conda-forge/label/broken tensorflow
conda install -c conda-forge/label/cf201901 tensorflow
conda install -c conda-forge/label/cf202003 tensorflow
reference: https://anaconda.org/conda-forge/tensorflow
Try cleaning conda cache-files in case of error
Related
I have installed Anaconda from scratch and then installed pyomo, which worked.
When I try to install pyomo.extra, it runs about 1/2 hour and then fails.
I had tried:
conda install -c conda-forge pyomo.extras
conda install -c "conda-forge/label/cf201901" pyomo.extras
conda install -c "conda-forge/label/cf202003" pyomo.extras
as recommended on https://anaconda.org/conda-forge/pyomo.extras
and also
conda config --add channels conda-forge
conda install pyomo.extras
subsequently checking with
conda search pyomo.extras --channel conda-forge
as recommended on https://github.com/conda-forge/pyomo.extras-feedstock
indicated that pyomo.extra isn't installed.
here is the generated output.
(base) PS C:\WINDOWS\system32> conda install pyomo.extras
Collecting package metadata (current_repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Solving environment: failed with repodata from current_repodata.json, will retry with next repodata source.
Collecting package metadata (repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Solving environment: -
Found conflicts! Looking for incompatible packages.
This can take several minutes. Press CTRL-C to abort.
Examining conflict for pyomo.extras: 75%|███████████████████████████████████▎ | 3/4 [00:03<00:01, 1.15s/it]\failed
UnsatisfiableError: The following specifications were found
to be incompatible with the existing python installation in your environment:
Specifications:
- pyomo.extras -> python[version='2.7.*|3.5.*|3.6.*|>=2.7,<2.8.0a0|>=3.6,<3.7.0a0|>=3.7,<3.8.0a0|3.4.*']
Your python: python=3.9
If python is on the left-most side of the chain, that's the version you've asked for.
When python appears to the right, that indicates that the thing on the left is somehow
not available for the python version you are constrained to. Note that conda will not
change your python version to a different minor version unless you explicitly specify
that.
The following specifications were found to be incompatible with your system:
- feature:/win-64::__win==0=0
- feature:|#/win-64::__win==0=0
- pyomo.extras -> ipython -> __osx
- pyomo.extras -> ipython -> __win
Your installed version is: 0
(base) PS C:\WINDOWS\system32> conda search pyomo.extras --channel conda-forge
Loading channels: -
(base) PS C:\WINDOWS\system32>
Questions:
what might be the reason why installation of pyomo.extra fails?
what alternatives for installing pyomo.extra can be recommende?
The package is only available for Python 2.7, 3.6, and 3.7, but your target environment currently has Python 3.9. Conda will not change the Python version in-place because it entails changing pretty much every package in the environment. Instead, try creating a new environment with a compatible Python version.
## include any other packages now
conda create --name foo python=3.7 pyomo pyomo.extras
Why only old versions? The feedstock repository has been failing to build updated versions since Python 3.8 and no maintainers have stepped in to fix it. If you would like to see updated versions built, then drop an Issue on the feedstock or, if you feel comfortable trying to fix it, send a Pull Request. The squeaky wheel get's the grease!
I think I have an environment problem with miniconda. Here is what I succeded(o)/mismanaged(x) to install python mayavi package.
The documentation writes to type :
(o) conda install vtk
(x) conda install pyqt=4
<(x) conda install mayavi> (not performed)
The output message for another cmd (conda install -c anaconda -c conda-forge mayavi) are:
Collecting package metadata (current_repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Solving environment: /
I understand that I must check the first failed package: pyqt. There are similar messages, before the shell windows simply breaks. I have read that an updated environment could work but my OS (fedora35 as virtual machine) and python 3.9 + miniconda3 are fast from best versions. Any suggestion or help to fix this or trying docker are welcome
I am getting the following error output when trying to install eomaps using Conda. I don't know how to solve this. I have also tried the same using pip but it didn't seem to solve the problem. Here is the error that Conda reports:
(base) C:\Windows\system32>conda install -c conda-forge eomaps
Collecting package metadata (current_repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Solving environment: failed with repodata from current_repodata.json, will retry with next repodata source.
Collecting package metadata (repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Solving environment: -
Found conflicts! Looking for incompatible packages.
This can take several minutes. Press CTRL-C to abort.
Here is the error that pip reports:
ERROR: Failed building wheel for cartopy
ERROR: Command errored out with exit status 1:
------
ERROR: Failed building wheel for fiona
ERROR: Could not build wheels for cartopy which use PEP 517 and cannot be installed directly
How can I get this to install?
EOmaps dev here...
I think the problem is that you're installing packages in your (base) environment! You most probably have some packages installed (previous versions of cartopy, pyproj etc.?) that conflict with the requirements of EOmaps which is why the conda-solver does not succeed.
I'd strongly suggest working in a fresh environment and NOT in the base environment!
(Checkout conda user-guide - managing environments)
As an example, the following should work without any problems:
conda create -n eomaps_env python=3.9
conda activate eomaps_env
conda install -c conda-forge eomaps
(the name eomaps_env and the python-version are both optional and you can change it to whatever name and version you like!)
To speed up solving complex dependencies you can also try to use mamba (a c++ reimplementation of the conda package manager) which does the very same thing as conda (but solving dependencies works a lot faster):
conda create -n eomaps_env python=3.9
conda activate eomaps_env
conda install -c conda-forge mamba
mamba install -c conda-forge eomaps
My Conda Environments are not showing up in my Jupiter Notebooks. I've reviewed all the (excellent) dialogue here, but it doesn't appear to work for python 3.9.2.
I tried the following, which didn't work:
conda install jupyter
conda install ipykernel
conda install nb_conda
Then when I try the following conda install nb_conda_kernels I get the following error:
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Solving environment: failed with repodata from current_repodata.json, will retry with next repodata source.
Collecting package metadata (repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Solving environment: |
Found conflicts! Looking for incompatible packages.
This can take several minutes. Press CTRL-C to abort.
failed
UnsatisfiableError: The following specifications were found
to be incompatible with the existing python installation in your environment:
Specifications:
- nb_conda_kernels -> python[version='>=2.7,<2.8.0a0|>=3.6,<3.7.0a0|>=3.7,<3.8.0a0|>=3.8,<3.9.0a0']
Your python: python=3.9
If python is on the left-most side of the chain, that's the version you've asked for.
When python appears to the right, that indicates that the thing on the left is somehow
not available for the python version you are constrained to. Note that conda will not
change your python version to a different minor version unless you explicitly specify
that.
How can I get my conda environments to show in jupyter notebooks in python 3.9.2? Thank you!
Try using the conda-forge channel, e.g.:
$ conda install -c conda-forge ipykernel
And you don't need to install jupyter if this conda env is just a kernel to add, if that's what you want to do.
After that, you need to run this in that conda env:
$ python -m ipykernel install --user --name vanilla --display-name "Python 3 (vanilla)"
Replace the name and display-name to your liking.
After installing Anaconda, at first, I installed tensorflow with conda, then I tried to install opencv with the following conda commands:
conda install -c conda-forge opencv
conda install -c conda-forge/label/gcc7 opencv
conda install -c conda-forge/label/broken opencv
conda install -c conda-forge/label/cf201901 opencv
conda install -c conda-forge/label/cf202003 opencv
But, none of the avobe worked, it failed to install opencv on my base (root) environment. (However, I became able to install it on a different environment.)
The error message I got is:
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Solving environment: failed with repodata from current_repodata.json, will retry with next repodata source.
I also tried to install opencv with Anaconda Navigator on base (root) environment. But, it didn't work also.
Then, I tried to install opencv with pip with the following commmand:
pip install opencv-python
But, it also didn't work.
Now, How can I install opencv on my base (root) environment?
Your (partial) error message (from the comments):
Solving environment: failed with initial frozen solve. Retrying with flexible solve. Solving environment: failed with repodata from current_repodata.json, will retry with next repodata source.
and the fact that you were able to install it into a different environment:
However, I became able to install it on a different environment.
Indicate that you have conflicts with other packages in your base environment, i.e. conda was not able to determine how to fit opencv and its dependencies together with other packages in your base environment.
The best practice here would be to set up a new environment and only install the packages that you need for a certain project.
Alternatively, you could also uninstall anaconda and install miniconda (which is the same as aanconda, but without the preinstalled packages) and then install into base only what you need.