1

I am trying to install tensorflow in one of my conda enviroment. and I got a lot of conflicts and eventually failed to install tensorflow.

But I need to keep this enviroment and add tensorflow, is there a way to fix this? Thanks

I do use 64bit python and I use conda install tensorflow. Thank you!


Examining intervaltree:  78%|███████████████████████████████████████████▋            | 181/232 [01:12<00:09,  5.29it/s]\Examining qstylizer:  79%|██████████████████████████████████████████████▊            | 184/232 [01:13<00:18,  2.64it/s]|Examining chardet:  84%|███████████████████████████████████████████████████▎         | 195/232 [01:16<00:15,  2.38it/s]\Examining jupyter_core:  91%|███████████████████████████████████████████████████▏    | 212/232 [01:24<00:14,  1.41it/s]\Examining m2w64-gcc-libs-core:  95%|██████████████████████████████████████████████▍  | 220/232 [01:31<00:08,  1.49it/s]-Examining python-cufflinks:  97%|██████████████████████████████████████████████████▍ | 225/232 [01:32<00:02,  2.70it/s]\                                                                                                                       |Examining conflict for babel jupyterlab-plotly-extension certifi markupsafe mistune pyflakes python-slugify traitlets c-Examining conflict for pandas-datareader babel sphinx jupyterlab_server python-cufflinks matplotlib pytz chart-studio p-Examining conflict for babel jupyterlab-plotly-extension pywin32-ctypes certifi markupsafe mistune pyflakes python-slug|Examining conflict for jupyterlab-plotly-extension jupyterlab_server jupyterlab:   7%| | 16/232 [02:23<27:22,  7.60s/it/Examining conflict for jupyterlab-plotly-extension jupyterlab spyder-kernels python-cufflinks ipykernel ipython jupyter|

after that,


UnsatisfiableError: The following specifications were found to be incompatible with each other:

Output in format: Requested package -> Available versions

Package pyopenssl conflicts for:
requests -> urllib3[version='>=1.21.1,<1.27'] -> pyopenssl[version='>=0.14']
pyopenssl
prometheus_client -> twisted -> pyopenssl[version='>=16.0.0']
urllib3 -> pyopenssl[version='>=0.14']

Package icc_rt conflicts for:
mkl_random -> numpy[version='>=1.16,<2.0a0'] -> icc_rt[version='>=13.1.6|>=2019.0.0|>=16.0.4']
ta -> numpy -> icc_rt[version='>=13.1.6|>=2019.0.0|>=16.0.4']
statsmodels -> scipy[version='>=1.3'] -> icc_rt[version='>=13.1.6|>=2019.0.0|>=16.0.4']
python-cufflinks -> numpy[version='>=1.9.2'] -> icc_rt[version='>=13.1.6|>=2019.0.0|>=16.0.4']
numpy -> icc_rt[version='>=13.1.6|>=2019.0.0|>=16.0.4']
numpy-base -> icc_rt[version='>=13.1.6|>=2019.0.0|>=16.0.4']
matplotlib-base -> numpy[version='>=1.16.6,<2.0a0'] -> icc_rt[version='>=13.1.6|>=2019.0.0|>=16.0.4']
patsy -> numpy[version='>=1.4.0'] -> icc_rt[version='>=13.1.6|>=2019.0.0|>=16.0.4']
matplotlib -> numpy[version='>=1.14.6,<2.0a0'] -> icc_rt[version='>=13.1.6|>=2019.0.0|>=16.0.4']
seaborn -> numpy[version='>=1.15'] -> icc_rt[version='>=13.1.6|>=2019.0.0|>=16.0.4']
pandas -> numpy[version='>=1.16.6,<2.0a0'] -> icc_rt[version='>=13.1.6|>=2019.0.0|>=16.0.4']
scipy -> icc_rt[version='>=13.1.6|>=2019.0.0|>=16.0.4']
roudan
  • 3,082
  • 5
  • 31
  • 72
  • 1
    Take a look at this issue: https://stackoverflow.com/questions/69754812/conda-unsatisfiableerror-even-when-seemingly-there-isnt-any-conflict In your case I also can't see any conflict, conda is not very helpful with the output. Try to make a minimal reproducible environment.yml file – psarka Jan 27 '22 at 08:04
  • 1
    Please create a fresh conda environment and Run "conda install tensorflow " to see the difference. But I suggest to follow instruction from here and make sure your pip version is updated. https://www.tensorflow.org/install/pip –  Feb 16 '22 at 09:10

0 Answers0