The conda environment does not work

[Carlos-Muniz]

I tried using the environment.yaml file on windows 10, but it did not work.
I was receiving multiple issues with the packages.
It seems there are bad dependencies.

Input:
conda env create -f .\environment.yml

Output:

Collecting package metadata (repodata.json): done
Solving environment: failed

ResolvePackageNotFound:
  - freeglut==3.0.0=hf484d3e_5
  - icu==58.2=h9c2bf20_1
  - libpng==1.6.36=hbc83047_0
  - pixman==0.38.0=h7b6447c_0
  - mistune==0.8.4=py36h7b6447c_0
  - readline==7.0=h7b6447c_5
  - pyzmq==18.0.0=py36he6710b0_0
  - h5py==2.8.0=py36h989c5e5_3
  - opencv==3.4.2=py36h6fd60c2_1
  - pyrsistent==0.15.6=py36h7b6447c_0
  - libxcb==1.13=h1bed415_1
  - matplotlib==3.0.3=py36h5429711_0
  - jasper==2.0.14=h07fcdf6_1
  - cython==0.29.6=py36he6710b0_0
  - pandas==0.25.2=py36he6710b0_0
  - jpeg==9b=h024ee3a_2
  - tensorflow==1.12.0=gpu_py36he74679b_0
  - tensorboard==1.12.2=py36he6710b0_0
  - cudnn==7.3.1=cuda9.2_0
  - glib==2.56.2=hd408876_0
  - ncurses==6.1=he6710b0_1
  - dbus==1.13.6=h746ee38_0
  - graphite2==1.3.13=h23475e2_0
  - numpy==1.16.2=py36h7e9f1db_0
  - opencv3==3.1.0=py36_0
  - zstd==1.3.7=h0b5b093_0
  - libtiff==4.0.10=h2733197_2
  - openssl==1.1.1d=h7b6447c_3
  - scikit-learn==0.20.3=py36hd81dba3_0
  - libedit==3.1.20181209=hc058e9b_0
  - mkl_random==1.0.2=py36hd81dba3_0
  - numpy-base==1.16.2=py36hde5b4d6_0
  - hdf5==1.10.2=hba1933b_1
  - gst-plugins-base==1.14.0=hbbd80ab_1
  - expat==2.2.6=he6710b0_0
  - libopus==1.3=h7b6447c_0
  - cupti==9.2.148=0
  - libvpx==1.7.0=h439df22_0
  - py-opencv==3.4.2=py36hb342d67_1
  - zeromq==4.3.1=he6710b0_3
  - python==3.6.8=h0371630_0
  - tornado==5.1.1=py36h7b6447c_0
  - pyqt==5.9.2=py36h05f1152_2
  - libopencv==3.4.2=hb342d67_1
  - kiwisolver==1.0.1=py36hf484d3e_0
  - yaml==0.1.7=had09818_2
  - libglu==9.0.0=hf484d3e_1
  - libxml2==2.9.9=he19cac6_0
  - pillow==6.2.1=py36h34e0f95_0
  - ptyprocess==0.6.0=py36_0
  - libsodium==1.0.16=h1bed415_0
  - markupsafe==1.1.1=py36h7b6447c_0
  - libgfortran-ng==7.3.0=hdf63c60_0
  - cairo==1.14.12=h8948797_3
  - pyyaml==3.13=py36h14c3975_0
  - c-ares==1.15.0=h7b6447c_1
  - cudatoolkit==9.2=0
  - libffi==3.2.1=hd88cf55_4
  - gmp==6.1.2=h6c8ec71_1
  - libuuid==1.0.3=h1bed415_2
  - xz==5.2.4=h14c3975_4
  - tensorflow-base==1.12.0=gpu_py36had579c0_0
  - libstdcxx-ng==8.2.0=hdf63c60_1
  - tk==8.6.8=hbc83047_0
  - sip==4.19.8=py36hf484d3e_0
  - libprotobuf==3.6.1=hd408876_0
  - scipy==1.2.1=py36h7c811a0_0
  - mkl_fft==1.0.10=py36ha843d7b_0
  - zlib==1.2.11=h7b6447c_3
  - cryptography==2.8=py36h1ba5d50_0
  - pcre==8.43=he6710b0_0
  - ffmpeg==4.0=hcdf2ecd_0
  - qt==5.9.7=h5867ecd_1
  - bzip2==1.0.6=h14c3975_5
  - sqlite==3.26.0=h7b6447c_0
  - cffi==1.13.2=py36h2e261b9_0
  - harfbuzz==1.8.8=hffaf4a1_0
  - grpcio==1.16.1=py36hf8bcb03_1
  - fontconfig==2.13.0=h9420a91_0
  - libgcc-ng==8.2.0=hdf63c60_1
  - gstreamer==1.14.0=hb453b48_1
  - freetype==2.9.1=h8a8886c_1

[bothe]

Thank you @Carlos-Muniz for trying the code, I mostly use Ubuntu where it works, I see that there is PyTorch in /conda_env/environment.yml perhaps that's a problem on Windows. On the other hand, I kept this script to remind the libraries and version I used, you may not need all of them and try running the code and install only the required ones. I hope this helps.