0

I try to create conda env with the following environment.yml file

name: myenv
channels:
  - rapidsai
  - conda-forge
  - nvidia
dependencies:
  - python=3.10
  - cudf=23.04
  - cuml=23.04
  - cugraph=23.04
  - cuspatial=23.04
  - cuxfilter=23.04
  - cucim=23.04
  - cusignal=23.04
  - cudatoolkit=11.2
  - tensorflow-gpu

The above version is copied from RAPIDS (except tensorflow-gpu)

Then create a docker image with the following Dockerfile

FROM continuumio/miniconda3:latest
RUN conda update conda 

COPY environment.yml .
RUN conda env create 
RUN conda init bash
RUN echo "source activate myenv" >> ~/.bashrc

While it is building, I got the following error.

...
                                                                                                                       0.1s
 => ERROR [ 4/18] RUN conda env create                                                                                                                                                                            1061.7s
------
 > [ 4/18] RUN conda env create:
#0 0.686 Retrieving notices: ...working... done
#0 0.744 Collecting package metadata (repodata.json): ...working... done
#0 97.73 Solving environment: ...working...
#0 371.2 Found conflicts! Looking for incompatible packages.
#0 371.2 This can take several minutes.  Press CTRL-C to abort.
failed
#0 426.7 Solving environment: ...working...
#0 982.2 Found conflicts! Looking for incompatible packages.
#0 982.2 This can take several minutes.  Press CTRL-C to abort.
failed
#0 1056.0
#0 1056.0 UnsatisfiableError: The following specifications were found to be incompatible with each other:
#0 1056.0
#0 1056.0 Output in format: Requested package -> Available versions
#0 1056.0
#0 1056.0 Package libstdcxx-ng conflicts for:
#0 1056.0 cudf=23.04 -> libstdcxx-ng[version='>=12']
#0 1056.0 cugraph=23.04 -> cuda-python[version='>=11.7.1,<12.0'] -> libstdcxx-ng[version='>=10.3.0|>=11.2.0|>=9.4.0|>=9.3.0|>=7.3.0|>=7.5.0']
#0 1056.0 cuml=23.04 -> libstdcxx-ng[version='>=12']
#0 1056.0 tensorflow-gpu -> tensorflow==2.6.2=cuda111py37hf54207c_2 -> libstdcxx-ng[version='>=5.4.0|>=7.5.0|>=9.4.0|>=7.2.0']
#0 1056.0 cuml=23.04 -> cuda-python[version='>=11.7.1,<12'] -> libstdcxx-ng[version='>=10.3.0|>=11.2.0|>=9.4.0|>=9.3.0|>=7.3.0|>=7.5.0']
#0 1056.0 cucim=23.04 -> libstdcxx-ng[version='>=12']
#0 1056.0 cugraph=23.04 -> libstdcxx-ng[version='>=12']
#0 1056.0 cudatoolkit=11.2 -> libstdcxx-ng[version='>=10.3.0|>=12|>=9.4.0|>=9.3.0']
#0 1056.0 cusignal=23.04 -> cupy[version='>=10,<12.0.0a0'] -> libstdcxx-ng[version='>=10.3.0|>=12|>=7.5.0|>=9.4.0|>=9.3.0|>=11.2.0|>=7.3.0|>=4.9|>=7.2.0']
#0 1056.0 python=3.10 -> libffi[version='>=3.4,<4.0a0'] -> libstdcxx-ng[version='>=11.2.0|>=9.4.0|>=7.5.0|>=7.3.0']
#0 1056.0 cucim=23.04 -> cudatoolkit[version='>=11,<12.0a0'] -> libstdcxx-ng[version='>=10.3.0|>=9.4.0|>=9.3.0|>=7.3.0|>=7.5.0|>=11.2.0|>=4.9|>=7.2.0']
#0 1056.0 cuspatial=23.04 -> libstdcxx-ng[version='>=12']
#0 1056.0 cuspatial=23.04 -> python[version='>=3.8,<3.9.0a0'] -> libstdcxx-ng[version='>=11.2.0|>=7.3.0|>=7.5.0|>=9.3.0|>=9.4.0']
#0 1056.0 cuxfilter=23.04 -> cudf=23.04 -> libstdcxx-ng[version='>=10.3.0|>=12|>=7.5.0|>=9.4.0|>=9.3.0|>=7.3.0|>=11.2.0']
#0 1056.0 cudf=23.04 -> cuda-python[version='>=11.7.1,<12.0'] -> libstdcxx-ng[version='>=10.3.0|>=11.2.0|>=9.4.0|>=9.3.0|>=7.3.0|>=7.5.0']
#0 1056.0
#0 1056.0 Package numba conflicts for:
#0 1056.0 cuxfilter=23.04 -> numba[version='>=0.54']
#0 1056.0 cuml=23.04 -> cudf=23.04 -> numba[version='>=0.56.4,<0.57']
#0 1056.0 cusignal=23.04 -> numba[version='>=0.49.0']
#0 1056.0 cudf=23.04 -> numba[version='>=0.56.4,<0.57']
#0 1056.0 cudf=23.04 -> rmm[version='>=23.4.1,<23.5.0a0'] -> numba[version='>=0.49']
#0 1056.0 cugraph=23.04 -> cudf=23.04 -> numba[version='>=0.54|>=0.56.4,<0.57']
#0 1056.0 cuxfilter=23.04 -> cudf=23.04 -> numba[version='>=0.51|>=0.51.0|>=0.56.4,<0.57']
#0 1056.0 cuspatial=23.04 -> cudf=23.04 -> numba[version='>=0.49|>=0.56.4,<0.57']
#0 1056.0
#0 1056.0 Package matplotlib-base conflicts for:
#0 1056.0 cuspatial=23.04 -> geopandas[version='>=0.11.0'] -> matplotlib-base[version='>=3.2.0']
#0 1056.0 cuxfilter=23.04 -> geopandas[version='>=0.11.0'] -> matplotlib-base[version='>=2.2|>=3.0|>=3.2.0|>=3']
#0 1056.0 cuml=23.04 -> seaborn -> matplotlib-base[version='>=2.1.2|>=3.1,!=3.6.1']
#0 1056.0
#0 1056.0 Package libraft conflicts for:
#0 1056.0 cugraph=23.04 -> pylibraft=23.04 -> libraft[version='23.04.00.*|23.04.01.*']
#0 1056.0 cugraph=23.04 -> libraft=23.04
#0 1056.0
#0 1056.0 Package python conflicts for:
#0 1056.0 cuspatial=23.04 -> geopandas[version='>=0.11.0'] -> python[version='3.10.*|3.8.*|>=3.8|>=3.9,<3.10.0a0']
#0 1056.0 python=3.10
#0 1056.0 cusignal=23.04 -> cupy[version='>=10,<12.0.0a0'] -> python[version='>=3.10,<3.11.0a0|>=3.11,<3.12.0a0|>=3.8,<3.9.0a0|>=3.9,<3.10.0a0|>=3.7,<3.8.0a0|>=3.6,<3.7.0a0']
#0 1056.0 cuxfilter=23.04 -> bokeh[version='>=2.4.2,<=2.5'] -> python[version='2.7.*|3.4.*|3.5.*|3.6.*|>=2.7|>=3.11,<3.12.0a0|>=3.7,<3.8.0a0|>=3.7|>=3.9,<3.10.0a0|>=3.8|>=3.6|>=3|>=2.7,<2.8.0a0|>=3.6,<3.7.0a0|>=3.5,<3.
6.0a0|3.10.*|3.8.*']
#0 1056.0 cudf=23.04 -> cachetools -> python[version='2.7.*|3.4.*|3.5.*|3.6.*|>=3.5,<3.6|>=3.5|>=3.7|>=3.9,<3.10.0a0|>=3.7,<3.8.0a0|>=3.11,<3.12.0a0|>=3.6,<3.7.0a0|>=2.7,<2.8.0a0|>=3.8|>=3.6|>=2.7|>=3.5,<3.6.0a0|>=3.7.
1,<3.8.0a0|3.10.*|3.8.*']
#0 1056.0 cudf=23.04 -> python[version='>=3.10,<3.11.0a0|>=3.8,<3.9.0a0']
#0 1056.0 cugraph=23.04 -> python[version='>=3.10,<3.11.0a0|>=3.8,<3.9.0a0']
#0 1056.0 cucim=23.04 -> click -> python[version='2.7.*|3.5.*|3.6.*|>=3.11,<3.12.0a0|>=3.8|>=3.9,<3.10.0a0|>=3.7,<3.8.0a0|>=3.6,<3.7.0a0|3.4.*|>=3|>=3.6|>=2.7,<2.8.0a0|>=3.5,<3.6.0a0|3.10.*|3.8.*']
#0 1056.0 cuml=23.04 -> cuda-python[version='>=11.7.1,<12'] -> python[version='2.7.*|3.5.*|3.6.*|>=3.11,<3.12.0a0|>=3.9,<3.10.0a0|>=3.7,<3.8.0a0|>=3.6,<3.7.0a0|>=3.8|>=3.6|>=3.7|>=2.7,<2.8.0a0|>=3.5,<3.6.0a0|3.10.*|3.4
.*|3.8.*']
#0 1056.0 cuxfilter=23.04 -> python[version='>=3.10,<3.11.0a0|>=3.8,<3.9.0a0']
#0 1056.0 cusignal=23.04 -> python
#0 1056.0 cugraph=23.04 -> cuda-python[version='>=11.7.1,<12.0'] -> python[version='3.10.*|>=3.11,<3.12.0a0|>=3.9,<3.10.0a0|>=3.7,<3.8.0a0|>=3.8|3.8.*']
#0 1056.0 cucim=23.04 -> python[version='>=3.10,<3.11.0a0|>=3.8,<3.9.0a0']
#0 1056.0 cuspatial=23.04 -> python[version='>=3.10,<3.11.0a0|>=3.8,<3.9.0a0']
#0 1056.0 tensorflow-gpu -> tensorflow==2.11.1=cuda112py39h01bd6f0_0 -> python[version='2.7.*|3.6.*|3.7.*|>=3.10,<3.11.0a0|>=3.9,<3.10.0a0|>=3.8,<3.9.0a0|>=3.7,<3.8.0a0|>=3.6,<3.7.0a0|3.8.*|3.9.*|>=2.7,<2.8.0a0|>=3.5,<
3.6.0a0|3.5.*']
#0 1056.0 cuml=23.04 -> python[version='>=3.10,<3.11.0a0|>=3.8,<3.9.0a0']
#0 1056.0
#0 1056.0 Package dask conflicts for:
#0 1056.0 cugraph=23.04 -> dask==2023.3.2
#0 1056.0 cuml=23.04 -> dask==2023.3.2
#0 1056.0 cuxfilter=23.04 -> dask-cuda=23.04 -> dask[version='2023.3.2|>=0.18.0']
#0 1056.0
#0 1056.0 Package pynvml conflicts for:
#0 1056.0 cuxfilter=23.04 -> dask-cuda=23.04 -> pynvml[version='>=11.0.0,<11.5']
#0 1056.0 cugraph=23.04 -> dask-cuda=23.04 -> pynvml[version='>=11.0.0,<11.5|>=11.4.1']
#0 1056.0
#0 1056.0 Package librmm conflicts for:
#0 1056.0 cudf=23.04 -> libcudf=23.04.01 -> librmm[version='23.04.*|>=23.4.1,<23.5.0a0|>=23.4.0,<23.5.0a0']
#0 1056.0 cuspatial=23.04 -> libcuspatial[version='>=23.4.0,<23.5.0a0'] -> librmm[version='23.04.*|>=23.4.1,<23.5.0a0|>=23.4.0,<23.5.0a0']
#0 1056.0 cugraph=23.04 -> libcugraph=23.04.01 -> librmm=23.04
#0 1056.0


... (skipped similar error)


#0 1056.0
#0 1056.0 Package libblas conflicts for:
#0 1056.0 cucim=23.04 -> numpy[version='>=1.21.6,<2.0a0'] -> libblas[version='>=3.8.0,<4.0a0|>=3.9.0,<4.0a0']
#0 1056.0 cudf=23.04 -> numpy[version='>=1.21'] -> libblas[version='>=3.8.0,<4.0a0|>=3.9.0,<4.0a0']
#0 1056.0 cusignal=23.04 -> numpy[version='>=1.19,<1.24'] -> libblas[version='>=3.8.0,<4.0a0|>=3.9.0,<4.0a0']
#0 1056.0
#0 1056.0 Package cudatoolkit conflicts for:
#0 1056.0 cudf=23.04 -> cudatoolkit[version='>=11,<12.0a0']
#0 1056.0 cugraph=23.04 -> cuda-python[version='>=11.7.1,<12.0'] -> cudatoolkit[version='10.0|10.0.*|10.1|10.1.*|10.2|10.2.*|11.0|11.0.*|>=11.2,<12|11.1|11.1.*|9.2|9.2.*']
#0 1056.0 tensorflow-gpu -> tensorflow==2.6.2=cuda111py37hf54207c_2 -> cudatoolkit[version='10.2|10.2.*|11.0|11.0.*|11.1|11.1.*|>=11.2,<12|9.0.*|8.0.*']
#0 1056.0 cuml=23.04 -> cudatoolkit[version='>=11,<12.0a0']
#0 1056.0 cusignal=23.04 -> cupy[version='>=10,<12.0.0a0'] -> cudatoolkit[version='10.2|10.2.*|11.0|11.0.*|>=11.2,<12|11.1|11.1.*']
#0 1056.0 cuxfilter=23.04 -> cudf=23.04 -> cudatoolkit[version='10.0|10.0.*|10.1|10.1.*|10.2|10.2.*|11.0|11.0.*|>=11,<12.0a0|>=11.2,<12|11.1|11.1.*|9.2|9.2.*']
#0 1056.0 cudf=23.04 -> cuda-python[version='>=11.7.1,<12.0'] -> cudatoolkit[version='10.0|10.0.*|10.1|10.1.*|10.2|10.2.*|11.0|11.0.*|>=11.2,<12|11.1|11.1.*|9.2|9.2.*|>=11.0,<=11.8|>=11.0,<=11.7']
#0 1056.0 cuml=23.04 -> cuda-python[version='>=11.7.1,<12'] -> cudatoolkit[version='10.0|10.0.*|10.1|10.1.*|10.2|10.2.*|11.0|11.0.*|11.4|11.4.*|>=11.2,<12|11.1|11.1.*|9.2|9.2.*|>=11.2,<12.0a0']
#0 1056.0 cugraph=23.04 -> cudatoolkit[version='>=11,<12.0a0']
#0 1056.0 cucim=23.04 -> cupy[version='>=10,<12.0.0a0'] -> cudatoolkit[version='10.2|10.2.*|11.0|11.0.*|>=11.2,<12|11.1|11.1.*']
#0 1056.0 cudatoolkit=11.2
#0 1056.0 cucim=23.04 -> cudatoolkit[version='>=11,<12.0a0']
#0 1056.0 cuspatial=23.04 -> cudf=23.04 -> cudatoolkit[version='>=11,<12.0a0']
#0 1056.0
#0 1056.0 Package importlib_metadata conflicts for:
#0 1056.0 cugraph=23.04 -> dask-core==2023.3.2 -> importlib_metadata[version='>=4.13.0']
#0 1056.0 cudf=23.04 -> numba[version='>=0.56.4,<0.57'] -> importlib_metadata
#0 1056.0 cuml=23.04 -> dask-core==2023.3.2 -> importlib_metadata[version='>=4.13.0']
#0 1056.0 cusignal=23.04 -> numba[version='>=0.49.0'] -> importlib_metadata
#0 1056.0 cuxfilter=23.04 -> numba[version='>=0.54'] -> importlib_metadata
#0 1056.0
#0 1056.0 Package toolz conflicts for:
#0 1056.0 cucim=23.04 -> scikit-image[version='>=0.19.0,<0.21.0a0'] -> toolz[version='>=0.7.3']
#0 1056.0 cuml=23.04 -> dask-core==2023.3.2 -> toolz[version='>=0.10.0|>=0.8.2']
#0 1056.0 cugraph=23.04 -> dask-core==2023.3.2 -> toolz[version='>=0.10.0|>=0.8.2']
#0 1056.0 cuxfilter=23.04 -> datashader[version='>=0.14,<=0.14.4'] -> toolz[version='>=0.7.4']
#0 1056.0
#0 1056.0 Package python_abi conflicts for:
#0 1056.0 cuxfilter=23.04 -> bokeh[version='>=2.4.2,<=2.5'] -> python_abi[version='3.11.*|3.7|3.7.*|3.8|3.9.*|3.9|3.6.*|3.8.*',build='*_pypy38_pp73|*_cp39|*_cp37m|*_pypy39_pp73|*_pypy37_pp73|*_cp311|*_cp36m|*_graalpy22
3_38_native']
#0 1056.0 cuxfilter=23.04 -> python_abi[version='3.10.*|3.8.*',build='*_cp310|*_cp38']
#0 1056.0
#0 1056.0 Package libpng conflicts for:
#0 1056.0 tensorflow-gpu -> tensorflow==1.14.0 -> libpng[version='>=1.6.37,<1.7.0a0']
#0 1056.0 cuxfilter=23.04 -> libwebp -> libpng[version='>=1.6.32,<1.7.0a0|>=1.6.34,<1.7.0a0|>=1.6.35,<1.7.0a0|>=1.6.37,<1.7.0a0|>=1.6.39,<1.7.0a0']
#0 1056.0
#0 1056.0 Package tbb conflicts for:
#0 1056.0 cusignal=23.04 -> numba[version='>=0.49.0'] -> tbb[version='>=2020.0|>=2020.2|>=2020.3|>=2021.4.0|>=2021.5.0']
#0 1056.0 cudf=23.04 -> numba[version='>=0.56.4,<0.57'] -> tbb[version='>=2021.5.0']
#0 1056.0 cuxfilter=23.04 -> numba[version='>=0.54'] -> tbb[version='>=2021.4.0|>=2021.5.0']
#0 1056.0
#0 1056.0 Package protobuf conflicts for:
#0 1056.0 cuxfilter=23.04 -> cudf=23.04 -> protobuf[version='>=4.21.6,<4.22']
#0 1056.0 cuspatial=23.04 -> cudf=23.04 -> protobuf[version='>=4.21.6,<4.22']
#0 1056.0 cugraph=23.04 -> cudf=23.04 -> protobuf[version='>=4.21.6,<4.22']
#0 1056.0 cudf=23.04 -> protobuf[version='>=4.21.6,<4.22']
#0 1056.0 cuml=23.04 -> cudf=23.04 -> protobuf[version='>=4.21.6,<4.22']
#0 1056.0 tensorflow-gpu -> tensorflow==1.13.1 -> protobuf[version='>=3.3.0|>=3.4.0|>=3.6.0|>=3.6.1']
#0 1056.0
#0 1056.0 Package pypy3.6 conflicts for:
#0 1056.0 cudf=23.04 -> fastavro[version='>=0.22.0'] -> pypy3.6[version='>=7.3.1|>=7.3.2|>=7.3.3']
#0 1056.0 cucim=23.04 -> click -> pypy3.6[version='>=7.3.1|>=7.3.2|>=7.3.3']
#0 1056.0 cusignal=23.04 -> numpy[version='>=1.19,<1.24'] -> pypy3.6[version='7.3.0.*|7.3.1.*|7.3.2.*|7.3.3.*|>=7.3.1|>=7.3.2|>=7.3.3']
#0 1056.0 cuml=23.04 -> cupy[version='>=7.8.0,<12.0.0a0'] -> pypy3.6[version='>=7.3.1|>=7.3.2|>=7.3.3']
#0 1056.0
#0 1056.0 Package cuspatial conflicts for:
#0 1056.0 cuspatial=23.04
#0 1056.0 cuxfilter=23.04 -> cuspatial=23.04
#0 1056.0
#0 1056.0 Package xz conflicts for:
#0 1056.0 cuxfilter=23.04 -> python[version='>=3.10,<3.11.0a0'] -> xz[version='>=5.2.10,<6.0a0|>=5.2.5,<5.3.0a0|>=5.2.6,<5.3.0a0|>=5.2.6,<6.0a0|>=5.2.8,<6.0a0|>=5.2.5,<6.0a0|>=5.2.4,<5.3.0a0|>=5.2.4,<6.0a0']
#0 1056.0 cudf=23.04 -> python[version='>=3.10,<3.11.0a0'] -> xz[version='>=5.2.10,<6.0a0|>=5.2.5,<5.3.0a0|>=5.2.6,<5.3.0a0|>=5.2.6,<6.0a0|>=5.2.8,<6.0a0|>=5.2.5,<6.0a0|>=5.2.4,<5.3.0a0|>=5.2.4,<6.0a0']
#0 1056.0 cucim=23.04 -> libcucim=23.04.01 -> xz[version='>=5.2.10,<6.0a0|>=5.2.5,<5.3.0a0|>=5.2.6,<5.3.0a0|>=5.2.6,<6.0a0|>=5.2.8,<6.0a0|>=5.2.5,<6.0a0|>=5.2.4,<5.3.0a0|>=5.2.4,<6.0a0']
#0 1056.0 cuspatial=23.04 -> python[version='>=3.8,<3.9.0a0'] -> xz[version='>=5.2.10,<6.0a0|>=5.2.4,<5.3.0a0|>=5.2.5,<5.3.0a0|>=5.2.6,<6.0a0|>=5.2.5,<6.0a0|>=5.2.4,<6.0a0|>=5.2.6,<5.3.0a0|>=5.2.8,<6.0a0']
#0 1056.0 python=3.10 -> xz[version='>=5.2.10,<6.0a0|>=5.2.5,<5.3.0a0|>=5.2.6,<5.3.0a0|>=5.2.6,<6.0a0|>=5.2.8,<6.0a0|>=5.2.5,<6.0a0']
#0 1056.0 cusignal=23.04 -> python -> xz[version='5.0.*|5.2.*|>=5.2.3,<5.3.0a0|>=5.2.4,<5.3.0a0|>=5.2.5,<5.3.0a0|>=5.2.6,<5.3.0a0|>=5.2.6,<6.0a0|>=5.2.10,<6.0a0|>=5.2.8,<6.0a0|>=5.2.5,<6.0a0|>=5.2.4,<6.0a0|>=5.2.3,<6.0
a0']
#0 1056.0 cugraph=23.04 -> python[version='>=3.10,<3.11.0a0'] -> xz[version='>=5.2.10,<6.0a0|>=5.2.5,<5.3.0a0|>=5.2.6,<5.3.0a0|>=5.2.6,<6.0a0|>=5.2.8,<6.0a0|>=5.2.5,<6.0a0|>=5.2.4,<5.3.0a0|>=5.2.4,<6.0a0']
#0 1056.0 cuml=23.04 -> python[version='>=3.10,<3.11.0a0'] -> xz[version='>=5.2.10,<6.0a0|>=5.2.5,<5.3.0a0|>=5.2.6,<5.3.0a0|>=5.2.6,<6.0a0|>=5.2.8,<6.0a0|>=5.2.5,<6.0a0|>=5.2.4,<5.3.0a0|>=5.2.4,<6.0a0']
#0 1056.0
#0 1056.0 Package mkl conflicts for:
#0 1056.0 tensorflow-gpu -> tensorflow-gpu-base==1.7.0 -> mkl[version='>=2018.0.2,<2019.0a0']
#0 1056.0 cudf=23.04 -> numpy[version='>=1.21'] -> mkl[version='>=2021.3.0,<2022.0a0|>=2021.4.0,<2022.0a0']
#0 1056.0 cusignal=23.04 -> numpy[version='>=1.19,<1.24'] -> mkl[version='>=2019.4,<2021.0a0|>=2021.2.0,<2022.0a0|>=2021.3.0,<2022.0a0|>=2021.4.0,<2022.0a0']
#0 1056.0 cucim=23.04 -> numpy[version='>=1.21.6,<2.0a0'] -> mkl[version='>=2018.0.0,<2019.0a0|>=2018.0.2,<2019.0a0|>=2018.0.3,<2019.0a0|>=2019.1,<2021.0a0|>=2019.4,<2021.0a0|>=2021.2.0,<2022.0a0|>=2021.3.0,<2022.0a0|>
=2021.4.0,<2022.0a0']
#0 1056.0
#0 1056.0 Package cudnn conflicts for:
#0 1056.0 tensorflow-gpu -> tensorflow-gpu-base==1.7.0 -> cudnn[version='6.0.*|>=7.0.5,<=8.0a0|>=7.1.2,<=8.0a0']
#0 1056.0 cuml=23.04 -> cupy[version='>=7.8.0,<12.0.0a0'] -> cudnn[version='8.0.0.*|>=8.0,<9.0a0|>=8.1.0.77,<9.0a0|>=7.6.5.32,<8.0a0|>=8.0.5.39,<9.0a0|>=7.6,<8.0a0']
#0 1056.0
#0 1056.0 Package pylibraft conflicts for:
#0 1056.0 cugraph=23.04 -> raft-dask=23.04 -> pylibraft[version='23.04.00.*|23.04.01.*']
#0 1056.0 cugraph=23.04 -> pylibraft=23.04
#0 1056.0
#0 1056.0 Package six conflicts for:
#0 1056.0 cudf=23.04 -> packaging -> six
#0 1056.0 cuxfilter=23.04 -> packaging -> six
#0 1056.0 tensorflow-gpu -> tensorflow==1.13.1 -> six[version='>=1.10.0']
#0 1056.0 cuml=23.04 -> cupy[version='>=7.8.0,<12.0.0a0'] -> six[version='>=1.9.0']
#0 1056.0
#0 1056.0 Package libgfortran5 conflicts for:
#0 1056.0 cucim=23.04 -> scipy -> libgfortran5[version='>=10.3.0|>=10.4.0|>=11.3.0|>=9.4.0|>=9.3.0|>=11.2.0']
#0 1056.0 cusignal=23.04 -> scipy[version='>=1.6.0'] -> libgfortran5[version='>=10.3.0|>=10.4.0|>=11.3.0|>=9.4.0|>=9.3.0|>=11.2.0']
#0 1056.0
#0 1056.0 Package libprotobuf conflicts for:
#0 1056.0 tensorflow-gpu -> tensorflow==1.13.1 -> libprotobuf[version='>=3.6.1,<3.7.0a0|>=3.8.0,<3.9.0a0']
#0 1056.0 cudf=23.04 -> protobuf[version='>=4.21.6,<4.22'] -> libprotobuf[version='3.21.10.*|3.21.11.*|3.21.12.*|>=3.21.12,<3.22.0a0|>=3.21.11,<3.22.0a0|>=3.21.10,<3.22.0a0|3.21.9.*|>=3.21.9,<3.22.0a0|3.21.8.*|>=3.21.8
,<3.22.0a0|3.21.7.*|>=3.21.7,<3.22.0a0|3.21.6.*|>=3.21.6,<3.22.0a0']
#0 1056.0
#0 1056.0 Package cuda-toolkit conflicts for:
#0 1056.0 cugraph=23.04 -> cuda-python[version='>=11.7.1,<12.0'] -> cuda-toolkit[version='>=11,<12.0a0']
#0 1056.0 cudf=23.04 -> cuda-python[version='>=11.7.1,<12.0'] -> cuda-toolkit[version='>=11,<12.0a0']
#0 1056.0 cuml=23.04 -> cuda-python[version='>=11.7.1,<12'] -> cuda-toolkit[version='>=11,<12.0a0']
#0 1056.0
#0 1056.0 Package bokeh conflicts for:
#0 1056.0 cuxfilter=23.04 -> datashader[version='>=0.14,<=0.14.4'] -> bokeh[version='<3.0.0|>=1.1.0,<3.0|>=1.1.0|>=2.4.3,<3.0|>=2.4.3,<2.5.0|>=2.4.3|>=2.4,<2.5|>=2.4.0,<2.5.0']
#0 1056.0 cuxfilter=23.04 -> bokeh[version='>=2.4.2,<=2.5']
#0 1056.0
#0 1056.0 Package ncurses conflicts for:
#0 1056.0 python=3.10 -> ncurses[version='>=6.2,<6.3.0a0|>=6.3,<7.0a0|>=6.2,<7.0.0a0|>=6.4,<7.0a0|>=6.2,<7.0a0']
#0 1056.0 python=3.10 -> readline[version='>=8.0,<9.0a0'] -> ncurses[version='>=6.1,<7.0.0a0|>=6.1,<7.0a0']
#0 1056.0
#0 1056.0 Package locket conflicts for:
#0 1056.0 cuml=23.04 -> distributed==2023.3.2.1 -> locket[version='>=1.0.0']
#0 1056.0 cugraph=23.04 -> distributed==2023.3.2.1 -> locket[version='>=1.0.0']The following specifications were found to be incompatible with your system:
#0 1056.0
#0 1056.0   - feature:/linux-64::__glibc==2.31=0
#0 1056.0   - feature:/linux-64::__linux==5.15.0=0
#0 1056.0   - feature:/linux-64::__unix==0=0
#0 1056.0   - feature:|@/linux-64::__glibc==2.31=0
#0 1056.0   - feature:|@/linux-64::__linux==5.15.0=0
#0 1056.0   - feature:|@/linux-64::__unix==0=0
#0 1056.0   - cucim=23.04 -> __glibc[version='>=2.17,<3.0.a0']
#0 1056.0   - cucim=23.04 -> click -> __unix
#0 1056.0   - cucim=23.04 -> click -> __win
#0 1056.0   - cucim=23.04 -> cupy[version='>=10,<12.0.0a0'] -> __glibc[version='>=2.17']
#0 1056.0   - cudf=23.04 -> __glibc[version='>=2.17,<3.0.a0']
#0 1056.0   - cudf=23.04 -> cuda-python[version='>=11.7.1,<12.0'] -> __glibc[version='>=2.17']
#0 1056.0   - cugraph=23.04 -> __glibc[version='>=2.17,<3.0.a0']
#0 1056.0   - cugraph=23.04 -> cuda-python[version='>=11.7.1,<12.0'] -> __glibc[version='>=2.17']
#0 1056.0   - cugraph=23.04 -> dask-cuda=23.04 -> __linux
#0 1056.0   - cuml=23.04 -> __glibc[version='>=2.17,<3.0.a0']
#0 1056.0   - cuml=23.04 -> cuda-python[version='>=11.7.1,<12'] -> __glibc[version='>=2.17']
#0 1056.0   - cusignal=23.04 -> cupy[version='>=10,<12.0.0a0'] -> __glibc[version='>=2.17|>=2.17,<3.0.a0']
#0 1056.0   - cuspatial=23.04 -> __glibc[version='>=2.17,<3.0.a0']
#0 1056.0   - cuxfilter=23.04 -> cudf=23.04 -> __glibc[version='>=2.17|>=2.17,<3.0.a0']
#0 1056.0   - cuxfilter=23.04 -> dask-cuda=23.04 -> __linux
#0 1056.0   - python=3.10 -> libgcc-ng[version='>=10.3.0'] -> __glibc[version='>=2.17']
#0 1056.0   - tensorflow-gpu -> tensorflow==2.11.1=cuda112py39h01bd6f0_0 -> __cuda
#0 1056.0 cuml=23.04 -> cupy[version='>=7.8.0,<12.0.0a0'] -> cudnn[version='8.0.0.*|>=8.0,<9.0a0|>=8.1.0.77,<9.0a0|>=7.6.5.32,<8.0a0|>=8.0.5.39,<9.0a0|>=7.6,<8.0a0']
#0 1056.0
#0 1056.0 Package pylibraft conflicts for:
#0 1056.0 cugraph=23.04 -> raft-dask=23.04 -> pylibraft[version='23.04.00.*|23.04.01.*']
#0 1056.0 cugraph=23.04 -> pylibraft=23.04
#0 1056.0
#0 1056.0 Package six conflicts for:
#0 1056.0 cudf=23.04 -> packaging -> six
#0 1056.0 cuxfilter=23.04 -> packaging -> six
#0 1056.0 tensorflow-gpu -> tensorflow==1.13.1 -> six[version='>=1.10.0']
#0 1056.0 cuml=23.04 -> cupy[version='>=7.8.0,<12.0.0a0'] -> six[version='>=1.9.0']
#0 1056.0
#0 1056.0 Package libgfortran5 conflicts for:
#0 1056.0 cucim=23.04 -> scipy -> libgfortran5[version='>=10.3.0|>=10.4.0|>=11.3.0|>=9.4.0|>=9.3.0|>=11.2.0']
#0 1056.0 cusignal=23.04 -> scipy[version='>=1.6.0'] -> libgfortran5[version='>=10.3.0|>=10.4.0|>=11.3.0|>=9.4.0|>=9.3.0|>=11.2.0']
#0 1056.0
#0 1056.0 Package libprotobuf conflicts for:
#0 1056.0 tensorflow-gpu -> tensorflow==1.13.1 -> libprotobuf[version='>=3.6.1,<3.7.0a0|>=3.8.0,<3.9.0a0']
#0 1056.0 cudf=23.04 -> protobuf[version='>=4.21.6,<4.22'] -> libprotobuf[version='3.21.10.*|3.21.11.*|3.21.12.*|>=3.21.12,<3.22.0a0|>=3.21.11,<3.22.0a0|>=3.21.10,<3.22.0a0|3.21.9.*|>=3.21.9,<3.22.0a0|3.21.8.*|>=3.21.8
,<3.22.0a0|3.21.7.*|>=3.21.7,<3.22.0a0|3.21.6.*|>=3.21.6,<3.22.0a0']
#0 1056.0
#0 1056.0 Package cuda-toolkit conflicts for:
#0 1056.0 cugraph=23.04 -> cuda-python[version='>=11.7.1,<12.0'] -> cuda-toolkit[version='>=11,<12.0a0']
#0 1056.0 cudf=23.04 -> cuda-python[version='>=11.7.1,<12.0'] -> cuda-toolkit[version='>=11,<12.0a0']
#0 1056.0 cuml=23.04 -> cuda-python[version='>=11.7.1,<12'] -> cuda-toolkit[version='>=11,<12.0a0']
#0 1056.0
#0 1056.0 Package bokeh conflicts for:
#0 1056.0 cuxfilter=23.04 -> datashader[version='>=0.14,<=0.14.4'] -> bokeh[version='<3.0.0|>=1.1.0,<3.0|>=1.1.0|>=2.4.3,<3.0|>=2.4.3,<2.5.0|>=2.4.3|>=2.4,<2.5|>=2.4.0,<2.5.0']
#0 1056.0 cuxfilter=23.04 -> bokeh[version='>=2.4.2,<=2.5']
#0 1056.0
#0 1056.0 Package ncurses conflicts for:
#0 1056.0 python=3.10 -> ncurses[version='>=6.2,<6.3.0a0|>=6.3,<7.0a0|>=6.2,<7.0.0a0|>=6.4,<7.0a0|>=6.2,<7.0a0']
#0 1056.0 python=3.10 -> readline[version='>=8.0,<9.0a0'] -> ncurses[version='>=6.1,<7.0.0a0|>=6.1,<7.0a0']
#0 1056.0
#0 1056.0 Package locket conflicts for:
#0 1056.0 cuml=23.04 -> distributed==2023.3.2.1 -> locket[version='>=1.0.0']
#0 1056.0 cugraph=23.04 -> distributed==2023.3.2.1 -> locket[version='>=1.0.0']The following specifications were found to be incompatible with your system:
#0 1056.0
#0 1056.0   - feature:/linux-64::__glibc==2.31=0
#0 1056.0   - feature:/linux-64::__linux==5.15.0=0
#0 1056.0   - feature:/linux-64::__unix==0=0
#0 1056.0   - feature:|@/linux-64::__glibc==2.31=0
#0 1056.0   - feature:|@/linux-64::__linux==5.15.0=0
#0 1056.0   - feature:|@/linux-64::__unix==0=0
#0 1056.0   - cucim=23.04 -> __glibc[version='>=2.17,<3.0.a0']
#0 1056.0   - cucim=23.04 -> click -> __unix
#0 1056.0   - cucim=23.04 -> click -> __win
#0 1056.0   - cucim=23.04 -> cupy[version='>=10,<12.0.0a0'] -> __glibc[version='>=2.17']
#0 1056.0   - cudf=23.04 -> __glibc[version='>=2.17,<3.0.a0']
#0 1056.0   - cudf=23.04 -> cuda-python[version='>=11.7.1,<12.0'] -> __glibc[version='>=2.17']
#0 1056.0   - cugraph=23.04 -> __glibc[version='>=2.17,<3.0.a0']
#0 1056.0   - cugraph=23.04 -> cuda-python[version='>=11.7.1,<12.0'] -> __glibc[version='>=2.17']
#0 1056.0   - cugraph=23.04 -> dask-cuda=23.04 -> __linux
#0 1056.0   - cuml=23.04 -> __glibc[version='>=2.17,<3.0.a0']
#0 1056.0   - cuml=23.04 -> cuda-python[version='>=11.7.1,<12'] -> __glibc[version='>=2.17']
#0 1056.0   - cusignal=23.04 -> cupy[version='>=10,<12.0.0a0'] -> __glibc[version='>=2.17|>=2.17,<3.0.a0']
#0 1056.0   - cuspatial=23.04 -> __glibc[version='>=2.17,<3.0.a0']
#0 1056.0   - cuxfilter=23.04 -> cudf=23.04 -> __glibc[version='>=2.17|>=2.17,<3.0.a0']
#0 1056.0   - cuxfilter=23.04 -> dask-cuda=23.04 -> __linux
#0 1056.0   - python=3.10 -> libgcc-ng[version='>=10.3.0'] -> __glibc[version='>=2.17']
#0 1056.0   - tensorflow-gpu -> tensorflow==2.11.1=cuda112py39h01bd6f0_0 -> __cuda
#0 1056.0   - tensorflow-gpu -> tensorflow==2.6.2=cuda111py37hf54207c_2 -> __glibc[version='>=2.17']
#0 1056.0
#0 1056.0 Your installed version is: 2.31
#0 1056.0
#0 1056.0
------
Dockerfile:7
--------------------
   5 |
   6 |     COPY environment.yml .
   7 | >>> RUN conda env create
   8 |     RUN conda init bash
   9 |     RUN echo "source activate myenv" >> ~/.bashrc
--------------------
ERROR: failed to solve: process "/bin/sh -c conda env create" did not complete successfully: exit code: 1
ERROR: Service 'ai' failed to build : Build failed
(venv) ray@raymond-U:~/dev/ray-docker/python.cuml$ sudo cp *.csv ../../data/cuml/aac/

I have tried to replace the tensorflow-gpu with tensorflow in the environment.yml. The environment can build. However after start the docker container with the following docker-compose.yml, the tensorflow cannot see any GPU.

version: '3'
services:
  ai:
    build: .
    image: cuml-image:1.0
    container_name: cuml
    command:  /bin/bash -c "source ~/.bashrc && jupyter notebook --ip=0.0.0.0 --no-browser --certfile=/root/.jupyter/jupyter.crt --keyfile=/root/.jupyter/jupyter.key --NotebookApp.token=xyz --allow-root /data > jupyter.log 2>&1 " 
    ports:
      - "8889:8888"    
    runtime: nvidia
    environment:
      - NVIDIA_VISIBLE_DEVICES=all

PS. I had also another docker container that creates by conda environment with tensorflow-gpu and pytorch-gpu as dependencies. In that environement tensorflow can see the my GPU correctly. However, I need to try cuml together with tensorflow and need to create the above environment.

raymond.mh.ng
  • 343
  • 2
  • 3
  • 21
  • In general, you should be fine using the standard `tensorflow` package. If you're able to get things working bare metal but not in Docker, something may not be working with your docker setup. On my system, I'm able to build your environment bare metal and Tensorflow can see the GPUs. If you provide some info about your system, docker version, and how you're launching the container, we may be able to better weigh in. – Nick Becker Apr 24 '23 at 17:01
  • After starting the container, try print(tf.config.list_physical_devices('GPU')), it return an empty list. Also, tf.test.is_built_with_cuda(), return False. I try the same comment with tensorflow-gpu conda alone, The first line return, [PhysicalDevice(name='/physical_device:GPU:0', device_type='GPU')]. The 2nd line return True. – raymond.mh.ng Apr 26 '23 at 00:21

0 Answers0