azureml-examples/python-sdk/tutorials/automl-with-azureml/Setup using a Local Conda environment/Windows

[mariabardon]

I am following the tutorial to set up a automl (with azureml) conda enviroment in Windows. I downloaded and extracted the github folder, started an anaconda prompt window, cd to automl-with-azureml folder, and run automl_setup.

I got a warning, and an error:

DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead. versionInvalid = LooseVersion(conda.version) < LooseVersion(minimumVersion) Collecting package metadata (repodata.json): done Solving environment: failed

ResolvePackageNotFound:

• matplotlib==2.1.0
• py-xgboost[version='<=0.90']

I use windows 10 and my conda info is: conda version : 4.12.0 conda-build version : 3.21.8

[mikecowgill]

I'm also having this issue on Mac M1. Running "bash automl_setup_mac.sh" Conda version: 4.12.0

Conda Info: active environment : base active env location : /Users/mike/miniconda3 shell level : 1 user config file : /Users/mike/.condarc populated config files : conda version : 4.12.0 conda-build version : not installed python version : 3.9.12.final.0 virtual packages : __osx=12.5.1=0 __unix=0=0 __archspec=1=arm64 base environment : /Users/mike/miniconda3 (writable) conda av data dir : /Users/mike/miniconda3/etc/conda conda av metadata url : None channel URLs : https://repo.anaconda.com/pkgs/main/osx-arm64 https://repo.anaconda.com/pkgs/main/noarch https://repo.anaconda.com/pkgs/r/osx-arm64 https://repo.anaconda.com/pkgs/r/noarch package cache : /Users/mike/miniconda3/pkgs /Users/mike/.conda/pkgs envs directories : /Users/mike/miniconda3/envs /Users/mike/.conda/envs platform : osx-arm64 user-agent : conda/4.12.0 requests/2.27.1 CPython/3.9.12 Darwin/21.6.0 OSX/12.5.1 UID:GID : 501:20 netrc file : None offline mode : False

Solving environment: failed

ResolvePackageNotFound:

• python[version='>=3.5.2,<3.8']
• conda-forge::fbprophet==0.5
• lightgbm[version='<=2.3.0']
• matplotlib==2.1.0
• scikit-learn==0.22.1
• cudatoolkit=9.0
• pytorch::pytorch=1.4.0
• pandas==0.25.1
• numpy==1.18.5
• urllib3[version='<1.24']
• py-xgboost[version='<=0.90']

[mikkeyboi]

I'm also having this issue on Mac M1. Running "bash automl_setup_mac.sh" Conda version: 4.12.0

Conda Info: active environment : base active env location : /Users/mike/miniconda3 shell level : 1 user config file : /Users/mike/.condarc populated config files : conda version : 4.12.0 conda-build version : not installed python version : 3.9.12.final.0 virtual packages : __osx=12.5.1=0 __unix=0=0 __archspec=1=arm64 base environment : /Users/mike/miniconda3 (writable) conda av data dir : /Users/mike/miniconda3/etc/conda conda av metadata url : None channel URLs : https://repo.anaconda.com/pkgs/main/osx-arm64 https://repo.anaconda.com/pkgs/main/noarch https://repo.anaconda.com/pkgs/r/osx-arm64 https://repo.anaconda.com/pkgs/r/noarch package cache : /Users/mike/miniconda3/pkgs /Users/mike/.conda/pkgs envs directories : /Users/mike/miniconda3/envs /Users/mike/.conda/envs platform : osx-arm64 user-agent : conda/4.12.0 requests/2.27.1 CPython/3.9.12 Darwin/21.6.0 OSX/12.5.1 UID:GID : 501:20 netrc file : None offline mode : False

Solving environment: failed

ResolvePackageNotFound:

• python[version='>=3.5.2,<3.8']
• conda-forge::fbprophet==0.5
• lightgbm[version='<=2.3.0']
• matplotlib==2.1.0
• scikit-learn==0.22.1
• cudatoolkit=9.0
• pytorch::pytorch=1.4.0
• pandas==0.25.1
• numpy==1.18.5
• urllib3[version='<1.24']
• py-xgboost[version='<=0.90']

The environment does not work on your system. python[version='>=3.5.2,<3.8'] Mac M1 does not support Python versions less than 3.8, unless you use Rosetta or something similar. Also, cudatoolkit=9.0 won't install on macOS nor Apple Silicon (they can't act as target systems), unless you intend to use your Mac as a host tool via cuda-gdb.

As for OP, there needs to be more information, such as your conda environment packages (activate your environment, then conda list).