How to setup pytorch in google-cloud-ml

前端 未结 2 609
無奈伤痛
無奈伤痛 2021-01-25 10:10

I try to throw job with Pytorch code in google-cloud-ml. so I code the \"setup.py\" file. And add option \"install_requires\"

\"setup.py\"

<
相关标签:
2条回答
  • 2021-01-25 10:34

    i find solution about setting up PYTORCH in google-cloud-ml

    first you have to get a .whl file about pytorch and store it to google storage bucket. and you will get the link for bucket link.

    gs://bucketname/directory/torch-0.3.0.post4-cp27-cp27mu-linux_x86_64.whl
    

    the .whl file is depend on your python version or cuda version....

    second you write the command line and setup.py because you have to set up the google-cloud-ml setting. related link is this submit_job_to_ml-engine you write the setup.py file to describe your setup. the related link is this write_setup.py_file

    this is my command code and setup.py file

    ===================================================================== "command"

    #commandline code
    JOB_NAME="run_ml_engine_pytorch_test_$(date +%Y%m%d_%H%M%S)"
    REGION=us-central1
    OUTPUT_PATH=gs://yourbucket
    gcloud ml-engine jobs submit training $JOB_NAME \
        --job-dir $OUTPUT_PATH \
        --runtime-version 1.4 \
        --module-name models.pytorch_test \
        --package-path models/ \
        --packages gs://yourbucket/directory/torch-0.3.0.post4-cp27-cp27mu-linux_x86_64.whl \
        --region $REGION \
        -- \
        --verbosity DEBUG
    

    ===================================================================== "setup.py"

    from setuptools import find_packages
    from setuptools import setup
    REQUIRED_PACKAGES = ['torchvision']
    setup(
        name='trainer',
        version='0.1',
        install_requires=REQUIRED_PACKAGES,
        packages=find_packages(),
        include_package_data=True,
        description='My pytorch trainer application package.'
    )
    

    =====================================================================

    third if you have experience submitting job to the ml-engine. you might know the file structure about submitting ml-engine packaging_training_model. you have to follow above link and know how to pack files.

    0 讨论(0)
  • 2021-01-25 11:00

    The actual error message is a bit buried, but it is this:

    'install_requires' must be a string or list of strings containing valid project/version requirement specifiers; Invalid requirement, parse error at "'://downl'"

    To use packages not hosted on PyPI, you need to use dependency_links (see this documentation). Something like this ought to work:

    from setuptools import find_packages
    from setuptools import setup
    
    REQUIRED_PACKAGES = ['torchvision']
    DEPENDENCY_LINKS = ['http://download.pytorch.org/whl/cpu/torch-0.3.0.post4-cp27-cp27mu-linux_x86_64.whl']
    
    setup(
        name='trainer',
        version='0.1',
        install_requires=REQUIRED_PACKAGES,
        dependency_links=DEPENDENCY_LINKS,
        packages=find_packages(),
        include_package_data=True,
        description='My keras trainer application package.'
    )
    
    0 讨论(0)
提交回复
热议问题