0 / 0
Customizing with third-party and private Python libraries
Customizing with third-party and private Python libraries

Customizing with third-party and private Python libraries

If your model requires custom components such as user-defined transformers, estimators, or user-defined tensors, you can create a custom software specification that is derived from a base, or a predefined specification. Python functions and Python scripts also support custom software specifications.

You can use custom software specification to reference any third-party libraries, user-created Python packages, or both. Third-party libraries or user-created Python packages must be specified as package extensions that can then be referenced in a custom software specification.

You can customize deployment runtimes in these ways:

For troubleshooting information, refer to Troubleshooting.

Defining customizations in a Watson Studio project and then promoting them to a deployment space

Environments in Watson Studio projects can be customized to include third-party libraries that can be installed from Anaconda or from the PyPI repository.

For details on creating environments with software specifications, refer to Environments.

As part of custom environment creation, these steps are performed internally (transparent to the user):

  • A package extension containing the details of third-party libraries is created in conda YAML format.
  • A custom software specification with the same name as that of the custom environment is created and the package extension that has been created is associated with this custom software specification.

The models or Python functions/scripts created with the custom environment must reference the custom software specification when they are saved in Watson Machine Learning repository in the project scope.

Propagating software specifications and package extensions from projects to deployment spaces

To export custom software specifications and package extensions that were created in a Watson Studio project to a deployment space:

  1. From your project interface, click the Manage tab.
  2. Select Environments.
  3. Click the Templates tab.
  4. From your custom environment's Options menu, select Promote to space.

Selecting "Promote to space" for a custom environment in Watson Studio interface

Alternatively, when you promote any model or Python function that is associated with a custom environment from a Watson Studio project to a deployment space, the associated custom software specification and package extension is also promoted to the deployment space.

If you want to update software specifications and package extensions after they were promoted to the deployment space, follow these steps:

  1. In the deployment space, delete the software specifications, package extensions, and associated models (optional) by using the Watson Machine Learning Python client.
  2. In a Watson Studio project, promote the model, function, or script that is associated with the changed custom software specification and package extension to the space.

Software specifications are also included when you import a project or space that includes one.

Creating package extensions and custom software specifications in a deployment space by using the Watson Machine Learning Python client

You can use the Watson Machine Learning APIs or Python client to define a custom software specification that is derived from a base specification.

High-level steps to create a custom software specification that uses third-party libraries and/or user-created Python packages:

  1. Optional: Save a conda YAML file that contains a list of third-party libraries or save a user-created Python library and create a package extension.
    Note: This step is not required if the model does not have any dependency on a third-party library or a user-created Python library.
  2. Create a custom software specification
  3. Add a reference of the package extensions to the custom software specification that you created.

Saving a conda YAML file that contains a list of third-party libraries

To save a conda YAML file that contains a list of third-party libraries as a package extension and create a custom software specification linked to the package extension:

  1. Authenticate and create the client.

    Refer to Authentication.

  2. Create and set the default deployment space, then list available software specifications.

     metadata = {
         wml_client.spaces.ConfigurationMetaNames.NAME:
             'examples-create-software-spec',
         wml_client.spaces.ConfigurationMetaNames.DESCRIPTION:
             'For my models'
     }
     space_details = wml_client.spaces.store(meta_props=metadata)
     space_uid = wml_client.spaces.get_id(space_details)
    
     # set the default space
     wml_client.set.default_space(space_uid)
    
     # see available meta names for software specs
     print('Available software specs configuration:', wml_client.software_specifications.ConfigurationMetaNames.get())
     wml_client.software_specifications.list()
    
     asset_id = 'undefined'
     pe_asset_id = 'undefined'
    
  3. Create the metadata for package extensions to add to the base specification.

     pe_metadata = {
         wml_client.package_extensions.ConfigurationMetaNames.NAME:
             'My custom library',
         # optional:
         # wml_client.software_specifications.ConfigurationMetaNames.DESCRIPTION:
         wml_client.package_extensions.ConfigurationMetaNames.TYPE:
             'conda_yml'
     }
    
  4. Create a yaml file that contains the list of additional packages and then save it as customlibrary.yaml.

    Example yaml file:

     name: add-regex-package
     dependencies:
         - regex
    

    For information on the proper yaml file structure, refer to Examples of customizations.

  5. Store package extension information.

     pe_asset_details = wml_client.package_extensions.store(
         meta_props=pe_metadata,
         file_path='customlibrary.yaml'
     )
     pe_asset_id = wml_client.package_extensions.get_id(pe_asset_details)
    
  6. Create the metadata for the software specification and store the software specification.

     # Get the id of the base software specification
     base_id = wml_client.software_specifications.get_id_by_name('default_py3.9')
    
     # create the metadata for software specs
     ss_metadata = {
         wml_client.software_specifications.ConfigurationMetaNames.NAME:
             'Python 3.9 with pre-installed ML package',
         wml_client.software_specifications.ConfigurationMetaNames.DESCRIPTION:
             'Adding some custom libraries like regex', # optional
         wml_client.software_specifications.ConfigurationMetaNames.BASE_SOFTWARE_SPECIFICATION:
             {'guid': base_id},
         wml_client.software_specifications.ConfigurationMetaNames.PACKAGE_EXTENSIONS:
             [{'guid': pe_asset_id}]
     }
    
     # store the software spec
     ss_asset_details = wml_client.software_specifications.store(meta_props=ss_metadata)
    
     # get the id of the new asset
     asset_id = wml_client.software_specifications.get_id(ss_asset_details)
    
     # view new software specification details
     import pprint as pp
    
     ss_asset_details = wml_client.software_specifications.get_details(asset_id)
     print('Package extensions', pp.pformat(
         ss_asset_details['entity']['software_specification']['package_extensions']
     ))
    

Saving a user-created Python library and creating a package extension

For information on the requirements to support user-created Python libraries, refer to Requirements for using custom components in models.

To save a user-created Python package as a package extension and create a custom software specification linked to the package extension:

  1. Authenticate and create the client.

    Refer to Authentication.

  2. Create and set the default deployment space, then list available software specifications.

     metadata = {
         wml_client.spaces.ConfigurationMetaNames.NAME:
             'examples-create-software-spec',
         wml_client.spaces.ConfigurationMetaNames.DESCRIPTION:
             'For my models'
     }
     space_details = wml_client.spaces.store(meta_props=metadata)
     space_uid = wml_client.spaces.get_id(space_details)
    
     # set the default space
     wml_client.set.default_space(space_uid)
    
     # see available meta names for software specs
     print('Available software specs configuration:', wml_client.software_specifications.ConfigurationMetaNames.get())
     wml_client.software_specifications.list()
    
     asset_id = 'undefined'
     pe_asset_id = 'undefined'
    
  3. Create the metadata for package extensions to add to the base specification.

    Note: Only pip_zip can be specified as value for the wml_client.package_extensions.ConfigurationMetaNames.TYPE meta data property.

     pe_metadata = {
         wml_client.package_extensions.ConfigurationMetaNames.NAME:
             'My Python library',
         # optional:
         # wml_client.software_specifications.ConfigurationMetaNames.DESCRIPTION:
         wml_client.package_extensions.ConfigurationMetaNames.TYPE:
             'pip.zip'
     }
    
  4. Specify the path of the user-created Python library.

     python_lib_file_path="my-python-library-0.1.zip"
    

    For information on the requirements to support user-created Python libraries, refer to Requirements for using custom components in models.

  5. Store package extension information.

     pe_asset_details = wml_client.package_extensions.store(
         meta_props=pe_metadata,
         file_path=python_lib_file_path
     )
     pe_asset_id = wml_client.package_extensions.get_id(pe_asset_details)
    
  6. Create the metadata for the software specification and store the software specification.

     # Get the id of the base software specification
     base_id = wml_client.software_specifications.get_id_by_name('default_py3.9')
    
     # create the metadata for software specs
     ss_metadata = {
         wml_client.software_specifications.ConfigurationMetaNames.NAME:
             'Python 3.9 with pre-installed ML package',
         wml_client.software_specifications.ConfigurationMetaNames.DESCRIPTION:
             'Adding some custom libraries like regex', # optional
         wml_client.software_specifications.ConfigurationMetaNames.BASE_SOFTWARE_SPECIFICATION:
             {'guid': base_id},
         wml_client.software_specifications.ConfigurationMetaNames.PACKAGE_EXTENSIONS:
             [{'guid': pe_asset_id}]
     }
    
     # store the software spec
     ss_asset_details = wml_client.software_specifications.store(meta_props=ss_metadata)
    
     # get the id of the new asset
     asset_id = wml_client.software_specifications.get_id(ss_asset_details)
    
     # view new software specification details
     import pprint as pp
    
     ss_asset_details = wml_client.software_specifications.get_details(asset_id)
     print('Package extensions', pp.pformat(
         ss_asset_details['entity']['software_specification']['package_extensions']
     ))
    

Troubleshooting

When a conda yml based custom library installation fails with this error: Encountered error while installing custom library, try these alternatives:

  • Use a different version of the same package that is available in Anaconda for the concerned Python version.
  • Install the library from the pypi repository, by using pip. Edit the conda yml installation file contents:

     name: <conda yml file name>
     dependencies:
     - numpy
     - pip:
         - pandas==1.2.5
    

Parent topic: Customizing deployment runtimes