ibm-cpp-orchestration 0.1.0

Supports users in usage of CPD Orchestration

IBM CPD Orchestration Python Client

This package supports users in usage of CPD Orchestration. It's primary usage is to enable users to store artifact results of a Orchestration Flow run.

Usage

Client construction

CPDOrchestration client is constructed from IAM APIKEY, which can be provided in a few ways:

• explicitly:

  from ibm_cpp_orchestration import CPDOrchestration
  
  client = CPDOrchestration(apikey)
  # or
  client = CPDOrchestration.from_apikey(apikey)
  

• implicitly:

  APIKEY=...
  export APIKEY
  

  from ibm_cpp_orchestration import CPDOrchestration
  
  client = CPDOrchestration()
  # or
  client = CPDOrchestration.new_instance()
  # or
  client = CPDOrchestration.from_apikey()
  

All of the above may also define service_name and url.

Usage in Python notebooks

Notebooks run in Orchestration Platform get inputs and expose outputs as a node in Orchestration Flow:

{
  "id": ...,
  "type": "execution_node",
  "op": "run_container",
  "app_data": {
    "pipeline_data": {
      "name": ...,
      "config": {
        "link": {
          "component_id_ref": "run-notebook"
        }
      },
      "inputs": [
        ...,
        {
          "name": "model_name",
          "group": "env_variables",
          "type": "String",
          "value_from": ...
        }
      ],
      "outputs": [
        {
          "name": "trained_model",
          "group": "output_variables",
          "type": {
            "CPDPath": {
              "path_type": "resource",
              "resource_type": "asset",
              "asset_type": "wml_model"
            }
          }
        }
      ]
    }
  },
  ...
}

Inside of the notebook, inputs are available as environmental variables:

model_name = os.environ['model_name']

Outputs are exposed using sdk method, store_results:

client = CPDOrchestration.from_apikey(...)
client.store_results({
  "trained_model": ... // cpd path to the trained model
})

Other features

Client also provides a method to get WML instance credentials:

client.get_wml_credentials() # the scope passed in notebook
# or
client.get_wml_credentials("cpd://projects/123456789")

Contribution

See a separate document on contribution.