Testing AI services with Python client library

Copy link to section

To test the logic of your AI service locally by using the RuntimeContext class of the watsonx.ai Python client library, follow these steps:

1. Use the RuntimeContext class of the Python client library to test your AI service locally:

   from ibm_watsonx_ai.deployments import RuntimeContext
   
   context = RuntimeContext(
       api_client=client, request_payload_json={}
   )
   
   # custom is optional argument which is specified during the time of creation of deployment
   custom_object = {"space_id": space_id}
   
   generate, generate_stream, generate_batch = basic_generate_demo(context, **custom_object)
   
   Copy to clipboardCopied to clipboard

   For more information, see watsonx.ai Python client library documentation for using RuntimeContext for AI services.

2. Depending on your use case, you can test the generate(), generate_stream(), or generate_batch() functions as follows:

   • To test the generate() function:

     context.request_payload_json = { "test": "ai_service inference payload"}
     print(generate(context))
     Copy to clipboardCopied to clipboard

   • To test the generate_stream() function:

     context.request_payload_json = {"sse": ["ai_service_stream", "inference", "test"]}
     for data in generate_stream(context):
         print(data)
     Copy to clipboardCopied to clipboard

   • To test the generate_batch() function:

     input_data_references = [
         {
             "type": "connection_asset",
             "connection": {"id": "2d07a6b4-8fa9-43ab-91c8-befcd9dab8d2"},
             "location": {
                 "bucket": "wml-v4-fvt-batch-pytorch-connection-input",
                 "file_name": "testing-123",
             },
         }
     ]
     output_data_reference = {
         "type": "data_asset",
         "location": {"name": "nb-pytorch_output.zip"},
     }
     
     generate_batch(input_data_references, output_data_reference)
     Copy to clipboardCopied to clipboard

Learn more

Copy link to section

• Creating AI service assets

Parent topic: Deploying AI services with direct coding