[bug] Pipelien runs fail due to wrong input parameters type casting

[erhwenkuo]

Environment

• How did you deploy Kubeflow Pipelines (KFP)?

  Use https://github.com/kubeflow/manifests to deploy kubeflow.

  git clone https://github.com/kubeflow/manifests.git
  cd manifests
  git checkout v1.7-branch

  Modify manifests/apps/pipeline/upstream/base/pipeline/kustomization.yaml:

  ...
  images:
    - name: gcr.io/ml-pipeline/api-server
      newTag: 2.0.0-rc.1
    - name: gcr.io/ml-pipeline/persistenceagent
      newTag: 2.0.0-rc.1
    - name: gcr.io/ml-pipeline/scheduledworkflow
      newTag: 2.0.0-rc.1
    - name: gcr.io/ml-pipeline/frontend
      newTag: 2.0.0-rc.1
    - name: gcr.io/ml-pipeline/viewer-crd-controller
      newTag: 2.0.0-rc.1
    - name: gcr.io/ml-pipeline/visualization-server
      newTag: 2.0.0-rc.1
  

  Then, use one command to apply these manifest:

  while ! kustomize build example | awk '!/well-defined/' | kubectl apply -f -; do echo "Retrying to apply resources"; sleep 10; done

• KFP version:

  version: 2.0.0-rc.1

Steps to reproduce

1. Create sample pipeline (v2)

   from kfp import dsl
   
   @dsl.component
   def square(x: float) -> float:
       return x ** 2
   
   @dsl.component
   def add(x: float, y: float) -> float:
       return x + y
   
   @dsl.component
   def square_root(x: float) -> float:
       return x ** .5
   
   @dsl.pipeline
   def pythagorean(a: float, b: float) -> float:
       a_sq_task = square(x=a)
       b_sq_task = square(x=b)
       sum_task = add(x=a_sq_task.output, y=b_sq_task.output)
       return square_root(x=sum_task.output).output
   
   compiler.Compiler().compile(pythagorean, 'pipeline.yaml')

2. Click "Pipelines" menu, then click "Upload pipeline"

   

3. Click "Create experiment" button and input experiment name

   

4. Configure "Start a run" with pipeline input parameters and click "Start"

   

   Use browser developer tools and find out UI call "/pipeline/apis/v2beta1/runs" and submit below JSON payload:

   {
     "description": "",
     "display_name": "Run of pythagorean (3d5ba)",
     "experiment_id": "e86993e5-7338-49f0-a129-3f1aa1676a06",
     "pipeline_version_reference": {
       "pipeline_id": "165d1ce7-ed91-4def-ac7a-14aabbdb0704",
       "pipeline_version_id": "77477a12-0a2e-4f1b-860c-2089216ea83f"
     },
     "runtime_config": {
       "parameters": {
         "a": 3,
         "b": 4
       }
     },
     "service_account": ""
   }

5. After run is triggered and execution fail.

   

   Checking the logs the error is "VaueError: Function square returned value of type; want type ...

   

Expected result

UI should detect input parameters type and convert into right JSON number representation.

Pipeline API server should do the right type casting base on IR Yaml.

Reference

The execution log of 1st task (square):

time="2023-05-17T05:52:35.570Z" level=info msg="capturing logs" argo=true
time="2023-05-17T05:52:35.602Z" level=info msg="capturing logs" argo=true
I0517 05:52:35.639643      34 launcher_v2.go:84] input ComponentSpec:{
  "inputDefinitions": {
    "parameters": {
      "x": {
        "parameterType": "NUMBER_DOUBLE"
      }
    }
  },
  "outputDefinitions": {
    "parameters": {
      "Output": {
        "parameterType": "NUMBER_DOUBLE"
      }
    }
  },
  "executorLabel": "exec-square-2"
}
I0517 05:52:35.640545      34 cache.go:139] Cannot detect ml-pipeline in the same namespace, default to ml-pipeline.kubeflow:8887 as KFP endpoint.
I0517 05:52:35.640563      34 cache.go:116] Connecting to cache endpoint ml-pipeline.kubeflow:8887
I0517 05:52:35.777714      34 object_store.go:306] Cannot detect minio-service in the same namespace, default to minio-service.kubeflow:9000 as MinIO endpoint.
WARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: 
https://pip.pypa.io/warnings/venv
[KFP Executor 2023-05-17 05:53:02,601 INFO]: Looking for component `square` in --component_module_path `/tmp/tmp.NS1HEzRMO2/ephemeral_component.py`
[KFP Executor 2023-05-17 05:53:02,602 INFO]: Loading KFP component "square" from /tmp/tmp.NS1HEzRMO2/ephemeral_component.py (directory "/tmp/tmp.NS1HEzRMO2" and module name "ephemeral_component")
[KFP Executor 2023-05-17 05:53:02,603 INFO]: Got executor_input:
{
    "inputs": {
        "parameterValues": {
            "x": 4
        }
    },
    "outputs": {
        "parameters": {
            "Output": {
                "outputFile": "/tmp/kfp/outputs/Output"
            }
        },
        "outputFile": "/tmp/kfp_outputs/output_metadata.json"
    }
}
Traceback (most recent call last):
  File "/usr/local/lib/python3.7/runpy.py", line 193, in _run_module_as_main
    "__main__", mod_spec)
  File "/usr/local/lib/python3.7/runpy.py", line 85, in _run_code
    exec(code, run_globals)
  File "/usr/local/lib/python3.7/site-packages/kfp/components/executor_main.py", line 105, in <module>
    executor_main()
  File "/usr/local/lib/python3.7/site-packages/kfp/components/executor_main.py", line 101, in executor_main
    executor.execute()
  File "/usr/local/lib/python3.7/site-packages/kfp/components/executor.py", line 341, in execute
    self._write_executor_output(result)
  File "/usr/local/lib/python3.7/site-packages/kfp/components/executor.py", line 258, in _write_executor_output
    func_output)
  File "/usr/local/lib/python3.7/site-packages/kfp/components/executor.py", line 228, in _handle_single_return_value
    f'Function `{self._func.__name__}` returned value of type {type(return_value)}; want type {origin_type}'
ValueError: Function `square` returned value of type <class 'int'>; want type <class 'float'>
F0517 05:53:02.719437      34 main.go:49] failed to execute component: exit status 1
Error: exit status 1
Error: exit status 1

Impacted by this bug? Give it a 👍.