In [1]:

    

# To use the latest publish `kubeflow-metadata` library, you can run:
!pip install kubeflow-metadata --user
# Install other packages:
!pip install pandas --user
# Then restart the Notebook kernel.


    

In [2]:

    

import pandas
from kubeflow.metadata import metadata
from datetime import datetime
from uuid import uuid4


    

In [3]:

    

ws1 = metadata.Workspace(
    # Connect to metadata service in namespace kubeflow in k8s cluster.
    store=metadata.Store(grpc_host=METADATA_STORE_HOST, grpc_port=METADATA_STORE_PORT),
    name="workspace_1",
    description="a workspace for testing",
    labels={"n1": "v1"})


    

In [4]:

    

r = metadata.Run(
    workspace=ws1,
    name="run-" + datetime.utcnow().isoformat("T") ,
    description="a run in ws_1",
)


    

In [5]:

    

exec = metadata.Execution(
    name = "execution" + datetime.utcnow().isoformat("T") ,
    workspace=ws1,
    run=r,
    description="execution example",
)
print("An execution was created with id %s" % exec.id)


    

    




An execution was created with id 290

In [6]:

    

date_set_version = "data_set_version_" + str(uuid4())
data_set = exec.log_input(
        metadata.DataSet(
            description="an example data",
            name="mytable-dump",
            owner="owner@my-company.org",
            uri="file://path/to/dataset",
            version=date_set_version,
            query="SELECT * FROM mytable"))
print("Data set id is {0.id} with version '{0.version}'".format(data_set))


    

    




Data set id is 171 with version 'data_set_version_cbebc757-0d76-4e1e-bbd9-02b065e4c3ea'

In [7]:

    

model_version = "model_version_" + str(uuid4())
model = exec.log_output(
    metadata.Model(
            name="MNIST",
            description="model to recognize handwritten digits",
            owner="someone@kubeflow.org",
            uri="gcs://my-bucket/mnist",
            model_type="neural network",
            training_framework={
                "name": "tensorflow",
                "version": "v1.0"
            },
            hyperparameters={
                "learning_rate": 0.5,
                "layers": [10, 3, 1],
                "early_stop": True
            },
            version=model_version,
            labels={"mylabel": "l1"}))
print(model)
print("\nModel id is {0.id} and version is {0.version}".format(model))


    

    




kubeflow.metadata.metadata.Model(workspace=None, name='MNIST', description='model to recognize handwritten digits', owner='someone@kubeflow.org', uri='gcs://my-bucket/mnist', version='model_version_50b419e2-af69-4c0e-a251-78246d4c0578', model_type='neural network', training_framework={'name': 'tensorflow', 'version': 'v1.0'}, hyperparameters={'learning_rate': 0.5, 'layers': [10, 3, 1], 'early_stop': True}, labels={'mylabel': 'l1'}, id=172, create_time='2019-12-04T00:44:49.444411Z', kwargs={})

Model id is 172 and version is model_version_50b419e2-af69-4c0e-a251-78246d4c0578

In [8]:

    

metrics = exec.log_output(
    metadata.Metrics(
            name="MNIST-evaluation",
            description="validating the MNIST model to recognize handwritten digits",
            owner="someone@kubeflow.org",
            uri="gcs://my-bucket/mnist-eval.csv",
            data_set_id=str(data_set.id),
            model_id=str(model.id),
            metrics_type=metadata.Metrics.VALIDATION,
            values={"accuracy": 0.95},
            labels={"mylabel": "l1"}))
print("Metrics id is %s" % metrics.id)


    

    




Metrics id is 173

In [9]:

    

serving_application = metadata.Execution(
    name="serving model",
    workspace=ws1,
    description="an execution to represent model serving component",
)
# Noticed we use model name, version, uri to uniquely identify existing model.
served_model = metadata.Model(
    name="MNIST",
    uri="gcs://my-bucket/mnist",
    version=model.version,
)
m=serving_application.log_input(served_model)
print("Found the mode with id {0.id} and version '{0.version}'.".format(m))


    

    




Found the mode with id 172 and version 'model_version_50b419e2-af69-4c0e-a251-78246d4c0578'.

In [10]:

    

pandas.DataFrame.from_dict(ws1.list(metadata.Model.ARTIFACT_TYPE_NAME))


    

    
Out[10]:







  

    

      

      
id
      
workspace
      
run
      
create_time
      
version
      
owner
      
description
      
name
      
model_type
      
uri
      
training_framework
      
hyperparameters
      
labels
      
kwargs
    
  
  

    

      
0
      
152
      
workspace_1
      
run-2019-12-04T00:10:29.797245
      
2019-12-04T00:10:30.516009Z
      
model_version_68fd3ccd-bb61-496a-b6b3-b9a7ab7a...
      
someone@kubeflow.org
      
model to recognize handwritten digits
      
MNIST
      
neural network
      
gcs://my-bucket/mnist
      
{'name': 'tensorflow', 'version': 'v1.0'}
      
{'learning_rate': 0.5, 'layers': [10, 3, 1], '...
      
{'mylabel': 'l1'}
      
{}
    
    

      
1
      
154
      
workspace_1
      
run-2019-12-04T00:10:29.797245
      
2019-12-04T00:14:44.401822Z
      
model_version_4757fb07-20f5-4443-8000-ba0ad5ca...
      
someone@kubeflow.org
      
model to recognize handwritten digits
      
MNIST
      
neural network
      
gcs://my-bucket/mnist
      
{'name': 'tensorflow', 'version': 'v1.0'}
      
{'learning_rate': 0.5, 'layers': [10, 3, 1], '...
      
{'mylabel': 'l1'}
      
{}
    
    

      
2
      
156
      
workspace_1
      
run-2019-12-04T00:10:29.797245
      
2019-12-04T00:14:56.893195Z
      
model_version_aa162457-85e4-473d-9f20-8fe72637...
      
someone@kubeflow.org
      
model to recognize handwritten digits
      
MNIST
      
neural network
      
gcs://my-bucket/mnist
      
{'name': 'tensorflow', 'version': 'v1.0'}
      
{'learning_rate': 0.5, 'layers': [10, 3, 1], '...
      
{'mylabel': 'l1'}
      
{}
    
    

      
3
      
160
      
workspace_1
      
run-2019-12-04T00:15:41.059380
      
2019-12-04T00:15:41.766812Z
      
model_version_7af7615c-ec8d-4e30-b765-14dd1d81...
      
someone@kubeflow.org
      
model to recognize handwritten digits
      
MNIST
      
neural network
      
gcs://my-bucket/mnist
      
{'name': 'tensorflow', 'version': 'v1.0'}
      
{'learning_rate': 0.5, 'layers': [10, 3, 1], '...
      
{'mylabel': 'l1'}
      
{}
    
    

      
4
      
163
      
workspace_1
      
run-2019-12-04T00:43:16.177845
      
2019-12-04T00:43:16.907629Z
      
model_version_3d8540f1-f74f-4660-83dc-2b5a886d...
      
someone@kubeflow.org
      
model to recognize handwritten digits
      
MNIST
      
neural network
      
gcs://my-bucket/mnist
      
{'name': 'tensorflow', 'version': 'v1.0'}
      
{'learning_rate': 0.5, 'layers': [10, 3, 1], '...
      
{'mylabel': 'l1'}
      
{}
    
    

      
5
      
166
      
workspace_1
      
run-2019-12-04T00:43:23.426877
      
2019-12-04T00:43:24.138877Z
      
model_version_9987cd2f-d354-49b5-955e-952f21ff...
      
someone@kubeflow.org
      
model to recognize handwritten digits
      
MNIST
      
neural network
      
gcs://my-bucket/mnist
      
{'name': 'tensorflow', 'version': 'v1.0'}
      
{'learning_rate': 0.5, 'layers': [10, 3, 1], '...
      
{'mylabel': 'l1'}
      
{}
    
    

      
6
      
169
      
workspace_1
      
run-2019-12-04T00:43:47.136103
      
2019-12-04T00:43:47.851783Z
      
model_version_123141d7-a3ce-4ac4-ade9-e750ad18...
      
someone@kubeflow.org
      
model to recognize handwritten digits
      
MNIST
      
neural network
      
gcs://my-bucket/mnist
      
{'name': 'tensorflow', 'version': 'v1.0'}
      
{'learning_rate': 0.5, 'layers': [10, 3, 1], '...
      
{'mylabel': 'l1'}
      
{}
    
    

      
7
      
172
      
workspace_1
      
run-2019-12-04T00:44:48.687078
      
2019-12-04T00:44:49.444411Z
      
model_version_50b419e2-af69-4c0e-a251-78246d4c...
      
someone@kubeflow.org
      
model to recognize handwritten digits
      
MNIST
      
neural network
      
gcs://my-bucket/mnist
      
{'name': 'tensorflow', 'version': 'v1.0'}
      
{'learning_rate': 0.5, 'layers': [10, 3, 1], '...
      
{'mylabel': 'l1'}
      
{}
    
  

In [11]:

    

print("Model id is %s\n" % model.id)
    
model_events = ws1.store.get_events_by_artifact_ids([model.id])

execution_ids = set(e.execution_id for e in model_events)
print("All executions related to the model are {}".format(execution_ids))
# assert execution_ids == set([serving_application.id, exec.id])

trainer_events = ws1.store.get_events_by_execution_ids([exec.id])
artifact_ids = set(e.artifact_id for e in trainer_events)
print("All artifacts related to the training event are {}".format(artifact_ids))# assert artifact_ids == set([model.id, metrics.id, data_set.id])


    

    




Model id is 172

All executions related to the model are {290, 291}
All artifacts related to the training event are {171, 172, 173}