Use TensorFlow to predict handwritten digits Watson Machine Learning icon
Icon

This notebook contains steps and code to demonstrate Deep Learning model training and scoring in the Watson Machine Learning service. This notebook introduces commands for getting data, training_definition persistance to Watson Machine Learning repository, model training, deployment and scoring.

Some familiarity with Python is helpful. This notebook uses Python 3 and Watson Studio environments.

Learning goals¶

In this notebook, you will learn how to:

  • Work with Watson Machine Learning experiments to train Deep Learning models (Tensorflow)
  • Save trained models in the Watson Machine Learning repository
  • Deploy a trained model online and score

Contents¶

  1. Set up
  2. Create the training definitions
  3. Train the model
  4. Save the model
  5. Create the online deployment
  6. Score data
  7. Summary and next steps

1. Set up¶

Before you use the sample code in this notebook, you must perform the following setup tasks:

  • Create a Watson Machine Learning (WML) Service instance (a free plan is offered and information about how to create the instance is here).

  • Create a Cloud Object Storage (COS) instance (a lite plan is offered and information about how to order storage is here).
    Note: When using Watson Studio, you already have a COS instance associated with the project you are running the notebook in.

  • Create new credentials with HMAC:

    • Go to your COS dashboard.
    • In the Service credentials tab, click New Credential+.
    • Add the inline configuration parameter: {"HMAC":true}, click Add. (For more information, see HMAC.)

    This configuration parameter adds the following section to the instance credentials, (for use later in this notebook):

      "cos_hmac_keys": {
        "access_key_id": "-------",
        "secret_access_key": "-------"
   }

1.1 Work with Cloud Object Storage (COS)¶

Install the boto library. This library allows Python developers to manage Cloud Object Storage (COS).

Tip: If ibm_boto3 is not preinstalled in you environment, run the following command to install it:

In [1]:
# Run the command if ibm_boto3 is not installed.
!pip install ibm-cos-sdk

Requirement not upgraded as not directly required: ibm-cos-sdk in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages
Requirement not upgraded as not directly required: ibm-cos-sdk-s3transfer==2.*,>=2.0.0 in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from ibm-cos-sdk)
Requirement not upgraded as not directly required: ibm-cos-sdk-core==2.*,>=2.0.0 in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from ibm-cos-sdk)
Requirement not upgraded as not directly required: docutils>=0.10 in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from ibm-cos-sdk-core==2.*,>=2.0.0->ibm-cos-sdk)
Requirement not upgraded as not directly required: python-dateutil<3.0.0,>=2.1 in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from ibm-cos-sdk-core==2.*,>=2.0.0->ibm-cos-sdk)
Requirement not upgraded as not directly required: jmespath<1.0.0,>=0.7.1 in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from ibm-cos-sdk-core==2.*,>=2.0.0->ibm-cos-sdk)
Requirement not upgraded as not directly required: six>=1.5 in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from python-dateutil<3.0.0,>=2.1->ibm-cos-sdk-core==2.*,>=2.0.0->ibm-cos-sdk)
In [2]:
# Install the boto library.
import ibm_boto3
from ibm_botocore.client import Config

Enter your COS credentials in the following cell.

You can find these credentials in your COS instance dashboard under the Service credentials tab.

Note: the HMAC key, described in set up the environment is included in these credentials.

In [3]:
cos_credentials = {
  "apikey": "***",
  "cos_hmac_keys": {
    "access_key_id": "***",
    "secret_access_key": "***"
  },
  "endpoints": "https://cos-service.bluemix.net/endpoints",
  "iam_apikey_description": "Auto generated apikey during resource-key operation for Instance - crn:v1:bluemix:public:cloud-object-storage:global:a/07a95aa44e6124e8b320b70cf88033fa:876e5285-4bef-4cf3-a89b-595e19648c7c::",
  "iam_apikey_name": "auto-generated-apikey-19a79dae-6a58-4b4f-878f-6839b711523f",
  "iam_role_crn": "crn:v1:bluemix:public:iam::::serviceRole:Writer",
  "iam_serviceid_crn": "crn:v1:bluemix:public:iam-identity::a/07a95aa44e6124e8b320b70cf88033fa::serviceid:ServiceId-3f2cccee-61ec-4147-8732-9f58479ba26a",
  "resource_instance_id": "crn:v1:bluemix:public:cloud-object-storage:global:a/07a95aa44e6124e8b320b70cf88033fa:876e5285-4bef-4cf3-a89b-595e19648c7c::"
}
In [4]:
# The code was removed by Watson Studio for sharing.

Define the endpoint.

To do this, go to the Endpoint tab in the COS instance's dashboard to get the endpoint information, then enter it in the cell below:

In [5]:
# Define endpoint information.
service_endpoint = 'https://s3-api.us-geo.objectstorage.softlayer.net'

You also need the IBM Cloud authorization endpoint to be able to create COS resource object.

In [6]:
# Define the authorization endpoint.
auth_endpoint = 'https://iam.bluemix.net/oidc/token'

Create a Boto resource to be able to write data to COS.

In [7]:
# Create a COS resource.
cos = ibm_boto3.resource('s3',
                         ibm_api_key_id=cos_credentials['apikey'],
                         ibm_service_instance_id=cos_credentials['resource_instance_id'],
                         ibm_auth_endpoint=auth_endpoint,
                         config=Config(signature_version='oauth'),
                         endpoint_url=service_endpoint)

Create two buckets, which you will use to store training data and training results.

Note: The bucket names must be unique.

In [8]:
from uuid import uuid4

bucket_uid = str(uuid4())
buckets = ['training-data-' + bucket_uid, 'training-results-' + bucket_uid]

for bucket in buckets:
    if not cos.Bucket(bucket) in cos.buckets.all():
        print('Creating bucket "{}"...'.format(bucket))
        try:
            cos.create_bucket(Bucket=bucket)
        except ibm_boto3.exceptions.ibm_botocore.client.ClientError as e:
            print('Error: {}.'.format(e.response['Error']['Message']))

Creating bucket "training-data-eacf90dc-f69a-437c-8f43-5d36aa59abf4"...
Creating bucket "training-results-eacf90dc-f69a-437c-8f43-5d36aa59abf4"...

Now you should have 2 buckets.

In [9]:
# Display a list of created buckets.
print(list(cos.buckets.all()))

[s3.Bucket(name='9fff2508-60c5-4e75-835b-210aa00ff010-style-data'), s3.Bucket(name='9fff2508-60c5-4e75-835b-210aa00ff010-style-results'), s3.Bucket(name='aiopenscalepythonclient-donotdelete-pr-hp7rm9ptqfgzw7'), s3.Bucket(name='aispheredemo-donotdelete-pr-s7lhzejm05qqne'), s3.Bucket(name='aispherescenario-datacatalog-sydghgvlq'), s3.Bucket(name='ale-kupa-in'), s3.Bucket(name='cars4you-train-data'), s3.Bucket(name='cifar-keras-nnm-input'), s3.Bucket(name='coreml-donotdelete-pr-5lprxel1c6jrfp'), s3.Bucket(name='coreml-mnist-input-data'), s3.Bucket(name='creditcardfraud545b518d4ec34681aa7e8e8680b486d5'), s3.Bucket(name='datasphereworkshops-donotdelete-pr-xfdm0p7mjo5ms6'), s3.Bucket(name='mnist-keras-data-example'), s3.Bucket(name='mnist-keras-data-example2018-06-01-19-36'), s3.Bucket(name='mnist-keras-results-example'), s3.Bucket(name='mnist-keras-results-example2018-06-01-19-36'), s3.Bucket(name='style-data-example'), s3.Bucket(name='style-data-example-workshops'), s3.Bucket(name='style-results-example'), s3.Bucket(name='style-results-example-workshops'), s3.Bucket(name='train-data-experiment-2018-03-28-09-34'), s3.Bucket(name='train-data-experiment-2018-05-30-14-26'), s3.Bucket(name='train-data-experiment-2018-08-16-13-11'), s3.Bucket(name='train-result-experiment-2018-03-28-09-34'), s3.Bucket(name='train-result-experiment-2018-05-30-14-26'), s3.Bucket(name='train-result-experiment-2018-08-16-13-11'), s3.Bucket(name='training-data-38a61fd8-a250-4a82-9277-ef51f88bfadb'), s3.Bucket(name='training-data-47687ebc-1597-4cdb-b692-51600c0fa77e'), s3.Bucket(name='training-data-eacf90dc-f69a-437c-8f43-5d36aa59abf4'), s3.Bucket(name='training-results-38a61fd8-a250-4a82-9277-ef51f88bfadb'), s3.Bucket(name='training-results-47687ebc-1597-4cdb-b692-51600c0fa77e'), s3.Bucket(name='training-results-eacf90dc-f69a-437c-8f43-5d36aa59abf4'), s3.Bucket(name='watsonstudiosamplenotebooks2ec122d7327248859e8d39c03f931082'), s3.Bucket(name='xxx-donotdelete-pr-ly02ysmxmvyqcr')]

1.2 Download the training data and upload it to the COS buckets¶

Download the training data and upload it to the 'training-data' bucket. First, create a list of links for the training data set.

In [10]:
# Create a list of links.
data_links = ['http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz',
              'http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz',
              'http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz',
              'http://yann.lecun.com/exdb/mnist/t10k-labels-idx1-ubyte.gz']

The code in the next cell uploads files from links to your COS.

In [11]:
# Upload files from the links to COS
from urllib.request import urlopen

bucket_obj = cos.Bucket(buckets[0])

for data_link in data_links:
    filename=data_link.split('/')[-1]
    print('Uploading data {}...'.format(filename))
    with urlopen(data_link) as data:
        bucket_obj.upload_fileobj(data, filename)
        print('{} is uploaded.'.format(filename))

Uploading data train-images-idx3-ubyte.gz...
train-images-idx3-ubyte.gz is uploaded.
Uploading data train-labels-idx1-ubyte.gz...
train-labels-idx1-ubyte.gz is uploaded.
Uploading data t10k-images-idx3-ubyte.gz...
t10k-images-idx3-ubyte.gz is uploaded.
Uploading data t10k-labels-idx1-ubyte.gz...
t10k-labels-idx1-ubyte.gz is uploaded.

Have a look at the list of the created buckets and their contents.

In [12]:
for bucket_name in buckets:
    print(bucket_name)
    bucket_obj = cos.Bucket(bucket_name)
    for obj in bucket_obj.objects.all():
        print("  File: {}, {:4.2f}kB".format(obj.key, obj.size/1024))

training-data-eacf90dc-f69a-437c-8f43-5d36aa59abf4
  File: t10k-images-idx3-ubyte.gz, 1610.23kB
  File: t10k-labels-idx1-ubyte.gz, 4.44kB
  File: train-images-idx3-ubyte.gz, 9680.10kB
  File: train-labels-idx1-ubyte.gz, 28.20kB
training-results-eacf90dc-f69a-437c-8f43-5d36aa59abf4

You are done with COS, and you are ready to train your model!

1.3. Work with the WML service instance¶

Import the libraries you need to work with your WML instance.

Hint: You may also need to install wget using the following command !pip install wget

In [13]:
import urllib3, requests, json, base64, time, os, wget
!pip install dask --upgrade

Authenticate to the Watson Machine Learning (WML) service on IBM Cloud.

Tip: Authentication information (your credentials) can be found in the Service credentials tab of the service instance that you created on IBM Cloud. If there are no credentials listed for your instance in Service credentials, click New credential (+) and enter the information required to generate new authentication information.

Action: Enter your WML service instance credentials here.

In [14]:
wml_credentials = {
  "url": "https://ibm-watson-ml.mybluemix.net",
  "access_key": "vgXPA8k5Ld613WJP/S0ijQZtiitjuwsj9qpVt+nvl9DFl8PK/67XgmnreKabd2ZPpxSFKe9cZoFYLlzgPf++qpWZYcc+6fawL9S0V+2V79Adc+zik+ZHJYrsBRl9GAcs",
  "username": "***",
  "password": "***",
  "instance_id": "***"
}
In [15]:
# The code was removed by Watson Studio for sharing.

Install the watson-machine-learning-client library from pypi.¶

In [16]:
!rm -rf $PIP_BUILD/watson-machine-learning-client
In [17]:
!pip install watson-machine-learning-client --upgrade

Requirement already up-to-date: watson-machine-learning-client in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages
Requirement not upgraded as not directly required: requests in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from watson-machine-learning-client)
Requirement not upgraded as not directly required: tqdm in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from watson-machine-learning-client)
Requirement not upgraded as not directly required: ibm-cos-sdk in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from watson-machine-learning-client)
Requirement not upgraded as not directly required: lomond in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from watson-machine-learning-client)
Requirement not upgraded as not directly required: urllib3 in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from watson-machine-learning-client)
Requirement not upgraded as not directly required: tabulate in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from watson-machine-learning-client)
Requirement not upgraded as not directly required: pandas in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from watson-machine-learning-client)
Requirement not upgraded as not directly required: certifi in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from watson-machine-learning-client)
Requirement not upgraded as not directly required: chardet<3.1.0,>=3.0.2 in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from requests->watson-machine-learning-client)
Requirement not upgraded as not directly required: idna<2.7,>=2.5 in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from requests->watson-machine-learning-client)
Requirement not upgraded as not directly required: ibm-cos-sdk-core==2.*,>=2.0.0 in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from ibm-cos-sdk->watson-machine-learning-client)
Requirement not upgraded as not directly required: ibm-cos-sdk-s3transfer==2.*,>=2.0.0 in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from ibm-cos-sdk->watson-machine-learning-client)
Requirement not upgraded as not directly required: six>=1.10.0 in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from lomond->watson-machine-learning-client)
Requirement not upgraded as not directly required: python-dateutil>=2 in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from pandas->watson-machine-learning-client)
Requirement not upgraded as not directly required: pytz>=2011k in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from pandas->watson-machine-learning-client)
Requirement not upgraded as not directly required: numpy>=1.9.0 in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from pandas->watson-machine-learning-client)
Requirement not upgraded as not directly required: jmespath<1.0.0,>=0.7.1 in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from ibm-cos-sdk-core==2.*,>=2.0.0->ibm-cos-sdk->watson-machine-learning-client)
Requirement not upgraded as not directly required: docutils>=0.10 in /opt/conda/envs/DSX-Python35/lib/python3.5/site-packages (from ibm-cos-sdk-core==2.*,>=2.0.0->ibm-cos-sdk->watson-machine-learning-client)

Import the watson-machine-learning-client and authenticate to the service instance.¶

In [18]:
from watson_machine_learning_client import WatsonMachineLearningAPIClient

2018-08-20 13:06:56,609 - watson_machine_learning_client.metanames - WARNING - 'AUTHOR_EMAIL' meta prop is deprecated. It will be ignored.
2018-08-20 13:06:59,394 - watson_machine_learning_client.metanames - WARNING - 'AUTHOR_EMAIL' meta prop is deprecated. It will be ignored.
2018-08-20 13:10:08,158 - watson_machine_learning_client.metanames - WARNING - 'AUTHOR_EMAIL' meta prop is deprecated. It will be ignored.

Note: A deprecation warning is returned from scikit-learn package that does not impact watson machine learning client functionalities.

In [19]:
client = WatsonMachineLearningAPIClient(wml_credentials)
In [20]:
# Display the client version number.
print(client.version)

1.0.302

Note: watson-machine-learning-client documentation can be found here.

2. Create the training definitions¶

In this section you:

  • 2.1 Prepare the training definition metadata
  • 2.2 Get the sample model definition content files from Git
  • 2.3 Store the training definition in the WML repository

2.1 Prepare the training definition metadata¶

In [21]:
model_definition_metadata = {
            client.repository.DefinitionMetaNames.NAME: "My definition name",
            client.repository.DefinitionMetaNames.DESCRIPTION: "My description",
            client.repository.DefinitionMetaNames.AUTHOR_NAME: "John Smith",
            client.repository.DefinitionMetaNames.FRAMEWORK_NAME: "tensorflow",
            client.repository.DefinitionMetaNames.FRAMEWORK_VERSION: "1.5",
            client.repository.DefinitionMetaNames.RUNTIME_NAME: "python",
            client.repository.DefinitionMetaNames.RUNTIME_VERSION: "3.5",
            client.repository.DefinitionMetaNames.EXECUTION_COMMAND: "python3 tensorflow_mnist_softmax.py --trainingIters 20"
            }

2.2 Get the sample model definition content file from Git ¶

In [22]:
filename='tf-softmax-model.zip'

if not os.path.isfile(filename):
    filename = wget.download('https://github.com/pmservice/wml-sample-models/raw/master/tensorflow/hand-written-digit-recognition/definition/tf-softmax-model.zip')
    
print(filename)

tf-softmax-model.zip

2.3 Store the training definition in the WML repository¶

In [23]:
definition_details = client.repository.store_definition(filename, model_definition_metadata)
In [24]:
definition_uid = client.repository.get_definition_uid(definition_details)
In [25]:
# Display the training definition uid.
print(definition_uid)

65481c9c-840b-46d7-8dfa-4376ef9c7f4a

3. Train the model¶

In this section, learn how to:

  • 3.1 Enter training configuration metadata
  • 3.2 Train the model in the background
  • 3.3 Monitor the training log
  • 3.4 Cancel the training run

3.1 Enter training configuration metadata¶

  • TRAINING_DATA_REFERENCE - references the uploaded training data.
  • TRAINING_RESULTS_REFERENCE - location where trained model will be saved.

Note Your COS credentials are referenced in this code.

In [26]:
# Configure the training metadata for the TRAINING_DATA_REFERENCE and TRAINING_RESULTS_REFERENCE.
training_configuration_metadata = {
            client.training.ConfigurationMetaNames.NAME: "Hand-written Digit Recognition", 
            client.training.ConfigurationMetaNames.AUTHOR_NAME: "John Smith",              
            client.training.ConfigurationMetaNames.DESCRIPTION: "Hand-written Digit Recognition training",
            client.training.ConfigurationMetaNames.COMPUTE_CONFIGURATION: {"name": "k80"},
            client.training.ConfigurationMetaNames.TRAINING_DATA_REFERENCE: {
                    "connection": {
                        "endpoint_url": service_endpoint,
                        "access_key_id": cos_credentials['cos_hmac_keys']['access_key_id'],
                        "secret_access_key": cos_credentials['cos_hmac_keys']['secret_access_key']
                    },
                    "source": {
                        "bucket": buckets[0],
                    },
                    "type": "s3"
                },
            client.training.ConfigurationMetaNames.TRAINING_RESULTS_REFERENCE: {
                "connection": {
                    "endpoint_url": service_endpoint,
                    "access_key_id": cos_credentials['cos_hmac_keys']['access_key_id'],
                    "secret_access_key": cos_credentials['cos_hmac_keys']['secret_access_key']
                },
                "target": {
                    "bucket": buckets[1],
                },
                "type": "s3"
            },
        }

3.2 Train the model in the background¶

To run the training in the background, set the optional parameter asynchronous=True (or remove it). In this case the parameter has been removed.

Note: To run the training in active mode, set asynchronous=False.

In [27]:
training_run_details = client.training.run(definition_uid, training_configuration_metadata)
In [28]:
training_run_guid_async = client.training.get_run_uid(training_run_details)

Check the status of the training run by calling the method the next cell:

In [29]:
# Get training run status.
status = client.training.get_status(training_run_guid_async)
print(json.dumps(status, indent=2))

{
  "submitted_at": "2018-08-20T13:07:03Z",
  "state": "pending",
  "current_at": "2018-08-20T13:07:04Z",
  "metrics": []
}

3.3 Monitor the training log¶

Run the cell below to monitor the training log.

In [30]:
client.training.monitor_logs(training_run_guid_async)

########################################################

Log monitor started for training run: training-C00cT2tiR

########################################################


Training with training/test data at:

  DATA_DIR: /mnt/data/training-data-eacf90dc-f69a-437c-8f43-5d36aa59abf4

  MODEL_DIR: /job/model-code

  TRAINING_JOB: 

  TRAINING_COMMAND: python3 tensorflow_mnist_softmax.py --trainingIters 20

Storing trained model at:

  RESULT_DIR: /mnt/results/training-results-eacf90dc-f69a-437c-8f43-5d36aa59abf4/training-C00cT2tiR

Mon Aug 20 13:08:12 UTC 2018: Running Tensorflow job

/usr/local/lib/python3.5/dist-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.

  from ._conv import register_converters as _register_converters

WARNING:tensorflow:From tensorflow_mnist_softmax.py:66: softmax_cross_entropy_with_logits (from tensorflow.python.ops.nn_ops) is deprecated and will be removed in a future version.

Instructions for updating:



Future major versions of TensorFlow will allow gradients to flow

into the labels input on backprop by default.



See tf.nn.softmax_cross_entropy_with_logits_v2.



2018-08-20 13:08:16.928487: I tensorflow/core/platform/cpu_feature_guard.cc:137] Your CPU supports instructions that this TensorFlow binary was not compiled to use: SSE4.1 SSE4.2 AVX AVX2 FMA

2018-08-20 13:08:17.106586: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1105] Found device 0 with properties: 

name: Tesla K80 major: 3 minor: 7 memoryClockRate(GHz): 0.8235

pciBusID: 0000:83:00.0

totalMemory: 11.17GiB freeMemory: 11.10GiB

2018-08-20 13:08:17.106632: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1195] Creating TensorFlow device (/device:GPU:0) -> (device: 0, name: Tesla K80, pci bus id: 0000:83:00.0, compute capability: 3.7)

Start model training

Optimization Finished!

Testing Accuracy:  0.9168

classification_signature content:

inputs {

  key: "inputs"

  value {

    name: "Placeholder:0"

    dtype: DT_FLOAT

    tensor_shape {

      dim {

        size: -1

      }

      dim {

        size: 784

      }

    }

  }

}

outputs {

  key: "classes"

  value {

    name: "predictor:0"

    dtype: DT_INT64

    tensor_shape {

      dim {

        size: -1

      }

    }

  }

}

method_name: "tensorflow/serving/classify"





-----------------
Log monitor done.
-----------------

3.4 Cancel the training run¶

You can cancel the training run by calling the method below.

client.training.cancel(training_run_guid_async)

4. Save the model¶

Call this method to the save model to the WML repository.¶

In [31]:
# Save model to WML repository.
saved_model_details = client.repository.store_model(training_run_guid_async, {"name": "My cool mnist model"})
In [32]:
# Display location of saved model.
print("Url: " + client.repository.get_model_url(saved_model_details))

Url: https://us-south.ml.cloud.ibm.com/v3/wml_instances/3f6e5c2b-4880-46aa-9d79-62e90ccc9d56/published_models/9039c349-b6e5-49ce-9212-bab6d5a7a72e

Tip: Run the code below to extract the stored model uid from the dictionary.

In [33]:
# Get and display model uid.
model_uid = client.repository.get_model_uid(saved_model_details)
print("Saved model uid: " + model_uid)

Saved model uid: 9039c349-b6e5-49ce-9212-bab6d5a7a72e

List the stored models.¶

In [34]:
client.repository.list_models()

------------------------------------  ------------------------------------  ------------------------  -----------------
GUID                                  NAME                                  CREATED                   FRAMEWORK
9039c349-b6e5-49ce-9212-bab6d5a7a72e  My cool mnist model                   2018-08-20T13:10:08.469Z  tensorflow-1.5
45180921-025b-4881-9ef0-2cff85c716d6  MNIST best model                      2018-08-20T12:09:12.954Z  tensorflow-1.5
9982062e-f598-4cc6-b6a6-fb11c0427aba  MNIST best model                      2018-08-08T17:13:28.977Z  tensorflow-1.5
19e95da9-520f-49bc-9e0c-227a5e74e130  Advisor productivity model            2018-06-13T18:02:54.960Z  mllib-2.1
cd7602c8-6630-4c89-8438-2b41314c1af8  Advisor productivity model            2018-06-13T17:19:08.751Z  mllib-2.1
3dddfb49-51de-471e-b9df-cf7174630845  Advisor productivity model            2018-06-13T16:07:48.140Z  mllib-2.1
84a49f46-8e4e-41e4-958b-7e3073820d62  Advisor productivity model            2018-06-13T04:13:06.731Z  mllib-2.1
fe00574f-95a2-4dd3-b29f-30d071734662  Advisor productivity model            2018-06-12T21:07:23.922Z  mllib-2.1
82d124ab-f174-424b-bf93-8b4240df2875  Advisor productivity model            2018-06-12T18:11:49.350Z  mllib-2.1
0498a7d4-56fd-43bf-92ef-1bebd6b057ef  Advisor productivity model            2018-06-12T16:30:22.586Z  mllib-2.1
1ab43c12-4313-46e9-8a1c-09bdb8004af4  Product line model                    2018-06-12T07:56:20.689Z  mllib-2.1
414fd52b-74a3-4d5e-ab5f-1288a0df83db  Product line model                    2018-06-11T08:19:51.753Z  mllib-2.1
eec29270-e1f1-416a-9dce-9cf794f083c3  Product line model                    2018-06-09T11:34:21.889Z  mllib-2.1
3dffd8f3-e1e3-43f9-9497-fe2d5714ff8d  Product line model                    2018-06-05T17:27:29.780Z  mllib-2.1
2ff256ec-1e40-45a5-8aba-7d83f2087ae7  Product line model                    2018-05-31T20:09:26.907Z  mllib-2.1
6ad5ae64-25ba-402d-8f54-384b4f9b9d08  Product line model                    2018-05-30T11:21:37.085Z  mllib-2.1
23f06316-1c4e-492a-9760-e0f8ccdbac33  Product line model                    2018-05-30T06:13:21.145Z  mllib-2.1
2c065114-e8ca-4f5a-819f-b05ca8dbe1f0  Product line model                    2018-05-29T11:52:52.012Z  mllib-2.1
6e589f3c-4b70-4816-a47a-4a345cc81024  Product line model                    2018-05-29T11:29:25.962Z  mllib-2.1
e963442c-5445-4fd8-bcaa-a7f8590f1eb3  Product line model                    2018-05-24T15:46:27.617Z  mllib-2.1
e6bc671d-7731-4149-ae28-62be0d09ac93  MNIST best model                      2018-05-24T12:59:05.837Z  tensorflow-1.5
267e805c-5639-435b-b4de-8c7be5e12700  Core ML - keras mnist model           2018-04-24T14:36:39.346Z  tensorflow-1.5
b7bcb71e-302b-4087-bcde-3be6e4188e03  Core ML - keras mnist model           2018-04-23T13:55:25.119Z  tensorflow-1.5
2e1434e3-21dd-4f38-b227-cf113ea9d6c7  Core ML - keras mnist model           2018-04-23T13:42:59.695Z  tensorflow-1.5
ec476762-d41b-4483-8e92-c13f3df9b356  Core ML - keras mnist model           2018-04-23T13:36:47.517Z  tensorflow-1.5
49f64c52-3cc3-47d9-bca5-ec9aea294c74  Core ML - keras mnist model           2018-04-23T13:33:14.643Z  tensorflow-1.5
b790fb08-a7a3-4c4b-9639-494e8038d924  Credit Card Fraud Detection Scikit    2018-04-20T09:21:56.225Z  scikit-learn-0.19
6e7aaadd-b86f-4751-b0ee-a74f38938b10  Fraud Detection - Credit Card (SPSS)  2018-04-09T18:40:01.883Z  spss-modeler-18.1
51aa4ef2-81e7-44bb-854d-1e459ac9f7ca  Credit Card Fraud Detection           2018-04-06T12:17:29.820Z  scikit-learn-0.19
9faf9502-87ca-4a15-8581-060cc84ccc5b  Sentiment Prediction Model            2018-03-29T10:37:16.069Z  mllib-2.1
70a8b8ad-ef04-4d9f-a1e7-679719577b62  Product line model                    2018-03-29T09:21:46.826Z  mllib-2.1
298832f7-d21d-4a07-a028-b5716dba1264  Best Heart Drug Selection             2018-03-29T08:35:31.303Z  mllib-2.1
0da45aa2-5378-4dc1-8617-3d209499f466  my spark model                        2018-03-28T14:49:01.055Z  mllib-2.1
9fb96bc1-bc12-4717-a03f-9b009a32d578  SPSS model for Churn prediction       2018-03-28T13:20:40.791Z  spss-modeler-18.1
5bcf67ab-cd88-49c9-99fd-99dde60eff70  SPSS model for Churn prediction       2018-03-28T13:11:28.164Z  spss-modeler-17.1
91149430-5588-4ee6-ac6a-695cd2392aa8  My cool mnist model tf-ddl            2018-03-28T11:15:59.813Z  tensorflow-1.5
b6e59ce5-7d24-4ce1-90f8-f6a918ad71ca  My cool mnist model tf-distributed    2018-03-28T11:15:56.061Z  tensorflow-1.5
84070d90-6db7-4a12-8da6-26aaf18e9378  My cool mnist model tf-horovod        2018-03-28T11:15:53.446Z  tensorflow-1.5
97822727-4bd1-4df0-95df-8ff9536ea01d  My cool mnist model                   2018-03-28T09:02:31.509Z  tensorflow-1.5
a0cb79ca-e937-4cbf-b7e5-23dad2d2ecec  drug-selection                        2018-03-16T18:10:25.046Z  mllib-2.0
327242a2-d2c0-400f-b673-4f33ad7ec17d  product-line-prediction               2018-03-16T18:08:31.260Z  mllib-2.0
f7219c8f-fccb-4d26-a1bd-f51544090eae  training-S3ttnlgig_2 best model       2018-03-16T13:32:33.747Z  tensorflow-1.5
861b6b73-6861-49c0-867c-d171d6d9d545  MNIST best model                      2018-03-16T13:00:40.461Z  tensorflow-1.5
c755c740-dab6-4d5d-b873-1060885d55d5  MNIST best model                      2018-03-16T10:12:31.990Z  tensorflow-1.5
a5a952e1-fa6f-4c64-88f7-293b367b7e1f  MNIST best model                      2018-03-16T10:12:16.773Z  tensorflow-1.5
43986a55-0f0e-493b-9b98-0712f918b0ba  MNIST CNN model                       2018-03-15T16:12:06.503Z  tensorflow-1.5
------------------------------------  ------------------------------------  ------------------------  -----------------

Delete the training run and its results.¶

Run the code in the cell below to delete train runs and results:

client.training.delete(training_run_uid_async)

5. Create the online deployment¶

Run the code in the following cell to deploy the stored model as a web service (online):

In [35]:
# Create online deployment.
deployment_details = client.deployments.create(model_uid, "Mnist model deployment")

#######################################################################################

Synchronous deployment creation for uid: '9039c349-b6e5-49ce-9212-bab6d5a7a72e' started

#######################################################################################


INITIALIZING
DEPLOY_SUCCESS


------------------------------------------------------------------------------------------------
Successfully finished deployment creation, deployment_uid='0befc8fd-3f16-4e8d-9f14-3b890be4b4c8'
------------------------------------------------------------------------------------------------

6. Score data¶

Now, extract the url endpoint, scoring_url, which will be used to send scoring requests.

In [36]:
#  Extract endpoint url and display it.
scoring_url = client.deployments.get_scoring_url(deployment_details)
print(scoring_url)

https://ibm-watson-ml.mybluemix.net/v3/wml_instances/3f6e5c2b-4880-46aa-9d79-62e90ccc9d56/deployments/0befc8fd-3f16-4e8d-9f14-3b890be4b4c8/online

Prepare sample scoring data to score deployed model.

In [37]:
image1 = [0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.639,0.757,0.596,0.361,0.200,0.200,0.200,0.200,0.122,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.082,0.875,0.992,0.988,0.992,0.988,0.992,0.988,0.992,0.753,0.322,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.200,0.992,0.400,0.000,0.082,0.400,0.243,0.400,0.400,0.239,0.718,0.157,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.200,0.988,0.400,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.239,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.200,0.992,0.400,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.361,0.988,0.400,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.678,0.992,0.400,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.992,0.988,0.875,0.796,0.796,0.796,0.322,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.482,0.996,0.992,0.878,0.796,0.796,0.875,0.996,0.278,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.161,0.953,0.992,0.510,0.078,0.000,0.000,0.078,0.992,0.910,0.161,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.596,0.992,0.718,0.000,0.000,0.000,0.000,0.000,0.518,0.992,0.400,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.200,0.592,0.078,0.000,0.000,0.000,0.000,0.000,0.200,0.988,0.400,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.082,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.443,0.992,0.400,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.322,0.718,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.757,0.988,0.400,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.796,0.718,0.000,0.000,0.000,0.000,0.000,0.000,0.082,0.996,0.992,0.161,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.082,0.875,0.400,0.000,0.000,0.000,0.000,0.000,0.082,0.796,0.992,0.510,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.800,0.482,0.000,0.000,0.000,0.000,0.161,0.678,0.992,0.796,0.078,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.635,0.875,0.400,0.082,0.400,0.639,0.953,0.992,0.671,0.078,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.161,0.918,0.992,1.000,0.992,1.000,0.675,0.322,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.118,0.514,0.753,0.435,0.196,0.039,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000]
image2 = [0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.196,0.878,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.275,0.114,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.475,0.906,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.580,0.659,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.016,0.765,0.906,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.376,0.824,0.043,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.271,0.988,0.525,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.447,0.988,0.082,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.176,0.925,0.851,0.047,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.753,0.988,0.082,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.659,0.969,0.208,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.071,1.000,0.992,0.082,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.329,0.949,0.827,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.553,0.992,0.741,0.020,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.663,0.988,0.416,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.125,0.910,0.980,0.259,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.059,0.882,0.988,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.525,0.988,0.827,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.086,0.988,0.643,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.663,0.988,0.655,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.035,0.800,0.820,0.071,0.000,0.000,0.000,0.000,0.000,0.000,0.086,0.992,0.992,0.420,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.663,0.988,0.780,0.333,0.333,0.333,0.333,0.506,0.643,0.765,0.988,0.988,0.416,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.161,0.667,0.961,0.988,0.988,0.988,0.988,0.910,0.906,0.984,0.988,0.988,0.035,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.192,0.329,0.329,0.329,0.329,0.000,0.000,0.631,0.988,0.988,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.498,0.988,0.988,0.176,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.502,0.992,0.992,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.498,0.988,0.988,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.529,0.988,0.957,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.910,0.925,0.435,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.702,0.259,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000,0.000]
In [38]:
%matplotlib inline
import matplotlib.pyplot as plt
import numpy as np
In [39]:
for i, image in enumerate([image1, image2]):
    plt.subplot(2, 2, i + 1)
    plt.axis('off')
    plt.imshow( (np.reshape(image, (28, 28)) * 255).astype(np.uint8), cmap=plt.cm.gray_r, interpolation='nearest')

Build a scoring dictionary consisting of two digits and send it to the deployed model to get predictions.

In [40]:
# Build scoring dictionary which has two digits (image1 and image29.
scoring_data = {'values': [image1, image2]}
In [41]:
# send scoring dictionary to deployed model to get predictions
predictions = client.deployments.score(scoring_url, scoring_data)
print("Scoring result: " + str(predictions))

Scoring result: {'values': [5, 4]}

6. Summary and next steps¶

You successfully completed this notebook!

You learned how to use watson-machine-learning-client to train and score TensorFlow models.

Check out our Online Documentation for a tutorial and more samples, documentation, how-tos, and blog posts.

Citations¶

Y. LeCun, L. Bottou, Y. Bengio, and P. Haffner. "Gradient-based learning applied to document recognition." Proceedings of the IEEE, 86(11):2278-2324, November 1998.

Authors¶

Lukasz Cmielowski, PhD, is a Automation Architect and Data Scientist at IBM with a track record of developing enterprise-level applications that substantially increase the clients' ability to turn data into actionable knowledge.

Copyright © 2017, 2018 IBM. This notebook and its source code are released under the terms of the MIT License.

Love this notebook? Don't have an account yet?
Share it with your colleagues and help them discover the power of Watson Studio! Sign Up