CODEWITHSUNDEEP
pythontensorflowmachine-learninggoogle-cloud-platformgoogle-cloud-ml
Jennings Mc Enroe
Jennings Mc Enroe

Reputation: 21

ML Engine: Prediction Error while executing local predict command

I have uploaded a version of the model in the Google ML Engine with saved_model.pb and a variables folder. When I try to execute the command:

gcloud ml-engine local predict --model-dir=saved_model --json-instances=request.json

It shows the following error:

ERROR: (gcloud.ml-engine.local.predict) 2018-09-11 19:06:39.770396: I tensorflow/core/platform/cpu_feature_guard.cc:141] 
Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA

Traceback (most recent call last):
  File "lib/googlecloudsdk/command_lib/ml_engine/local_predict.py", line 172, in <module>
    main()
  File "lib/googlecloudsdk/command_lib/ml_engine/local_predict.py", line 167, in main
    signature_name=args.signature_name)
  File "/usr/lib/google-cloud-sdk/lib/third_party/ml_sdk/cloud/ml/prediction/prediction_lib.py", line 106, in local_predict
    predictions = model.predict(instances, signature_name=signature_name)
  File "/usr/lib/google-cloud-sdk/lib/third_party/ml_sdk/cloud/ml/prediction/prediction_utils.py", line 230, in predict
    preprocessed = self.preprocess(instances, stats=stats, **kwargs)
  File "/usr/lib/google-cloud-sdk/lib/third_party/ml_sdk/cloud/ml/prediction/frameworks/tf_prediction_lib.py", line 436, in preprocess
    preprocessed = self._canonicalize_input(instances, signature)
  File "/usr/lib/google-cloud-sdk/lib/third_party/ml_sdk/cloud/ml/prediction/frameworks/tf_prediction_lib.py", line 453, in _canonicalize_input
    return canonicalize_single_tensor_input(instances, tensor_name)
  File "/usr/lib/google-cloud-sdk/lib/third_party/ml_sdk/cloud/ml/prediction/frameworks/tf_prediction_lib.py", line 166, in canonicalize_single_tensor_input
    instances = [parse_single_tensor(x, tensor_name) for x in instances]
  File "/usr/lib/google-cloud-sdk/lib/third_party/ml_sdk/cloud/ml/prediction/frameworks/tf_prediction_lib.py", line 162, in parse_single_tensor
    (tensor_name, list(x.keys())))


cloud.ml.prediction.prediction_utils.PredictionError: Invalid inputs: Expected tensor name: inputs, got tensor name: [u'inputs', u'key']. (Error code: 1)

My request.json file is

{"inputs": {"b64": "/9j/4AAQSkZJRgABAQAAAQABAAD/2wBDAAgGBgcGBQgHBwcJCQgKDBQNDAsLDBkSEw8UHRofHh0aHBwgJC4nICIsIxwcKDcpLDAxNDQ0Hyc5PTgyPC4zNDL/2wBDAQkJCQwLDBgNDRgyIRwhMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjL/wAARCAHVArwDASIAAhEBAxEB/8QAHwAAAQUBAQEBA....."}, "key": "841bananas.jpg"}

Thanks in advance.

Upvotes: 1

Views: 564

Answers (1)

rhaertel80
rhaertel80

Reputation: 8399

It appears your model was exported with only one input named "inputs". In that case, you shouldn't be sending "key" in the JSON, i.e., (scroll to the end to see I've removed "keys"):

{"inputs": {"b64": "/9j/4AAQSkZJRgABAQAAAQABAAD/2wBDAAgGBgcGBQgHBwcJCQgKDBQNDAsLDBkSEw8UHRofHh0aHBwgJC4nICIsIxwcKDcpLDAxNDQ0Hyc5PTgyPC4zNDL/2wBDAQkJCQwLDBgNDRgyIRwhMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjL/wAARCAHVArwDASIAAhEBAxEB/8QAHwAAAQUBAQEBA....."}}

Upvotes: 1

Related Questions