Class: Google::Cloud::AutoML::V1beta1::PredictionServiceClient

Inherits:
Object
  • Object
show all
Defined in:
lib/google/cloud/automl/v1beta1/prediction_service_client.rb

Overview

AutoML Prediction API.

On any input that is documented to expect a string parameter in snake_case or kebab-case, either of those cases is accepted.

Defined Under Namespace

Classes: OperationsClient

Constant Summary collapse

SERVICE_ADDRESS =

The default address of the service.

"automl.googleapis.com".freeze
DEFAULT_SERVICE_PORT =

The default port of the service.

443
GRPC_INTERCEPTORS =

The default set of gRPC interceptors.

[]
DEFAULT_TIMEOUT = 
30
ALL_SCOPES =

The scopes needed to make gRPC calls to all of the methods defined in this service.

[
  "https://www.googleapis.com/auth/cloud-platform"
].freeze

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(credentials: nil, scopes: ALL_SCOPES, client_config: {}, timeout: DEFAULT_TIMEOUT, metadata: nil, service_address: nil, service_port: nil, exception_transformer: nil, lib_name: nil, lib_version: "") ⇒ PredictionServiceClient

Returns a new instance of PredictionServiceClient.

Parameters:

  • credentials (Google::Auth::Credentials, String, Hash, GRPC::Core::Channel, GRPC::Core::ChannelCredentials, Proc) (defaults to: nil) —

    Provides the means for authenticating requests made by the client. This parameter can be many types. A Google::Auth::Credentials uses a the properties of its represented keyfile for authenticating requests made by this client. A String will be treated as the path to the keyfile to be used for the construction of credentials for this client. A Hash will be treated as the contents of a keyfile to be used for the construction of credentials for this client. A GRPC::Core::Channel will be used to make calls through. A GRPC::Core::ChannelCredentials for the setting up the RPC client. The channel credentials should already be composed with a GRPC::Core::CallCredentials object. A Proc will be used as an updater_proc for the Grpc channel. The proc transforms the metadata for requests, generally, to give OAuth credentials.

  • scopes (Array<String>) (defaults to: ALL_SCOPES) —

    The OAuth scopes for this service. This parameter is ignored if an updater_proc is supplied.

  • client_config (Hash) (defaults to: {}) —

    A Hash for call options for each method. See Google::Gax#construct_settings for the structure of this data. Falls back to the default config if not specified or the specified config is missing data points.

  • timeout (Numeric) (defaults to: DEFAULT_TIMEOUT) —

    The default timeout, in seconds, for calls made through this client.

  • metadata (Hash) (defaults to: nil) —

    Default metadata to be sent with each request. This can be overridden on a per call basis.

  • service_address (String) (defaults to: nil) —

    Override for the service hostname, or nil to leave as the default.

  • service_port (Integer) (defaults to: nil) —

    Override for the service port, or nil to leave as the default.

  • exception_transformer (Proc) (defaults to: nil) —

    An optional proc that intercepts any exceptions raised during an API call to inject custom error handling.



123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
 
# File 'lib/google/cloud/automl/v1beta1/prediction_service_client.rb', line 123

def initialize \
    credentials: nil,
    scopes: ALL_SCOPES,
    client_config: {},
    timeout: DEFAULT_TIMEOUT,
    metadata: nil,
    service_address: nil,
    service_port: nil,
    exception_transformer: nil,
    lib_name: nil,
    lib_version: ""
  # These require statements are intentionally placed here to initialize
  # the gRPC module only when it's required.
  # See https://github.com/googleapis/toolkit/issues/446
  require "google/gax/grpc"
  require "google/cloud/automl/v1beta1/prediction_service_services_pb"

  credentials ||= Google::Cloud::AutoML::V1beta1::Credentials.default

  @operations_client = OperationsClient.new(
    credentials: credentials,
    scopes: scopes,
    client_config: client_config,
    timeout: timeout,
    lib_name: lib_name,
    service_address: service_address,
    service_port: service_port,
    lib_version: lib_version,
    metadata: metadata,
  )

  if credentials.is_a?(String) || credentials.is_a?(Hash)
    updater_proc = Google::Cloud::AutoML::V1beta1::Credentials.new(credentials).updater_proc
  end
  if credentials.is_a?(GRPC::Core::Channel)
    channel = credentials
  end
  if credentials.is_a?(GRPC::Core::ChannelCredentials)
    chan_creds = credentials
  end
  if credentials.is_a?(Proc)
    updater_proc = credentials
  end
  if credentials.is_a?(Google::Auth::Credentials)
    updater_proc = credentials.updater_proc
  end

  package_version = Google::Cloud::AutoML::VERSION

  google_api_client = "gl-ruby/#{RUBY_VERSION}"
  google_api_client << " #{lib_name}/#{lib_version}" if lib_name
  google_api_client << " gapic/#{package_version} gax/#{Google::Gax::VERSION}"
  google_api_client << " grpc/#{GRPC::VERSION}"
  google_api_client.freeze

  headers = { :"x-goog-api-client" => google_api_client }
  headers.merge!(metadata) unless metadata.nil?
  client_config_file = Pathname.new(__dir__).join(
    "prediction_service_client_config.json"
  )
  defaults = client_config_file.open do |f|
    Google::Gax.construct_settings(
      "google.cloud.automl.v1beta1.PredictionService",
      JSON.parse(f.read),
      client_config,
      Google::Gax::Grpc::STATUS_CODE_NAMES,
      timeout,
      errors: Google::Gax::Grpc::API_ERRORS,
      metadata: headers
    )
  end

  # Allow overriding the service path/port in subclasses.
  service_path = service_address || self.class::SERVICE_ADDRESS
  port = service_port || self.class::DEFAULT_SERVICE_PORT
  interceptors = self.class::GRPC_INTERCEPTORS
  @prediction_service_stub = Google::Gax::Grpc.create_stub(
    service_path,
    port,
    chan_creds: chan_creds,
    channel: channel,
    updater_proc: updater_proc,
    scopes: scopes,
    interceptors: interceptors,
    &Google::Cloud::AutoML::V1beta1::PredictionService::Stub.method(:new)
  )

  @predict = Google::Gax.create_api_call(
    @prediction_service_stub.method(:predict),
    defaults["predict"],
    exception_transformer: exception_transformer,
    params_extractor: proc do |request|
      {'name' => request.name}
    end
  )
  @batch_predict = Google::Gax.create_api_call(
    @prediction_service_stub.method(:batch_predict),
    defaults["batch_predict"],
    exception_transformer: exception_transformer,
    params_extractor: proc do |request|
      {'name' => request.name}
    end
  )
end

Class Method Details

.model_path(project, location, model) ⇒ String

Returns a fully-qualified model resource name string.

Parameters:

  • project (String)
  • location (String)
  • model (String)

Returns:

  • (String) 


82
83
84
85
86
87
88
 
# File 'lib/google/cloud/automl/v1beta1/prediction_service_client.rb', line 82

def self.model_path project, location, model
  MODEL_PATH_TEMPLATE.render(
    :"project" => project,
    :"location" => location,
    :"model" => model
  )
end

Instance Method Details

#batch_predict(name, input_config, output_config, params: nil, options: nil) ⇒ Google::Gax::Operation

Perform a batch prediction. Unlike the online Predict, batch prediction result won't be immediately available in the response. Instead, a long running operation object is returned. User can poll the operation result via GetOperation method. Once the operation is done, BatchPredictResult is returned in the response field. Available for following ML problems:

  • Image Classification
  • Image Object Detection
  • Video Classification
  • Video Object Tracking * Text Extraction
  • Tables

Examples:

require "google/cloud/automl"

prediction_client = Google::Cloud::AutoML::Prediction.new(version: :v1beta1)
formatted_name = Google::Cloud::AutoML::V1beta1::PredictionServiceClient.model_path("[PROJECT]", "[LOCATION]", "[MODEL]")

# TODO: Initialize `input_config`:
input_config = {}

# TODO: Initialize `output_config`:
output_config = {}

# Register a callback during the method call.
operation = prediction_client.batch_predict(formatted_name, input_config, output_config) do |op|
  raise op.results.message if op.error?
  op_results = op.results
  # Process the results.

  metadata = op.metadata
  # Process the metadata.
end

# Or use the return value to register a callback.
operation.on_done do |op|
  raise op.results.message if op.error?
  op_results = op.results
  # Process the results.

  metadata = op.metadata
  # Process the metadata.
end

# Manually reload the operation.
operation.reload!

# Or block until the operation completes, triggering callbacks on
# completion.
operation.wait_until_done!

Parameters:

  • name (String) —

    Name of the model requested to serve the batch prediction.

  • input_config (Google::Cloud::AutoML::V1beta1::BatchPredictInputConfig | Hash) —

    Required. The input configuration for batch prediction. A hash of the same form as Google::Cloud::AutoML::V1beta1::BatchPredictInputConfig can also be provided.

  • output_config (Google::Cloud::AutoML::V1beta1::BatchPredictOutputConfig | Hash) —

    Required. The Configuration specifying where output predictions should be written. A hash of the same form as Google::Cloud::AutoML::V1beta1::BatchPredictOutputConfig can also be provided.

  • params (Hash{String => String}) (defaults to: nil) —

    Additional domain-specific parameters for the predictions, any string must be up to 25000 characters long.

    • For Text Classification:

    score_threshold - (float) A value from 0.0 to 1.0. When the model makes predictions for a text snippet, it will only produce results that have at least this confidence score. The default is 0.5.

    • For Image Classification:

    score_threshold - (float) A value from 0.0 to 1.0. When the model makes predictions for an image, it will only produce results that have at least this confidence score. The default is 0.5.

    • For Image Object Detection:

    score_threshold - (float) When Model detects objects on the image, it will only produce bounding boxes which have at least this confidence score. Value in 0 to 1 range, default is 0.5. max_bounding_box_count - (int64) No more than this number of bounding boxes will be produced per image. Default is 100, the requested value may be limited by server.

    • For Video Classification : score_threshold - (float) A value from 0.0 to 1.0. When the model makes predictions for a video, it will only produce results that have at least this confidence score. The default is 0.5. segment_classification - (boolean) Set to true to request segment-level classification. AutoML Video Intelligence returns labels and their confidence scores for the entire segment of the video that user specified in the request configuration. The default is "true". shot_classification - (boolean) Set to true to request shot-level classification. AutoML Video Intelligence determines the boundaries for each camera shot in the entire segment of the video that user specified in the request configuration. AutoML Video Intelligence then returns labels and their confidence scores for each detected shot, along with the start and end time of the shot. WARNING: Model evaluation is not done for this classification type, the quality of it depends on training data, but there are no metrics provided to describe that quality. The default is "false". 1s_interval_classification - (boolean) Set to true to request classification for a video at one-second intervals. AutoML Video Intelligence returns labels and their confidence scores for each second of the entire segment of the video that user specified in the request configuration. WARNING: Model evaluation is not done for this classification type, the quality of it depends on training data, but there are no metrics provided to describe that quality. The default is "false".

    • For Video Object Tracking: score_threshold - (float) When Model detects objects on video frames, it will only produce bounding boxes which have at least this confidence score. Value in 0 to 1 range, default is 0.5. max_bounding_box_count - (int64) No more than this number of bounding boxes will be returned per frame. Default is 100, the requested value may be limited by server. min_bounding_box_size - (float) Only bounding boxes with shortest edge at least that long as a relative value of video frame size will be returned. Value in 0 to 1 range. Default is 0.

  • options (Google::Gax::CallOptions) (defaults to: nil) —

    Overrides the default settings for this call, e.g, timeout, retries, etc.

Returns:

  • (Google::Gax::Operation)

Raises:

  • (Google::Gax::GaxError) —

    if the RPC is aborted.



448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
 
# File 'lib/google/cloud/automl/v1beta1/prediction_service_client.rb', line 448

def batch_predict \
    name,
    input_config,
    output_config,
    params: nil,
    options: nil
  req = {
    name: name,
    input_config: input_config,
    output_config: output_config,
    params: params
  }.delete_if { |_, v| v.nil? }
  req = Google::Gax::to_proto(req, Google::Cloud::AutoML::V1beta1::BatchPredictRequest)
  operation = Google::Gax::Operation.new(
    @batch_predict.call(req, options),
    @operations_client,
    Google::Cloud::AutoML::V1beta1::BatchPredictResult,
    Google::Cloud::AutoML::V1beta1::OperationMetadata,
    call_options: options
  )
  operation.on_done { |operation| yield(operation) } if block_given?
  operation
end

#predict(name, payload, params: nil, options: nil) {|result, operation| ... } ⇒ Google::Cloud::AutoML::V1beta1::PredictResponse

Perform an online prediction. The prediction result will be directly returned in the response. Available for following ML problems, and their expected request payloads:

  • Image Classification - Image in .JPEG, .GIF or .PNG format, image_bytes up to 30MB.
  • Image Object Detection - Image in .JPEG, .GIF or .PNG format, image_bytes up to 30MB.
  • Text Classification - TextSnippet, content up to 60,000 characters, UTF-8 encoded.
  • Text Extraction - TextSnippet, content up to 30,000 characters, UTF-8 NFC encoded.
  • Translation - TextSnippet, content up to 25,000 characters, UTF-8 encoded.
  • Tables - Row, with column values matching the columns of the model, up to 5MB. Not available for FORECASTING

prediction_type.

  • Text Sentiment - TextSnippet, content up 500 characters, UTF-8 encoded.

Examples:

require "google/cloud/automl"

prediction_client = Google::Cloud::AutoML::Prediction.new(version: :v1beta1)
formatted_name = Google::Cloud::AutoML::V1beta1::PredictionServiceClient.model_path("[PROJECT]", "[LOCATION]", "[MODEL]")

# TODO: Initialize `payload`:
payload = {}
response = prediction_client.predict(formatted_name, payload)

Parameters:

  • name (String) —

    Name of the model requested to serve the prediction.

  • payload (Google::Cloud::AutoML::V1beta1::ExamplePayload | Hash) —

    Required. Payload to perform a prediction on. The payload must match the problem type that the model was trained to solve. A hash of the same form as Google::Cloud::AutoML::V1beta1::ExamplePayload can also be provided.

  • params (Hash{String => String}) (defaults to: nil) —

    Additional domain-specific parameters, any string must be up to 25000 characters long.

    • For Image Classification:

    score_threshold - (float) A value from 0.0 to 1.0. When the model makes predictions for an image, it will only produce results that have at least this confidence score. The default is 0.5.

    • For Image Object Detection: score_threshold - (float) When Model detects objects on the image, it will only produce bounding boxes which have at least this confidence score. Value in 0 to 1 range, default is 0.5. max_bounding_box_count - (int64) No more than this number of bounding boxes will be returned in the response. Default is 100, the requested value may be limited by server.
    • For Tables: feature_importance - (boolean) Whether

    [feature_importance][[google.cloud.automl.v1beta1.TablesModelColumnInfo.feature_importance] should be populated in the returned

    [TablesAnnotation(-s)][[google.cloud.automl.v1beta1.TablesAnnotation]. The default is false.

  • options (Google::Gax::CallOptions) (defaults to: nil) —

    Overrides the default settings for this call, e.g, timeout, retries, etc.

Yields:

  • (result, operation) —

    Access the result along with the RPC operation

Yield Parameters:

Returns:

Raises:

  • (Google::Gax::GaxError) —

    if the RPC is aborted.



300
301
302
303
304
305
306
307
308
309
310
311
312
313
 
# File 'lib/google/cloud/automl/v1beta1/prediction_service_client.rb', line 300

def predict \
    name,
    payload,
    params: nil,
    options: nil,
    &block
  req = {
    name: name,
    payload: payload,
    params: params
  }.delete_if { |_, v| v.nil? }
  req = Google::Gax::to_proto(req, Google::Cloud::AutoML::V1beta1::PredictRequest)
  @predict.call(req, options, &block)
end