Gitiles
Code Review Sign In
gerrit-public.fairphone.software / platform / external / python / google-api-python-client / c30d2b5d4478b230246dcaaa9f1c533ca3304387 / . / docs / dyn / ml_v1beta1.projects.jobs.html
blob: 57245ca8fb8ba342fbc02d0915d77220eedfa2c5 [file] [log] [blame]
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]1<html><body>
2<style>
3
4body, h1, h2, h3, div, span, p, pre, a {
5 margin: 0;
6 padding: 0;
7 border: 0;
8 font-weight: inherit;
9 font-style: inherit;
10 font-size: 100%;
11 font-family: inherit;
12 vertical-align: baseline;
13}
14
15body {
16 font-size: 13px;
17 padding: 1em;
18}
19
20h1 {
21 font-size: 26px;
22 margin-bottom: 1em;
23}
24
25h2 {
26 font-size: 24px;
27 margin-bottom: 1em;
28}
29
30h3 {
31 font-size: 20px;
32 margin-bottom: 1em;
33 margin-top: 1em;
34}
35
36pre, code {
37 line-height: 1.5;
38 font-family: Monaco, 'DejaVu Sans Mono', 'Bitstream Vera Sans Mono', 'Lucida Console', monospace;
39}
40
41pre {
42 margin-top: 0.5em;
43}
44
45h1, h2, h3, p {
46 font-family: Arial, sans serif;
47}
48
49h1, h2, h3 {
50 border-bottom: solid #CCC 1px;
51}
52
53.toc_element {
54 margin-top: 0.5em;
55}
56
57.firstline {
58 margin-left: 2 em;
59}
60
61.method {
62 margin-top: 1em;
63 border: solid 1px #CCC;
64 padding: 1em;
65 background: #EEE;
66}
67
68.details {
69 font-weight: bold;
70 font-size: 14px;
71}
72
73</style>
74
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]75<h1><a href="ml_v1beta1.html">Google Cloud Machine Learning Engine</a> . <a href="ml_v1beta1.projects.html">projects</a> . <a href="ml_v1beta1.projects.jobs.html">jobs</a></h1>
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]76<h2>Instance Methods</h2>
77<p class="toc_element">
78 <code><a href="#cancel">cancel(name=None, body, x__xgafv=None)</a></code></p>
79<p class="firstline">Cancels a running job.</p>
80<p class="toc_element">
81 <code><a href="#create">create(parent=None, body, x__xgafv=None)</a></code></p>
82<p class="firstline">Creates a training or a batch prediction job.</p>
83<p class="toc_element">
84 <code><a href="#get">get(name=None, x__xgafv=None)</a></code></p>
85<p class="firstline">Describes a job.</p>
86<p class="toc_element">
87 <code><a href="#list">list(parent=None, pageSize=None, filter=None, pageToken=None, x__xgafv=None)</a></code></p>
88<p class="firstline">Lists the jobs in the project.</p>
89<p class="toc_element">
90 <code><a href="#list_next">list_next(previous_request, previous_response)</a></code></p>
91<p class="firstline">Retrieves the next page of results.</p>
92<h3>Method Details</h3>
93<div class="method">
94 <code class="details" id="cancel">cancel(name=None, body, x__xgafv=None)</code>
95 <pre>Cancels a running job.
96
97Args:
98 name: string, Required. The name of the job to cancel.
99
100Authorization: requires `Editor` role on the parent project. (required)
101 body: object, The request body. (required)
102 The object takes the form of:
103
104{ # Request message for the CancelJob method.
105 }
106
107 x__xgafv: string, V1 error format.
108 Allowed values
109 1 - v1 error format
110 2 - v2 error format
111
112Returns:
113 An object of the form:
114
115 { # A generic empty message that you can re-use to avoid defining duplicated
116 # empty messages in your APIs. A typical example is to use it as the request
117 # or the response type of an API method. For instance:
118 #
119 # service Foo {
120 # rpc Bar(google.protobuf.Empty) returns (google.protobuf.Empty);
121 # }
122 #
123 # The JSON representation for `Empty` is empty JSON object `{}`.
124 }</pre>
125</div>
126
127<div class="method">
128 <code class="details" id="create">create(parent=None, body, x__xgafv=None)</code>
129 <pre>Creates a training or a batch prediction job.
130
131Args:
132 parent: string, Required. The project name.
133
134Authorization: requires `Editor` role on the specified project. (required)
135 body: object, The request body. (required)
136 The object takes the form of:
137
138{ # Represents a training or prediction job.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]139 "trainingOutput": { # Represents results of a training job. Output only. # The current training job result.
140 "completedTrialCount": "A String", # The number of hyperparameter tuning trials that completed successfully.
141 # Only set for hyperparameter tuning jobs.
142 "isHyperparameterTuningJob": True or False, # Whether this job is a hyperparameter tuning job.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]143 "trials": [ # Results for individual Hyperparameter trials.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]144 # Only set for hyperparameter tuning jobs.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]145 { # Represents the result of a single hyperparameter tuning trial from a
146 # training job. The TrainingOutput object that is returned on successful
147 # completion of a training job with hyperparameter tuning includes a list
148 # of HyperparameterOutput objects, one for each successful trial.
149 "hyperparameters": { # The hyperparameters given to this trial.
150 "a_key": "A String",
151 },
152 "trialId": "A String", # The trial id for these results.
153 "allMetrics": [ # All recorded object metrics for this trial.
154 { # An observed value of a metric.
155 "trainingStep": "A String", # The global training step for this metric.
156 "objectiveValue": 3.14, # The objective value at this training step.
157 },
158 ],
159 "finalMetric": { # An observed value of a metric. # The final objective metric seen for this trial.
160 "trainingStep": "A String", # The global training step for this metric.
161 "objectiveValue": 3.14, # The objective value at this training step.
162 },
163 },
164 ],
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]165 "consumedMLUnits": 3.14, # The amount of ML units consumed by the job.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]166 },
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]167 "predictionInput": { # Represents input parameters for a prediction job. # Input parameters to create a prediction job.
168 "modelName": "A String", # Use this field if you want to use the default version for the specified
169 # model. The string must use the following format:
170 #
Jon Wayne Parrottdc6c1ef2016-10-14 11:04:30 -0700[diff] [blame]171 # `"projects/<var>[YOUR_PROJECT]</var>/models/<var>[YOUR_MODEL]</var>"`
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]172 "runtimeVersion": "A String", # Optional. The Google Cloud ML runtime version to use for this batch
173 # prediction. If not set, Google Cloud ML will pick the runtime version used
174 # during the CreateVersion request for this model version, or choose the
175 # latest stable version when model version information is not available
176 # such as when the model is specified by uri.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]177 "inputPaths": [ # Required. The Google Cloud Storage location of the input data files.
178 # May contain wildcards.
179 "A String",
180 ],
Jon Wayne Parrottdc6c1ef2016-10-14 11:04:30 -0700[diff] [blame]181 "maxWorkerCount": "A String", # Optional. The maximum number of workers to be used for parallel processing.
182 # Defaults to 10 if not specified.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]183 "uri": "A String", # Use this field if you want to specify a Google Cloud Storage path for
184 # the model to use.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]185 "outputPath": "A String", # Required. The output Google Cloud Storage location.
186 "dataFormat": "A String", # Required. The format of the input data files.
187 "versionName": "A String", # Use this field if you want to specify a version of the model to use. The
188 # string is formatted the same way as `model_version`, with the addition
189 # of the version information:
190 #
Jon Wayne Parrottdc6c1ef2016-10-14 11:04:30 -0700[diff] [blame]191 # `"projects/<var>[YOUR_PROJECT]</var>/models/<var>YOUR_MODEL/versions/<var>[YOUR_VERSION]</var>"`
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]192 "region": "A String", # Required. The Google Compute Engine region to run the prediction job in.
193 },
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]194 "errorMessage": "A String", # Output only. The details of a failure or a cancellation.
195 "jobId": "A String", # Required. The user-specified id of the job.
196 "state": "A String", # Output only. The detailed state of a job.
197 "startTime": "A String", # Output only. When the job processing was started.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]198 "trainingInput": { # Represents input parameters for a training job. # Input parameters to create a training job.
199 "workerType": "A String", # Optional. Specifies the type of virtual machine to use for your training
200 # job's worker nodes.
201 #
202 # The supported values are the same as those described in the entry for
203 # `masterType`.
204 #
205 # This value must be present when `scaleTier` is set to `CUSTOM` and
206 # `workerCount` is greater than zero.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]207 "runtimeVersion": "A String", # Optional. The Google Cloud ML runtime version to use for training. If not
208 # set, Google Cloud ML will choose the latest stable version.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]209 "scaleTier": "A String", # Required. Specifies the machine types, the number of replicas for workers
210 # and parameter servers.
211 "masterType": "A String", # Optional. Specifies the type of virtual machine to use for your training
212 # job's master worker.
213 #
214 # The following types are supported:
215 #
216 # <dl>
217 # <dt>standard</dt>
218 # <dd>
219 # A basic machine configuration suitable for training simple models with
220 # small to moderate datasets.
221 # </dd>
222 # <dt>large_model</dt>
223 # <dd>
224 # A machine with a lot of memory, specially suited for parameter servers
225 # when your model is large (having many hidden layers or layers with very
226 # large numbers of nodes).
227 # </dd>
228 # <dt>complex_model_s</dt>
229 # <dd>
230 # A machine suitable for the master and workers of the cluster when your
231 # model requires more computation than the standard machine can handle
232 # satisfactorily.
233 # </dd>
234 # <dt>complex_model_m</dt>
235 # <dd>
236 # A machine with roughly twice the number of cores and roughly double the
237 # memory of <code suppresswarning="true">complex_model_s</code>.
238 # </dd>
239 # <dt>complex_model_l</dt>
240 # <dd>
241 # A machine with roughly twice the number of cores and roughly double the
242 # memory of <code suppresswarning="true">complex_model_m</code>.
243 # </dd>
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]244 # <dt>standard_gpu</dt>
245 # <dd>
246 # A machine equivalent to <code suppresswarning="true">standard</code> that
247 # also includes a
248 # <a href="ml/docs/how-tos/using-gpus">
249 # GPU that you can use in your trainer</a>.
250 # </dd>
251 # <dt>complex_model_m_gpu</dt>
252 # <dd>
253 # A machine equivalent to
254 # <code suppresswarning="true">coplex_model_m</code> that also includes
255 # four GPUs.
256 # </dd>
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]257 # </dl>
258 #
259 # You must set this value when `scaleTier` is set to `CUSTOM`.
260 "hyperparameters": { # Represents a set of hyperparameters to optimize. # Optional. The set of Hyperparameters to tune.
261 "maxTrials": 42, # Optional. How many training trials should be attempted to optimize
262 # the specified hyperparameters.
263 #
264 # Defaults to one.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]265 "hyperparameterMetricTag": "A String", # Optional. The Tensorflow summary tag name to use for optimizing trials. For
266 # current versions of Tensorflow, this tag name should exactly match what is
267 # shown in Tensorboard, including all scopes. For versions of Tensorflow
268 # prior to 0.12, this should be only the tag passed to tf.Summary.
269 # By default, "training/hptuning/metric" will be used.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]270 "params": [ # Required. The set of parameters to tune.
271 { # Represents a single hyperparameter to optimize.
272 "maxValue": 3.14, # Required if typeis `DOUBLE` or `INTEGER`. This field
273 # should be unset if type is `CATEGORICAL`. This value should be integers if
274 # type is `INTEGER`.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]275 "categoricalValues": [ # Required if type is `CATEGORICAL`. The list of possible categories.
276 "A String",
277 ],
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]278 "discreteValues": [ # Required if type is `DISCRETE`.
279 # A list of feasible points.
280 # The list should be in strictly increasing order. For instance, this
281 # parameter might have possible settings of 1.5, 2.5, and 4.0. This list
282 # should not contain more than 1,000 values.
283 3.14,
284 ],
285 "parameterName": "A String", # Required. The parameter name must be unique amongst all ParameterConfigs in
286 # a HyperparameterSpec message. E.g., "learning_rate".
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]287 "minValue": 3.14, # Required if type is `DOUBLE` or `INTEGER`. This field
288 # should be unset if type is `CATEGORICAL`. This value should be integers if
289 # type is INTEGER.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]290 "type": "A String", # Required. The type of the parameter.
291 "scaleType": "A String", # Optional. How the parameter should be scaled to the hypercube.
292 # Leave unset for categorical parameters.
293 # Some kind of scaling is strongly recommended for real or integral
294 # parameters (e.g., `UNIT_LINEAR_SCALE`).
295 },
296 ],
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]297 "goal": "A String", # Required. The type of goal to use for tuning. Available types are
298 # `MAXIMIZE` and `MINIMIZE`.
299 #
300 # Defaults to `MAXIMIZE`.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]301 "maxParallelTrials": 42, # Optional. The number of training trials to run concurrently.
302 # You can reduce the time it takes to perform hyperparameter tuning by adding
303 # trials in parallel. However, each trail only benefits from the information
304 # gained in completed trials. That means that a trial does not get access to
305 # the results of trials running at the same time, which could reduce the
306 # quality of the overall optimization.
307 #
308 # Each trial will use the same scale tier and machine types.
309 #
310 # Defaults to one.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]311 },
312 "region": "A String", # Required. The Google Compute Engine region to run the training job in.
313 "args": [ # Optional. Command line arguments to pass to the program.
314 "A String",
315 ],
316 "pythonModule": "A String", # Required. The Python module name to run after installing the packages.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]317 "jobDir": "A String", # Optional. A Google Cloud Storage path in which to store training outputs
318 # and other data needed for training. This path is passed to your TensorFlow
319 # program as the 'job_dir' command-line argument. The benefit of specifying
320 # this field is that Cloud ML validates the path for use in training.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]321 "packageUris": [ # Required. The Google Cloud Storage location of the packages with
322 # the training program and any additional dependencies.
323 "A String",
324 ],
325 "workerCount": "A String", # Optional. The number of worker replicas to use for the training job. Each
326 # replica in the cluster will be of the type specified in `worker_type`.
327 #
328 # This value can only be used when `scale_tier` is set to `CUSTOM`. If you
329 # set this value, you must also set `worker_type`.
330 "parameterServerType": "A String", # Optional. Specifies the type of virtual machine to use for your training
331 # job's parameter server.
332 #
333 # The supported values are the same as those described in the entry for
334 # `master_type`.
335 #
336 # This value must be present when `scaleTier` is set to `CUSTOM` and
337 # `parameter_server_count` is greater than zero.
338 "parameterServerCount": "A String", # Optional. The number of parameter server replicas to use for the training
339 # job. Each replica in the cluster will be of the type specified in
340 # `parameter_server_type`.
341 #
342 # This value can only be used when `scale_tier` is set to `CUSTOM`.If you
343 # set this value, you must also set `parameter_server_type`.
344 },
345 "endTime": "A String", # Output only. When the job processing was completed.
346 "predictionOutput": { # Represents results of a prediction job. # The current prediction job result.
347 "outputPath": "A String", # The output Google Cloud Storage location provided at the job creation time.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]348 "nodeHours": 3.14, # Node hours used by the batch prediction job.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]349 "predictionCount": "A String", # The number of generated predictions.
350 "errorCount": "A String", # The number of data instances which resulted in errors.
351 },
352 "createTime": "A String", # Output only. When the job was created.
353 }
354
355 x__xgafv: string, V1 error format.
356 Allowed values
357 1 - v1 error format
358 2 - v2 error format
359
360Returns:
361 An object of the form:
362
363 { # Represents a training or prediction job.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]364 "trainingOutput": { # Represents results of a training job. Output only. # The current training job result.
365 "completedTrialCount": "A String", # The number of hyperparameter tuning trials that completed successfully.
366 # Only set for hyperparameter tuning jobs.
367 "isHyperparameterTuningJob": True or False, # Whether this job is a hyperparameter tuning job.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]368 "trials": [ # Results for individual Hyperparameter trials.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]369 # Only set for hyperparameter tuning jobs.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]370 { # Represents the result of a single hyperparameter tuning trial from a
371 # training job. The TrainingOutput object that is returned on successful
372 # completion of a training job with hyperparameter tuning includes a list
373 # of HyperparameterOutput objects, one for each successful trial.
374 "hyperparameters": { # The hyperparameters given to this trial.
375 "a_key": "A String",
376 },
377 "trialId": "A String", # The trial id for these results.
378 "allMetrics": [ # All recorded object metrics for this trial.
379 { # An observed value of a metric.
380 "trainingStep": "A String", # The global training step for this metric.
381 "objectiveValue": 3.14, # The objective value at this training step.
382 },
383 ],
384 "finalMetric": { # An observed value of a metric. # The final objective metric seen for this trial.
385 "trainingStep": "A String", # The global training step for this metric.
386 "objectiveValue": 3.14, # The objective value at this training step.
387 },
388 },
389 ],
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]390 "consumedMLUnits": 3.14, # The amount of ML units consumed by the job.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]391 },
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]392 "predictionInput": { # Represents input parameters for a prediction job. # Input parameters to create a prediction job.
393 "modelName": "A String", # Use this field if you want to use the default version for the specified
394 # model. The string must use the following format:
395 #
Jon Wayne Parrottdc6c1ef2016-10-14 11:04:30 -0700[diff] [blame]396 # `"projects/<var>[YOUR_PROJECT]</var>/models/<var>[YOUR_MODEL]</var>"`
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]397 "runtimeVersion": "A String", # Optional. The Google Cloud ML runtime version to use for this batch
398 # prediction. If not set, Google Cloud ML will pick the runtime version used
399 # during the CreateVersion request for this model version, or choose the
400 # latest stable version when model version information is not available
401 # such as when the model is specified by uri.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]402 "inputPaths": [ # Required. The Google Cloud Storage location of the input data files.
403 # May contain wildcards.
404 "A String",
405 ],
Jon Wayne Parrottdc6c1ef2016-10-14 11:04:30 -0700[diff] [blame]406 "maxWorkerCount": "A String", # Optional. The maximum number of workers to be used for parallel processing.
407 # Defaults to 10 if not specified.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]408 "uri": "A String", # Use this field if you want to specify a Google Cloud Storage path for
409 # the model to use.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]410 "outputPath": "A String", # Required. The output Google Cloud Storage location.
411 "dataFormat": "A String", # Required. The format of the input data files.
412 "versionName": "A String", # Use this field if you want to specify a version of the model to use. The
413 # string is formatted the same way as `model_version`, with the addition
414 # of the version information:
415 #
Jon Wayne Parrottdc6c1ef2016-10-14 11:04:30 -0700[diff] [blame]416 # `"projects/<var>[YOUR_PROJECT]</var>/models/<var>YOUR_MODEL/versions/<var>[YOUR_VERSION]</var>"`
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]417 "region": "A String", # Required. The Google Compute Engine region to run the prediction job in.
418 },
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]419 "errorMessage": "A String", # Output only. The details of a failure or a cancellation.
420 "jobId": "A String", # Required. The user-specified id of the job.
421 "state": "A String", # Output only. The detailed state of a job.
422 "startTime": "A String", # Output only. When the job processing was started.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]423 "trainingInput": { # Represents input parameters for a training job. # Input parameters to create a training job.
424 "workerType": "A String", # Optional. Specifies the type of virtual machine to use for your training
425 # job's worker nodes.
426 #
427 # The supported values are the same as those described in the entry for
428 # `masterType`.
429 #
430 # This value must be present when `scaleTier` is set to `CUSTOM` and
431 # `workerCount` is greater than zero.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]432 "runtimeVersion": "A String", # Optional. The Google Cloud ML runtime version to use for training. If not
433 # set, Google Cloud ML will choose the latest stable version.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]434 "scaleTier": "A String", # Required. Specifies the machine types, the number of replicas for workers
435 # and parameter servers.
436 "masterType": "A String", # Optional. Specifies the type of virtual machine to use for your training
437 # job's master worker.
438 #
439 # The following types are supported:
440 #
441 # <dl>
442 # <dt>standard</dt>
443 # <dd>
444 # A basic machine configuration suitable for training simple models with
445 # small to moderate datasets.
446 # </dd>
447 # <dt>large_model</dt>
448 # <dd>
449 # A machine with a lot of memory, specially suited for parameter servers
450 # when your model is large (having many hidden layers or layers with very
451 # large numbers of nodes).
452 # </dd>
453 # <dt>complex_model_s</dt>
454 # <dd>
455 # A machine suitable for the master and workers of the cluster when your
456 # model requires more computation than the standard machine can handle
457 # satisfactorily.
458 # </dd>
459 # <dt>complex_model_m</dt>
460 # <dd>
461 # A machine with roughly twice the number of cores and roughly double the
462 # memory of <code suppresswarning="true">complex_model_s</code>.
463 # </dd>
464 # <dt>complex_model_l</dt>
465 # <dd>
466 # A machine with roughly twice the number of cores and roughly double the
467 # memory of <code suppresswarning="true">complex_model_m</code>.
468 # </dd>
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]469 # <dt>standard_gpu</dt>
470 # <dd>
471 # A machine equivalent to <code suppresswarning="true">standard</code> that
472 # also includes a
473 # <a href="ml/docs/how-tos/using-gpus">
474 # GPU that you can use in your trainer</a>.
475 # </dd>
476 # <dt>complex_model_m_gpu</dt>
477 # <dd>
478 # A machine equivalent to
479 # <code suppresswarning="true">coplex_model_m</code> that also includes
480 # four GPUs.
481 # </dd>
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]482 # </dl>
483 #
484 # You must set this value when `scaleTier` is set to `CUSTOM`.
485 "hyperparameters": { # Represents a set of hyperparameters to optimize. # Optional. The set of Hyperparameters to tune.
486 "maxTrials": 42, # Optional. How many training trials should be attempted to optimize
487 # the specified hyperparameters.
488 #
489 # Defaults to one.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]490 "hyperparameterMetricTag": "A String", # Optional. The Tensorflow summary tag name to use for optimizing trials. For
491 # current versions of Tensorflow, this tag name should exactly match what is
492 # shown in Tensorboard, including all scopes. For versions of Tensorflow
493 # prior to 0.12, this should be only the tag passed to tf.Summary.
494 # By default, "training/hptuning/metric" will be used.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]495 "params": [ # Required. The set of parameters to tune.
496 { # Represents a single hyperparameter to optimize.
497 "maxValue": 3.14, # Required if typeis `DOUBLE` or `INTEGER`. This field
498 # should be unset if type is `CATEGORICAL`. This value should be integers if
499 # type is `INTEGER`.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]500 "categoricalValues": [ # Required if type is `CATEGORICAL`. The list of possible categories.
501 "A String",
502 ],
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]503 "discreteValues": [ # Required if type is `DISCRETE`.
504 # A list of feasible points.
505 # The list should be in strictly increasing order. For instance, this
506 # parameter might have possible settings of 1.5, 2.5, and 4.0. This list
507 # should not contain more than 1,000 values.
508 3.14,
509 ],
510 "parameterName": "A String", # Required. The parameter name must be unique amongst all ParameterConfigs in
511 # a HyperparameterSpec message. E.g., "learning_rate".
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]512 "minValue": 3.14, # Required if type is `DOUBLE` or `INTEGER`. This field
513 # should be unset if type is `CATEGORICAL`. This value should be integers if
514 # type is INTEGER.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]515 "type": "A String", # Required. The type of the parameter.
516 "scaleType": "A String", # Optional. How the parameter should be scaled to the hypercube.
517 # Leave unset for categorical parameters.
518 # Some kind of scaling is strongly recommended for real or integral
519 # parameters (e.g., `UNIT_LINEAR_SCALE`).
520 },
521 ],
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]522 "goal": "A String", # Required. The type of goal to use for tuning. Available types are
523 # `MAXIMIZE` and `MINIMIZE`.
524 #
525 # Defaults to `MAXIMIZE`.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]526 "maxParallelTrials": 42, # Optional. The number of training trials to run concurrently.
527 # You can reduce the time it takes to perform hyperparameter tuning by adding
528 # trials in parallel. However, each trail only benefits from the information
529 # gained in completed trials. That means that a trial does not get access to
530 # the results of trials running at the same time, which could reduce the
531 # quality of the overall optimization.
532 #
533 # Each trial will use the same scale tier and machine types.
534 #
535 # Defaults to one.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]536 },
537 "region": "A String", # Required. The Google Compute Engine region to run the training job in.
538 "args": [ # Optional. Command line arguments to pass to the program.
539 "A String",
540 ],
541 "pythonModule": "A String", # Required. The Python module name to run after installing the packages.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]542 "jobDir": "A String", # Optional. A Google Cloud Storage path in which to store training outputs
543 # and other data needed for training. This path is passed to your TensorFlow
544 # program as the 'job_dir' command-line argument. The benefit of specifying
545 # this field is that Cloud ML validates the path for use in training.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]546 "packageUris": [ # Required. The Google Cloud Storage location of the packages with
547 # the training program and any additional dependencies.
548 "A String",
549 ],
550 "workerCount": "A String", # Optional. The number of worker replicas to use for the training job. Each
551 # replica in the cluster will be of the type specified in `worker_type`.
552 #
553 # This value can only be used when `scale_tier` is set to `CUSTOM`. If you
554 # set this value, you must also set `worker_type`.
555 "parameterServerType": "A String", # Optional. Specifies the type of virtual machine to use for your training
556 # job's parameter server.
557 #
558 # The supported values are the same as those described in the entry for
559 # `master_type`.
560 #
561 # This value must be present when `scaleTier` is set to `CUSTOM` and
562 # `parameter_server_count` is greater than zero.
563 "parameterServerCount": "A String", # Optional. The number of parameter server replicas to use for the training
564 # job. Each replica in the cluster will be of the type specified in
565 # `parameter_server_type`.
566 #
567 # This value can only be used when `scale_tier` is set to `CUSTOM`.If you
568 # set this value, you must also set `parameter_server_type`.
569 },
570 "endTime": "A String", # Output only. When the job processing was completed.
571 "predictionOutput": { # Represents results of a prediction job. # The current prediction job result.
572 "outputPath": "A String", # The output Google Cloud Storage location provided at the job creation time.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]573 "nodeHours": 3.14, # Node hours used by the batch prediction job.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]574 "predictionCount": "A String", # The number of generated predictions.
575 "errorCount": "A String", # The number of data instances which resulted in errors.
576 },
577 "createTime": "A String", # Output only. When the job was created.
578 }</pre>
579</div>
580
581<div class="method">
582 <code class="details" id="get">get(name=None, x__xgafv=None)</code>
583 <pre>Describes a job.
584
585Args:
586 name: string, Required. The name of the job to get the description of.
587
588Authorization: requires `Viewer` role on the parent project. (required)
589 x__xgafv: string, V1 error format.
590 Allowed values
591 1 - v1 error format
592 2 - v2 error format
593
594Returns:
595 An object of the form:
596
597 { # Represents a training or prediction job.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]598 "trainingOutput": { # Represents results of a training job. Output only. # The current training job result.
599 "completedTrialCount": "A String", # The number of hyperparameter tuning trials that completed successfully.
600 # Only set for hyperparameter tuning jobs.
601 "isHyperparameterTuningJob": True or False, # Whether this job is a hyperparameter tuning job.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]602 "trials": [ # Results for individual Hyperparameter trials.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]603 # Only set for hyperparameter tuning jobs.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]604 { # Represents the result of a single hyperparameter tuning trial from a
605 # training job. The TrainingOutput object that is returned on successful
606 # completion of a training job with hyperparameter tuning includes a list
607 # of HyperparameterOutput objects, one for each successful trial.
608 "hyperparameters": { # The hyperparameters given to this trial.
609 "a_key": "A String",
610 },
611 "trialId": "A String", # The trial id for these results.
612 "allMetrics": [ # All recorded object metrics for this trial.
613 { # An observed value of a metric.
614 "trainingStep": "A String", # The global training step for this metric.
615 "objectiveValue": 3.14, # The objective value at this training step.
616 },
617 ],
618 "finalMetric": { # An observed value of a metric. # The final objective metric seen for this trial.
619 "trainingStep": "A String", # The global training step for this metric.
620 "objectiveValue": 3.14, # The objective value at this training step.
621 },
622 },
623 ],
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]624 "consumedMLUnits": 3.14, # The amount of ML units consumed by the job.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]625 },
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]626 "predictionInput": { # Represents input parameters for a prediction job. # Input parameters to create a prediction job.
627 "modelName": "A String", # Use this field if you want to use the default version for the specified
628 # model. The string must use the following format:
629 #
Jon Wayne Parrottdc6c1ef2016-10-14 11:04:30 -0700[diff] [blame]630 # `"projects/<var>[YOUR_PROJECT]</var>/models/<var>[YOUR_MODEL]</var>"`
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]631 "runtimeVersion": "A String", # Optional. The Google Cloud ML runtime version to use for this batch
632 # prediction. If not set, Google Cloud ML will pick the runtime version used
633 # during the CreateVersion request for this model version, or choose the
634 # latest stable version when model version information is not available
635 # such as when the model is specified by uri.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]636 "inputPaths": [ # Required. The Google Cloud Storage location of the input data files.
637 # May contain wildcards.
638 "A String",
639 ],
Jon Wayne Parrottdc6c1ef2016-10-14 11:04:30 -0700[diff] [blame]640 "maxWorkerCount": "A String", # Optional. The maximum number of workers to be used for parallel processing.
641 # Defaults to 10 if not specified.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]642 "uri": "A String", # Use this field if you want to specify a Google Cloud Storage path for
643 # the model to use.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]644 "outputPath": "A String", # Required. The output Google Cloud Storage location.
645 "dataFormat": "A String", # Required. The format of the input data files.
646 "versionName": "A String", # Use this field if you want to specify a version of the model to use. The
647 # string is formatted the same way as `model_version`, with the addition
648 # of the version information:
649 #
Jon Wayne Parrottdc6c1ef2016-10-14 11:04:30 -0700[diff] [blame]650 # `"projects/<var>[YOUR_PROJECT]</var>/models/<var>YOUR_MODEL/versions/<var>[YOUR_VERSION]</var>"`
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]651 "region": "A String", # Required. The Google Compute Engine region to run the prediction job in.
652 },
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]653 "errorMessage": "A String", # Output only. The details of a failure or a cancellation.
654 "jobId": "A String", # Required. The user-specified id of the job.
655 "state": "A String", # Output only. The detailed state of a job.
656 "startTime": "A String", # Output only. When the job processing was started.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]657 "trainingInput": { # Represents input parameters for a training job. # Input parameters to create a training job.
658 "workerType": "A String", # Optional. Specifies the type of virtual machine to use for your training
659 # job's worker nodes.
660 #
661 # The supported values are the same as those described in the entry for
662 # `masterType`.
663 #
664 # This value must be present when `scaleTier` is set to `CUSTOM` and
665 # `workerCount` is greater than zero.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]666 "runtimeVersion": "A String", # Optional. The Google Cloud ML runtime version to use for training. If not
667 # set, Google Cloud ML will choose the latest stable version.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]668 "scaleTier": "A String", # Required. Specifies the machine types, the number of replicas for workers
669 # and parameter servers.
670 "masterType": "A String", # Optional. Specifies the type of virtual machine to use for your training
671 # job's master worker.
672 #
673 # The following types are supported:
674 #
675 # <dl>
676 # <dt>standard</dt>
677 # <dd>
678 # A basic machine configuration suitable for training simple models with
679 # small to moderate datasets.
680 # </dd>
681 # <dt>large_model</dt>
682 # <dd>
683 # A machine with a lot of memory, specially suited for parameter servers
684 # when your model is large (having many hidden layers or layers with very
685 # large numbers of nodes).
686 # </dd>
687 # <dt>complex_model_s</dt>
688 # <dd>
689 # A machine suitable for the master and workers of the cluster when your
690 # model requires more computation than the standard machine can handle
691 # satisfactorily.
692 # </dd>
693 # <dt>complex_model_m</dt>
694 # <dd>
695 # A machine with roughly twice the number of cores and roughly double the
696 # memory of <code suppresswarning="true">complex_model_s</code>.
697 # </dd>
698 # <dt>complex_model_l</dt>
699 # <dd>
700 # A machine with roughly twice the number of cores and roughly double the
701 # memory of <code suppresswarning="true">complex_model_m</code>.
702 # </dd>
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]703 # <dt>standard_gpu</dt>
704 # <dd>
705 # A machine equivalent to <code suppresswarning="true">standard</code> that
706 # also includes a
707 # <a href="ml/docs/how-tos/using-gpus">
708 # GPU that you can use in your trainer</a>.
709 # </dd>
710 # <dt>complex_model_m_gpu</dt>
711 # <dd>
712 # A machine equivalent to
713 # <code suppresswarning="true">coplex_model_m</code> that also includes
714 # four GPUs.
715 # </dd>
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]716 # </dl>
717 #
718 # You must set this value when `scaleTier` is set to `CUSTOM`.
719 "hyperparameters": { # Represents a set of hyperparameters to optimize. # Optional. The set of Hyperparameters to tune.
720 "maxTrials": 42, # Optional. How many training trials should be attempted to optimize
721 # the specified hyperparameters.
722 #
723 # Defaults to one.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]724 "hyperparameterMetricTag": "A String", # Optional. The Tensorflow summary tag name to use for optimizing trials. For
725 # current versions of Tensorflow, this tag name should exactly match what is
726 # shown in Tensorboard, including all scopes. For versions of Tensorflow
727 # prior to 0.12, this should be only the tag passed to tf.Summary.
728 # By default, "training/hptuning/metric" will be used.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]729 "params": [ # Required. The set of parameters to tune.
730 { # Represents a single hyperparameter to optimize.
731 "maxValue": 3.14, # Required if typeis `DOUBLE` or `INTEGER`. This field
732 # should be unset if type is `CATEGORICAL`. This value should be integers if
733 # type is `INTEGER`.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]734 "categoricalValues": [ # Required if type is `CATEGORICAL`. The list of possible categories.
735 "A String",
736 ],
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]737 "discreteValues": [ # Required if type is `DISCRETE`.
738 # A list of feasible points.
739 # The list should be in strictly increasing order. For instance, this
740 # parameter might have possible settings of 1.5, 2.5, and 4.0. This list
741 # should not contain more than 1,000 values.
742 3.14,
743 ],
744 "parameterName": "A String", # Required. The parameter name must be unique amongst all ParameterConfigs in
745 # a HyperparameterSpec message. E.g., "learning_rate".
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]746 "minValue": 3.14, # Required if type is `DOUBLE` or `INTEGER`. This field
747 # should be unset if type is `CATEGORICAL`. This value should be integers if
748 # type is INTEGER.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]749 "type": "A String", # Required. The type of the parameter.
750 "scaleType": "A String", # Optional. How the parameter should be scaled to the hypercube.
751 # Leave unset for categorical parameters.
752 # Some kind of scaling is strongly recommended for real or integral
753 # parameters (e.g., `UNIT_LINEAR_SCALE`).
754 },
755 ],
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]756 "goal": "A String", # Required. The type of goal to use for tuning. Available types are
757 # `MAXIMIZE` and `MINIMIZE`.
758 #
759 # Defaults to `MAXIMIZE`.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]760 "maxParallelTrials": 42, # Optional. The number of training trials to run concurrently.
761 # You can reduce the time it takes to perform hyperparameter tuning by adding
762 # trials in parallel. However, each trail only benefits from the information
763 # gained in completed trials. That means that a trial does not get access to
764 # the results of trials running at the same time, which could reduce the
765 # quality of the overall optimization.
766 #
767 # Each trial will use the same scale tier and machine types.
768 #
769 # Defaults to one.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]770 },
771 "region": "A String", # Required. The Google Compute Engine region to run the training job in.
772 "args": [ # Optional. Command line arguments to pass to the program.
773 "A String",
774 ],
775 "pythonModule": "A String", # Required. The Python module name to run after installing the packages.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]776 "jobDir": "A String", # Optional. A Google Cloud Storage path in which to store training outputs
777 # and other data needed for training. This path is passed to your TensorFlow
778 # program as the 'job_dir' command-line argument. The benefit of specifying
779 # this field is that Cloud ML validates the path for use in training.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]780 "packageUris": [ # Required. The Google Cloud Storage location of the packages with
781 # the training program and any additional dependencies.
782 "A String",
783 ],
784 "workerCount": "A String", # Optional. The number of worker replicas to use for the training job. Each
785 # replica in the cluster will be of the type specified in `worker_type`.
786 #
787 # This value can only be used when `scale_tier` is set to `CUSTOM`. If you
788 # set this value, you must also set `worker_type`.
789 "parameterServerType": "A String", # Optional. Specifies the type of virtual machine to use for your training
790 # job's parameter server.
791 #
792 # The supported values are the same as those described in the entry for
793 # `master_type`.
794 #
795 # This value must be present when `scaleTier` is set to `CUSTOM` and
796 # `parameter_server_count` is greater than zero.
797 "parameterServerCount": "A String", # Optional. The number of parameter server replicas to use for the training
798 # job. Each replica in the cluster will be of the type specified in
799 # `parameter_server_type`.
800 #
801 # This value can only be used when `scale_tier` is set to `CUSTOM`.If you
802 # set this value, you must also set `parameter_server_type`.
803 },
804 "endTime": "A String", # Output only. When the job processing was completed.
805 "predictionOutput": { # Represents results of a prediction job. # The current prediction job result.
806 "outputPath": "A String", # The output Google Cloud Storage location provided at the job creation time.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]807 "nodeHours": 3.14, # Node hours used by the batch prediction job.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]808 "predictionCount": "A String", # The number of generated predictions.
809 "errorCount": "A String", # The number of data instances which resulted in errors.
810 },
811 "createTime": "A String", # Output only. When the job was created.
812 }</pre>
813</div>
814
815<div class="method">
816 <code class="details" id="list">list(parent=None, pageSize=None, filter=None, pageToken=None, x__xgafv=None)</code>
817 <pre>Lists the jobs in the project.
818
819Args:
820 parent: string, Required. The name of the project for which to list jobs.
821
822Authorization: requires `Viewer` role on the specified project. (required)
823 pageSize: integer, Optional. The number of jobs to retrieve per "page" of results. If there
824are more remaining results than this number, the response message will
825contain a valid value in the `next_page_token` field.
826
827The default value is 20, and the maximum page size is 100.
828 filter: string, Optional. Specifies the subset of jobs to retrieve.
829 pageToken: string, Optional. A page token to request the next page of results.
830
831You get the token from the `next_page_token` field of the response from
832the previous call.
833 x__xgafv: string, V1 error format.
834 Allowed values
835 1 - v1 error format
836 2 - v2 error format
837
838Returns:
839 An object of the form:
840
841 { # Response message for the ListJobs method.
842 "nextPageToken": "A String", # Optional. Pass this token as the `page_token` field of the request for a
843 # subsequent call.
844 "jobs": [ # The list of jobs.
845 { # Represents a training or prediction job.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]846 "trainingOutput": { # Represents results of a training job. Output only. # The current training job result.
847 "completedTrialCount": "A String", # The number of hyperparameter tuning trials that completed successfully.
848 # Only set for hyperparameter tuning jobs.
849 "isHyperparameterTuningJob": True or False, # Whether this job is a hyperparameter tuning job.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]850 "trials": [ # Results for individual Hyperparameter trials.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]851 # Only set for hyperparameter tuning jobs.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]852 { # Represents the result of a single hyperparameter tuning trial from a
853 # training job. The TrainingOutput object that is returned on successful
854 # completion of a training job with hyperparameter tuning includes a list
855 # of HyperparameterOutput objects, one for each successful trial.
856 "hyperparameters": { # The hyperparameters given to this trial.
857 "a_key": "A String",
858 },
859 "trialId": "A String", # The trial id for these results.
860 "allMetrics": [ # All recorded object metrics for this trial.
861 { # An observed value of a metric.
862 "trainingStep": "A String", # The global training step for this metric.
863 "objectiveValue": 3.14, # The objective value at this training step.
864 },
865 ],
866 "finalMetric": { # An observed value of a metric. # The final objective metric seen for this trial.
867 "trainingStep": "A String", # The global training step for this metric.
868 "objectiveValue": 3.14, # The objective value at this training step.
869 },
870 },
871 ],
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]872 "consumedMLUnits": 3.14, # The amount of ML units consumed by the job.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]873 },
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]874 "predictionInput": { # Represents input parameters for a prediction job. # Input parameters to create a prediction job.
875 "modelName": "A String", # Use this field if you want to use the default version for the specified
876 # model. The string must use the following format:
877 #
Jon Wayne Parrottdc6c1ef2016-10-14 11:04:30 -0700[diff] [blame]878 # `"projects/<var>[YOUR_PROJECT]</var>/models/<var>[YOUR_MODEL]</var>"`
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]879 "runtimeVersion": "A String", # Optional. The Google Cloud ML runtime version to use for this batch
880 # prediction. If not set, Google Cloud ML will pick the runtime version used
881 # during the CreateVersion request for this model version, or choose the
882 # latest stable version when model version information is not available
883 # such as when the model is specified by uri.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]884 "inputPaths": [ # Required. The Google Cloud Storage location of the input data files.
885 # May contain wildcards.
886 "A String",
887 ],
Jon Wayne Parrottdc6c1ef2016-10-14 11:04:30 -0700[diff] [blame]888 "maxWorkerCount": "A String", # Optional. The maximum number of workers to be used for parallel processing.
889 # Defaults to 10 if not specified.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]890 "uri": "A String", # Use this field if you want to specify a Google Cloud Storage path for
891 # the model to use.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]892 "outputPath": "A String", # Required. The output Google Cloud Storage location.
893 "dataFormat": "A String", # Required. The format of the input data files.
894 "versionName": "A String", # Use this field if you want to specify a version of the model to use. The
895 # string is formatted the same way as `model_version`, with the addition
896 # of the version information:
897 #
Jon Wayne Parrottdc6c1ef2016-10-14 11:04:30 -0700[diff] [blame]898 # `"projects/<var>[YOUR_PROJECT]</var>/models/<var>YOUR_MODEL/versions/<var>[YOUR_VERSION]</var>"`
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]899 "region": "A String", # Required. The Google Compute Engine region to run the prediction job in.
900 },
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]901 "errorMessage": "A String", # Output only. The details of a failure or a cancellation.
902 "jobId": "A String", # Required. The user-specified id of the job.
903 "state": "A String", # Output only. The detailed state of a job.
904 "startTime": "A String", # Output only. When the job processing was started.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]905 "trainingInput": { # Represents input parameters for a training job. # Input parameters to create a training job.
906 "workerType": "A String", # Optional. Specifies the type of virtual machine to use for your training
907 # job's worker nodes.
908 #
909 # The supported values are the same as those described in the entry for
910 # `masterType`.
911 #
912 # This value must be present when `scaleTier` is set to `CUSTOM` and
913 # `workerCount` is greater than zero.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]914 "runtimeVersion": "A String", # Optional. The Google Cloud ML runtime version to use for training. If not
915 # set, Google Cloud ML will choose the latest stable version.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]916 "scaleTier": "A String", # Required. Specifies the machine types, the number of replicas for workers
917 # and parameter servers.
918 "masterType": "A String", # Optional. Specifies the type of virtual machine to use for your training
919 # job's master worker.
920 #
921 # The following types are supported:
922 #
923 # <dl>
924 # <dt>standard</dt>
925 # <dd>
926 # A basic machine configuration suitable for training simple models with
927 # small to moderate datasets.
928 # </dd>
929 # <dt>large_model</dt>
930 # <dd>
931 # A machine with a lot of memory, specially suited for parameter servers
932 # when your model is large (having many hidden layers or layers with very
933 # large numbers of nodes).
934 # </dd>
935 # <dt>complex_model_s</dt>
936 # <dd>
937 # A machine suitable for the master and workers of the cluster when your
938 # model requires more computation than the standard machine can handle
939 # satisfactorily.
940 # </dd>
941 # <dt>complex_model_m</dt>
942 # <dd>
943 # A machine with roughly twice the number of cores and roughly double the
944 # memory of <code suppresswarning="true">complex_model_s</code>.
945 # </dd>
946 # <dt>complex_model_l</dt>
947 # <dd>
948 # A machine with roughly twice the number of cores and roughly double the
949 # memory of <code suppresswarning="true">complex_model_m</code>.
950 # </dd>
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]951 # <dt>standard_gpu</dt>
952 # <dd>
953 # A machine equivalent to <code suppresswarning="true">standard</code> that
954 # also includes a
955 # <a href="ml/docs/how-tos/using-gpus">
956 # GPU that you can use in your trainer</a>.
957 # </dd>
958 # <dt>complex_model_m_gpu</dt>
959 # <dd>
960 # A machine equivalent to
961 # <code suppresswarning="true">coplex_model_m</code> that also includes
962 # four GPUs.
963 # </dd>
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]964 # </dl>
965 #
966 # You must set this value when `scaleTier` is set to `CUSTOM`.
967 "hyperparameters": { # Represents a set of hyperparameters to optimize. # Optional. The set of Hyperparameters to tune.
968 "maxTrials": 42, # Optional. How many training trials should be attempted to optimize
969 # the specified hyperparameters.
970 #
971 # Defaults to one.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]972 "hyperparameterMetricTag": "A String", # Optional. The Tensorflow summary tag name to use for optimizing trials. For
973 # current versions of Tensorflow, this tag name should exactly match what is
974 # shown in Tensorboard, including all scopes. For versions of Tensorflow
975 # prior to 0.12, this should be only the tag passed to tf.Summary.
976 # By default, "training/hptuning/metric" will be used.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]977 "params": [ # Required. The set of parameters to tune.
978 { # Represents a single hyperparameter to optimize.
979 "maxValue": 3.14, # Required if typeis `DOUBLE` or `INTEGER`. This field
980 # should be unset if type is `CATEGORICAL`. This value should be integers if
981 # type is `INTEGER`.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]982 "categoricalValues": [ # Required if type is `CATEGORICAL`. The list of possible categories.
983 "A String",
984 ],
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]985 "discreteValues": [ # Required if type is `DISCRETE`.
986 # A list of feasible points.
987 # The list should be in strictly increasing order. For instance, this
988 # parameter might have possible settings of 1.5, 2.5, and 4.0. This list
989 # should not contain more than 1,000 values.
990 3.14,
991 ],
992 "parameterName": "A String", # Required. The parameter name must be unique amongst all ParameterConfigs in
993 # a HyperparameterSpec message. E.g., "learning_rate".
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]994 "minValue": 3.14, # Required if type is `DOUBLE` or `INTEGER`. This field
995 # should be unset if type is `CATEGORICAL`. This value should be integers if
996 # type is INTEGER.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]997 "type": "A String", # Required. The type of the parameter.
998 "scaleType": "A String", # Optional. How the parameter should be scaled to the hypercube.
999 # Leave unset for categorical parameters.
1000 # Some kind of scaling is strongly recommended for real or integral
1001 # parameters (e.g., `UNIT_LINEAR_SCALE`).
1002 },
1003 ],
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]1004 "goal": "A String", # Required. The type of goal to use for tuning. Available types are
1005 # `MAXIMIZE` and `MINIMIZE`.
1006 #
1007 # Defaults to `MAXIMIZE`.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]1008 "maxParallelTrials": 42, # Optional. The number of training trials to run concurrently.
1009 # You can reduce the time it takes to perform hyperparameter tuning by adding
1010 # trials in parallel. However, each trail only benefits from the information
1011 # gained in completed trials. That means that a trial does not get access to
1012 # the results of trials running at the same time, which could reduce the
1013 # quality of the overall optimization.
1014 #
1015 # Each trial will use the same scale tier and machine types.
1016 #
1017 # Defaults to one.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]1018 },
1019 "region": "A String", # Required. The Google Compute Engine region to run the training job in.
1020 "args": [ # Optional. Command line arguments to pass to the program.
1021 "A String",
1022 ],
1023 "pythonModule": "A String", # Required. The Python module name to run after installing the packages.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]1024 "jobDir": "A String", # Optional. A Google Cloud Storage path in which to store training outputs
1025 # and other data needed for training. This path is passed to your TensorFlow
1026 # program as the 'job_dir' command-line argument. The benefit of specifying
1027 # this field is that Cloud ML validates the path for use in training.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]1028 "packageUris": [ # Required. The Google Cloud Storage location of the packages with
1029 # the training program and any additional dependencies.
1030 "A String",
1031 ],
1032 "workerCount": "A String", # Optional. The number of worker replicas to use for the training job. Each
1033 # replica in the cluster will be of the type specified in `worker_type`.
1034 #
1035 # This value can only be used when `scale_tier` is set to `CUSTOM`. If you
1036 # set this value, you must also set `worker_type`.
1037 "parameterServerType": "A String", # Optional. Specifies the type of virtual machine to use for your training
1038 # job's parameter server.
1039 #
1040 # The supported values are the same as those described in the entry for
1041 # `master_type`.
1042 #
1043 # This value must be present when `scaleTier` is set to `CUSTOM` and
1044 # `parameter_server_count` is greater than zero.
1045 "parameterServerCount": "A String", # Optional. The number of parameter server replicas to use for the training
1046 # job. Each replica in the cluster will be of the type specified in
1047 # `parameter_server_type`.
1048 #
1049 # This value can only be used when `scale_tier` is set to `CUSTOM`.If you
1050 # set this value, you must also set `parameter_server_type`.
1051 },
1052 "endTime": "A String", # Output only. When the job processing was completed.
1053 "predictionOutput": { # Represents results of a prediction job. # The current prediction job result.
1054 "outputPath": "A String", # The output Google Cloud Storage location provided at the job creation time.
Sai Cheemalapatic30d2b52017-03-13 12:12:03 -0400[diff] [blame^]1055 "nodeHours": 3.14, # Node hours used by the batch prediction job.
Sai Cheemalapatiea3a5e12016-10-12 14:05:53 -0700[diff] [blame]1056 "predictionCount": "A String", # The number of generated predictions.
1057 "errorCount": "A String", # The number of data instances which resulted in errors.
1058 },
1059 "createTime": "A String", # Output only. When the job was created.
1060 },
1061 ],
1062 }</pre>
1063</div>
1064
1065<div class="method">
1066 <code class="details" id="list_next">list_next(previous_request, previous_response)</code>
1067 <pre>Retrieves the next page of results.
1068
1069Args:
1070 previous_request: The request for the previous page. (required)
1071 previous_response: The response from the request for the previous page. (required)
1072
1073Returns:
1074 A request object that you can call 'execute()' on to request the next
1075 page. Returns None if there are no more items in the collection.
1076 </pre>
1077</div>
1078
1079</body></html>