Blame - docs/dyn/ml_v1.projects.models.html - platform/external/python/google-api-python-client

2017-03-13 12:12:03 -0400

[diff] [blame]

76

<h2>Instance Methods</h2>

77

78

<code><a href="ml_v1.projects.models.versions.html">versions()</a></code>

79

</p>

80

<p class="firstline">Returns the versions Resource.</p>

81

82

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

83

<code><a href="#create">create(parent, body=None, x__xgafv=None)</a></code></p>

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

84

<p class="firstline">Creates a model which will later contain one or more versions.</p>

85

Thomas Coffee

2f24537

2017-03-27 10:39:26 -0700

[diff] [blame]

86

<code><a href="#delete">delete(name, x__xgafv=None)</a></code></p>

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

87

<p class="firstline">Deletes a model.</p>

88

Thomas Coffee

2f24537

2017-03-27 10:39:26 -0700

[diff] [blame]

89

<code><a href="#get">get(name, x__xgafv=None)</a></code></p>

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

90

<p class="firstline">Gets information about a model, including its name, the description (if</p>

91

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

92

<code><a href="#getIamPolicy">getIamPolicy(resource, options_requestedPolicyVersion=None, x__xgafv=None)</a></code></p>

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

93

<p class="firstline">Gets the access control policy for a resource.</p>

94

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

95

<code><a href="#list">list(parent, pageSize=None, pageToken=None, filter=None, x__xgafv=None)</a></code></p>

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

96

<p class="firstline">Lists the models in a project.</p>

97

98

<code><a href="#list_next">list_next(previous_request, previous_response)</a></code></p>

99

<p class="firstline">Retrieves the next page of results.</p>

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

100

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

101

<code><a href="#patch">patch(name, body=None, updateMask=None, x__xgafv=None)</a></code></p>

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

102

<p class="firstline">Updates a specific model resource.</p>

103

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

104

<code><a href="#setIamPolicy">setIamPolicy(resource, body=None, x__xgafv=None)</a></code></p>

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

105

<p class="firstline">Sets the access control policy on the specified resource. Replaces any</p>

106

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

107

<code><a href="#testIamPermissions">testIamPermissions(resource, body=None, x__xgafv=None)</a></code></p>

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

108

<p class="firstline">Returns permissions that a caller has on the specified resource.</p>

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

109

<h3>Method Details</h3>

110

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

111

<code class="details" id="create">create(parent, body=None, x__xgafv=None)</code>

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

112

<pre>Creates a model which will later contain one or more versions.

113

114

You must add at least one version before you can request predictions from

115

the model. Add versions by calling

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

116

projects.models.versions.create.

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

117

118

Args:

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

119

parent: string, Required. The project name. (required)

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

120

body: object, The request body.

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

121

The object takes the form of:

122

123

{ # Represents a machine learning solution.

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

124

#

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

125

# A model can have multiple versions, each of which is a deployed, trained

126

# model ready to receive prediction requests. The model itself is just a

127

# container.

128

"description": "A String", # Optional. The description specified for the model when it was created.

129

"regions": [ # Optional. The list of regions where the model is going to be deployed.

130

# Only one region per model is supported.

131

# Defaults to 'us-central1' if nothing is set.

132

# See the <a href="/ml-engine/docs/tensorflow/regions">available regions</a>

133

# for AI Platform services.

134

# Note:

135

# * No matter where a model is deployed, it can always be accessed by

136

# users from anywhere, both for online and batch prediction.

137

# * The region for a batch prediction job is set by the region field when

138

# submitting the batch prediction job and does not take its value from

139

# this field.

Bu Sun Kim

2020-05-27 12:20:54 -0700

[diff] [blame]

140

"A String",

141

],

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

142

"name": "A String", # Required. The name specified for the model when it was created.

143

#

144

# The model name must be unique within the project it is created in.

145

"onlinePredictionConsoleLogging": True or False, # Optional. If true, online prediction nodes send `stderr` and `stdout`

146

# streams to Stackdriver Logging. These can be more verbose than the standard

147

# access logs (see `onlinePredictionLogging`) and can incur higher cost.

148

# However, they are helpful for debugging. Note that

149

# [Stackdriver logs may incur a cost](/stackdriver/pricing), especially if

150

# your project receives prediction requests at a high QPS. Estimate your

151

# costs before enabling this option.

152

#

153

# Default is false.

Bu Sun Kim

2020-05-27 12:20:54 -0700

[diff] [blame]

154

"etag": "A String", # `etag` is used for optimistic concurrency control as a way to help

155

# prevent simultaneous updates of a model from overwriting each other.

156

# It is strongly suggested that systems make use of the `etag` in the

157

# read-modify-write cycle to perform model updates in order to avoid race

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

158

# conditions: An `etag` is returned in the response to `GetModel`, and

159

# systems are expected to put that etag in the request to `UpdateModel` to

Bu Sun Kim

2020-05-27 12:20:54 -0700

[diff] [blame]

160

# ensure that their change will be applied to the model as intended.

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

161

"labels": { # Optional. One or more labels that you can add, to organize your models.

162

# Each label is a key-value pair, where both the key and the value are

163

# arbitrary strings that you supply.

164

# For more information, see the documentation on

165

# <a href="/ml-engine/docs/tensorflow/resource-labels">using labels</a>.

166

"a_key": "A String",

167

},

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

168

"defaultVersion": { # Represents a version of the model. # Output only. The default version of the model. This version will be used to

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

169

# handle prediction requests that do not specify a version.

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

170

#

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

171

# You can change the default version by calling

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

172

# projects.models.versions.setDefault.

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

173

#

174

# Each version is a trained model deployed in the cloud, ready to handle

175

# prediction requests. A model can have multiple versions. You can get

176

# information about all of the versions of a given model by calling

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

177

# projects.models.versions.list.

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

178

"labels": { # Optional. One or more labels that you can add, to organize your model

179

# versions. Each label is a key-value pair, where both the key and the value

180

# are arbitrary strings that you supply.

181

# For more information, see the documentation on

182

# <a href="/ml-engine/docs/tensorflow/resource-labels">using labels</a>.

183

"a_key": "A String",

Bu Sun Kim

2020-05-27 12:20:54 -0700

[diff] [blame]

184

},

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

185

"machineType": "A String", # Optional. The type of machine on which to serve the model. Currently only

186

# applies to online prediction service. If this field is not specified, it

187

# defaults to `mls1-c1-m2`.

Bu Sun Kim

2020-05-27 12:20:54 -0700

[diff] [blame]

188

#

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

189

# Online prediction supports the following machine types:

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

190

#

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

# * `mls1-c1-m2`

# * `mls1-c4-m2`

# * `n1-standard-2`

# * `n1-standard-4`

# * `n1-standard-8`

# * `n1-standard-16`

# * `n1-standard-32`

# * `n1-highmem-2`

# * `n1-highmem-4`

# * `n1-highmem-8`

# * `n1-highmem-16`

# * `n1-highmem-32`

# * `n1-highcpu-2`

# * `n1-highcpu-4`

# * `n1-highcpu-8`

# * `n1-highcpu-16`

# * `n1-highcpu-32`

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

208

#

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

209

# `mls1-c1-m2` is generally available. All other machine types are available

210

# in beta. Learn more about the [differences between machine

211

# types](/ml-engine/docs/machine-types-online-prediction).

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

212

"packageUris": [ # Optional. Cloud Storage paths (`gs://…`) of packages for [custom

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

213

# prediction routines](/ml-engine/docs/tensorflow/custom-prediction-routines)

214

# or [scikit-learn pipelines with custom

215

# code](/ml-engine/docs/scikit/exporting-for-prediction#custom-pipeline-code).

216

#

217

# For a custom prediction routine, one of these packages must contain your

218

# Predictor class (see

219

# [`predictionClass`](#Version.FIELDS.prediction_class)). Additionally,

220

# include any dependencies used by your Predictor or scikit-learn pipeline

221

# uses that are not already included in your selected [runtime

222

# version](/ml-engine/docs/tensorflow/runtime-version-list).

223

#

224

# If you specify this field, you must also set

225

# [`runtimeVersion`](#Version.FIELDS.runtime_version) to 1.4 or greater.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

226

"A String",

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

227

],

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

228

"acceleratorConfig": { # Represents a hardware accelerator request config. # Optional. Accelerator config for using GPUs for online prediction (beta).

229

# Only specify this field if you have specified a Compute Engine (N1) machine

230

# type in the `machineType` field. Learn more about [using GPUs for online

231

# prediction](/ml-engine/docs/machine-types-online-prediction#gpus).

232

# Note that the AcceleratorConfig can be used in both Jobs and Versions.

233

# Learn more about [accelerators for training](/ml-engine/docs/using-gpus) and

234

# [accelerators for online

235

# prediction](/ml-engine/docs/machine-types-online-prediction#gpus).

236

"type": "A String", # The type of accelerator to use.

237

"count": "A String", # The number of accelerators to attach to each machine running the job.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

238

},

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

239

"state": "A String", # Output only. The state of a version.

240

"name": "A String", # Required. The name specified for the version when it was created.

241

#

242

# The version name must be unique within the model it is created in.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

243

"autoScaling": { # Options for automatically scaling a model. # Automatically scale the number of nodes used to serve the model in

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

244

# response to increases and decreases in traffic. Care should be

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

245

# taken to ramp up traffic according to the model's ability to scale

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

246

# or you will start seeing increases in latency and 429 response codes.

247

#

248

# Note that you cannot use AutoScaling if your version uses

249

# [GPUs](#Version.FIELDS.accelerator_config). Instead, you must use specify

250

# `manual_scaling`.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

251

"minNodes": 42, # Optional. The minimum number of nodes to allocate for this model. These

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

252

# nodes are always up, starting from the time the model is deployed.

253

# Therefore, the cost of operating this model will be at least

254

# `rate` * `min_nodes` * number of hours since last billing cycle,

255

# where `rate` is the cost per node-hour as documented in the

256

# [pricing guide](/ml-engine/docs/pricing),

257

# even if no predictions are performed. There is additional cost for each

258

# prediction performed.

259

#

260

# Unlike manual scaling, if the load gets too heavy for the nodes

261

# that are up, the service will automatically add nodes to handle the

262

# increased load as well as scale back as traffic drops, always maintaining

263

# at least `min_nodes`. You will be charged for the time in which additional

264

# nodes are used.

265

#

266

# If `min_nodes` is not specified and AutoScaling is used with a [legacy

267

# (MLS1) machine type](/ml-engine/docs/machine-types-online-prediction),

268

# `min_nodes` defaults to 0, in which case, when traffic to a model stops

269

# (and after a cool-down period), nodes will be shut down and no charges will

270

# be incurred until traffic to the model resumes.

271

#

272

# If `min_nodes` is not specified and AutoScaling is used with a [Compute

273

# Engine (N1) machine type](/ml-engine/docs/machine-types-online-prediction),

274

# `min_nodes` defaults to 1. `min_nodes` must be at least 1 for use with a

275

# Compute Engine machine type.

276

#

277

# Note that you cannot use AutoScaling if your version uses

278

# [GPUs](#Version.FIELDS.accelerator_config). Instead, you must use

279

# ManualScaling.

280

#

281

# You can set `min_nodes` when creating the model version, and you can also

282

# update `min_nodes` for an existing version:

283

# <pre>

284

# update_body.json:

285

# {

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

286

# 'autoScaling': {

287

# 'minNodes': 5

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

# }

# }

# </pre>

# HTTP request:

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

292

# <pre style="max-width: 626px;">

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

293

# PATCH

294

# https://ml.googleapis.com/v1/{name=projects/*/models/*/versions/*}?update_mask=autoScaling.minNodes

295

# -d @./update_body.json

296

# </pre>

297

},

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

298

"explanationConfig": { # Message holding configuration options for explaining model predictions. # Optional. Configures explainability features on the model's version.

299

# Some explanation features require additional metadata to be loaded

300

# as part of the model payload.

301

# There are two feature attribution methods supported for TensorFlow models:

302

# integrated gradients and sampled Shapley.

303

# [Learn more about feature

304

# attributions.](/ai-platform/prediction/docs/ai-explanations/overview)

305

"integratedGradientsAttribution": { # Attributes credit by computing the Aumann-Shapley value taking advantage # Attributes credit by computing the Aumann-Shapley value taking advantage

306

# of the model's fully differentiable structure. Refer to this paper for

307

# more details: http://proceedings.mlr.press/v70/sundararajan17a.html

308

# of the model's fully differentiable structure. Refer to this paper for

309

# more details: https://arxiv.org/abs/1703.01365

310

"numIntegralSteps": 42, # Number of steps for approximating the path integral.

311

# A good value to start is 50 and gradually increase until the

312

# sum to diff property is met within the desired error range.

313

},

314

"xraiAttribution": { # Attributes credit by computing the XRAI taking advantage # Attributes credit by computing the XRAI taking advantage

315

# of the model's fully differentiable structure. Refer to this paper for

316

# more details: https://arxiv.org/abs/1906.02825

317

# Currently only implemented for models with natural image inputs.

318

# of the model's fully differentiable structure. Refer to this paper for

319

# more details: https://arxiv.org/abs/1906.02825

320

# Currently only implemented for models with natural image inputs.

321

"numIntegralSteps": 42, # Number of steps for approximating the path integral.

322

# A good value to start is 50 and gradually increase until the

323

# sum to diff property is met within the desired error range.

324

},

325

"sampledShapleyAttribution": { # An attribution method that approximates Shapley values for features that # An attribution method that approximates Shapley values for features that

326

# contribute to the label being predicted. A sampling strategy is used to

327

# approximate the value rather than considering all subsets of features.

328

# contribute to the label being predicted. A sampling strategy is used to

329

# approximate the value rather than considering all subsets of features.

330

"numPaths": 42, # The number of feature permutations to consider when approximating the

331

# Shapley values.

332

},

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

333

},

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

334

"pythonVersion": "A String", # Required. The version of Python used in prediction.

335

#

336

# The following Python versions are available:

337

#

338

# * Python '3.7' is available when `runtime_version` is set to '1.15' or

339

# later.

340

# * Python '3.5' is available when `runtime_version` is set to a version

341

# from '1.4' to '1.14'.

342

# * Python '2.7' is available when `runtime_version` is set to '1.15' or

343

# earlier.

344

#

345

# Read more about the Python versions available for [each runtime

346

# version](/ml-engine/docs/runtime-version-list).

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

347

"requestLoggingConfig": { # Configuration for logging request-response pairs to a BigQuery table. # Optional. *Only* specify this field in a

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

348

# projects.models.versions.patch

349

# request. Specifying it in a

350

# projects.models.versions.create

351

# request has no effect.

352

#

353

# Configures the request-response pair logging on predictions from this

354

# Version.

355

# Online prediction requests to a model version and the responses to these

356

# requests are converted to raw strings and saved to the specified BigQuery

357

# table. Logging is constrained by [BigQuery quotas and

358

# limits](/bigquery/quotas). If your project exceeds BigQuery quotas or limits,

359

# AI Platform Prediction does not log request-response pairs, but it continues

360

# to serve predictions.

361

#

362

# If you are using [continuous

363

# evaluation](/ml-engine/docs/continuous-evaluation/), you do not need to

364

# specify this configuration manually. Setting up continuous evaluation

365

# automatically enables logging of request-response pairs.

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

366

"samplingPercentage": 3.14, # Percentage of requests to be logged, expressed as a fraction from 0 to 1.

367

# For example, if you want to log 10% of requests, enter `0.1`. The sampling

368

# window is the lifetime of the model version. Defaults to 0.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

369

"bigqueryTableName": "A String", # Required. Fully qualified BigQuery table name in the following format:

370

# "<var>project_id</var>.<var>dataset_name</var>.<var>table_name</var>"

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

371

#

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

372

# The specified table must already exist, and the "Cloud ML Service Agent"

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

373

# for your project must have permission to write to it. The table must have

374

# the following [schema](/bigquery/docs/schemas):

375

#

376

# <table>

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

377

# <tr><th>Field name</th><th style="display: table-cell">Type</th>

378

# <th style="display: table-cell">Mode</th></tr>

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

379

# <tr><td>model</td><td>STRING</td><td>REQUIRED</td></tr>

380

# <tr><td>model_version</td><td>STRING</td><td>REQUIRED</td></tr>

381

# <tr><td>time</td><td>TIMESTAMP</td><td>REQUIRED</td></tr>

382

# <tr><td>raw_data</td><td>STRING</td><td>REQUIRED</td></tr>

383

# <tr><td>raw_prediction</td><td>STRING</td><td>NULLABLE</td></tr>

384

# <tr><td>groundtruth</td><td>STRING</td><td>NULLABLE</td></tr>

385

# </table>

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

386

},

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

387

"manualScaling": { # Options for manually scaling a model. # Manually select the number of nodes to use for serving the

388

# model. You should generally use `auto_scaling` with an appropriate

389

# `min_nodes` instead, but this option is available if you want more

390

# predictable billing. Beware that latency and error rates will increase

391

# if the traffic exceeds that capability of the system to serve it based

392

# on the selected number of nodes.

393

"nodes": 42, # The number of nodes to allocate for this model. These nodes are always up,

394

# starting from the time the model is deployed, so the cost of operating

395

# this model will be proportional to `nodes` * number of hours since

396

# last billing cycle plus the cost for each prediction performed.

397

},

398

"createTime": "A String", # Output only. The time the version was created.

399

"lastUseTime": "A String", # Output only. The time the version was last used for prediction.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

400

"framework": "A String", # Optional. The machine learning framework AI Platform uses to train

401

# this version of the model. Valid values are `TENSORFLOW`, `SCIKIT_LEARN`,

402

# `XGBOOST`. If you do not specify a framework, AI Platform

403

# will analyze files in the deployment_uri to determine a framework. If you

404

# choose `SCIKIT_LEARN` or `XGBOOST`, you must also set the runtime version

405

# of the model to 1.4 or greater.

406

#

407

# Do **not** specify a framework if you're deploying a [custom

408

# prediction routine](/ml-engine/docs/tensorflow/custom-prediction-routines).

409

#

410

# If you specify a [Compute Engine (N1) machine

411

# type](/ml-engine/docs/machine-types-online-prediction) in the

412

# `machineType` field, you must specify `TENSORFLOW`

413

# for the framework.

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

414

"predictionClass": "A String", # Optional. The fully qualified name

415

# (<var>module_name</var>.<var>class_name</var>) of a class that implements

416

# the Predictor interface described in this reference field. The module

417

# containing this class should be included in a package provided to the

418

# [`packageUris` field](#Version.FIELDS.package_uris).

419

#

420

# Specify this field if and only if you are deploying a [custom prediction

421

# routine (beta)](/ml-engine/docs/tensorflow/custom-prediction-routines).

422

# If you specify this field, you must set

423

# [`runtimeVersion`](#Version.FIELDS.runtime_version) to 1.4 or greater and

424

# you must set `machineType` to a [legacy (MLS1)

425

# machine type](/ml-engine/docs/machine-types-online-prediction).

426

#

427

# The following code sample provides the Predictor interface:

428

#

429

# <pre style="max-width: 626px;">

430

# class Predictor(object):

431

# """Interface for constructing custom predictors."""

432

#

433

# def predict(self, instances, **kwargs):

434

# """Performs custom prediction.

435

#

436

# Instances are the decoded values from the request. They have already

437

# been deserialized from JSON.

438

#

439

# Args:

440

# instances: A list of prediction input instances.

441

# **kwargs: A dictionary of keyword args provided as additional

442

# fields on the predict request body.

443

#

444

# Returns:

445

# A list of outputs containing the prediction results. This list must

446

# be JSON serializable.

447

# """

448

# raise NotImplementedError()

449

#

450

# @classmethod

451

# def from_path(cls, model_dir):

452

# """Creates an instance of Predictor using the given path.

453

#

454

# Loading of the predictor should be done in this method.

455

#

456

# Args:

457

# model_dir: The local directory that contains the exported model

458

# file along with any additional files uploaded when creating the

# version resource.

#

# Returns:

# An instance implementing this Predictor class.

463

# """

464

# raise NotImplementedError()

465

# </pre>

466

#

467

# Learn more about [the Predictor interface and custom prediction

468

# routines](/ml-engine/docs/tensorflow/custom-prediction-routines).

469

"isDefault": True or False, # Output only. If true, this version will be used to handle prediction

470

# requests that do not specify a version.

471

#

472

# You can change the default version by calling

473

# projects.methods.versions.setDefault.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

474

"etag": "A String", # `etag` is used for optimistic concurrency control as a way to help

475

# prevent simultaneous updates of a model from overwriting each other.

476

# It is strongly suggested that systems make use of the `etag` in the

477

# read-modify-write cycle to perform model updates in order to avoid race

478

# conditions: An `etag` is returned in the response to `GetVersion`, and

479

# systems are expected to put that etag in the request to `UpdateVersion` to

480

# ensure that their change will be applied to the model as intended.

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

481

"serviceAccount": "A String", # Optional. Specifies the service account for resource access control.

482

"errorMessage": "A String", # Output only. The details of a failure or a cancellation.

483

"deploymentUri": "A String", # Required. The Cloud Storage location of the trained model used to

484

# create the version. See the

485

# [guide to model

486

# deployment](/ml-engine/docs/tensorflow/deploying-models) for more

487

# information.

488

#

489

# When passing Version to

490

# projects.models.versions.create

491

# the model service uses the specified location as the source of the model.

492

# Once deployed, the model version is hosted by the prediction service, so

493

# this location is useful only as a historical record.

494

# The total number of model files can't exceed 1000.

495

"runtimeVersion": "A String", # Required. The AI Platform runtime version to use for this deployment.

496

#

497

# For more information, see the

498

# [runtime version list](/ml-engine/docs/runtime-version-list) and

499

# [how to manage runtime versions](/ml-engine/docs/versioning).

500

"description": "A String", # Optional. The description specified for the version when it was created.

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

501

},

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

502

"onlinePredictionLogging": True or False, # Optional. If true, online prediction access logs are sent to StackDriver

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

503

# Logging. These logs are like standard server access logs, containing

504

# information like timestamp and latency for each request. Note that

505

# [Stackdriver logs may incur a cost](/stackdriver/pricing), especially if

506

# your project receives prediction requests at a high queries per second rate

507

# (QPS). Estimate your costs before enabling this option.

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

508

#

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

509

# Default is false.

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

510

}

511

512

x__xgafv: string, V1 error format.

Allowed values

1 - v1 error format

2 - v2 error format

Returns:

An object of the form:

519

520

{ # Represents a machine learning solution.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

521

#

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

522

# A model can have multiple versions, each of which is a deployed, trained

523

# model ready to receive prediction requests. The model itself is just a

524

# container.

525

"description": "A String", # Optional. The description specified for the model when it was created.

526

"regions": [ # Optional. The list of regions where the model is going to be deployed.

527

# Only one region per model is supported.

528

# Defaults to 'us-central1' if nothing is set.

529

# See the <a href="/ml-engine/docs/tensorflow/regions">available regions</a>

530

# for AI Platform services.

531

# Note:

532

# * No matter where a model is deployed, it can always be accessed by

533

# users from anywhere, both for online and batch prediction.

534

# * The region for a batch prediction job is set by the region field when

535

# submitting the batch prediction job and does not take its value from

# this field.

"A String",

],

"name": "A String", # Required. The name specified for the model when it was created.

540

#

541

# The model name must be unique within the project it is created in.

542

"onlinePredictionConsoleLogging": True or False, # Optional. If true, online prediction nodes send `stderr` and `stdout`

543

# streams to Stackdriver Logging. These can be more verbose than the standard

544

# access logs (see `onlinePredictionLogging`) and can incur higher cost.

545

# However, they are helpful for debugging. Note that

546

# [Stackdriver logs may incur a cost](/stackdriver/pricing), especially if

547

# your project receives prediction requests at a high QPS. Estimate your

548

# costs before enabling this option.

549

#

550

# Default is false.

551

"etag": "A String", # `etag` is used for optimistic concurrency control as a way to help

552

# prevent simultaneous updates of a model from overwriting each other.

553

# It is strongly suggested that systems make use of the `etag` in the

554

# read-modify-write cycle to perform model updates in order to avoid race

555

# conditions: An `etag` is returned in the response to `GetModel`, and

556

# systems are expected to put that etag in the request to `UpdateModel` to

557

# ensure that their change will be applied to the model as intended.

558

"labels": { # Optional. One or more labels that you can add, to organize your models.

559

# Each label is a key-value pair, where both the key and the value are

560

# arbitrary strings that you supply.

561

# For more information, see the documentation on

562

# <a href="/ml-engine/docs/tensorflow/resource-labels">using labels</a>.

563

"a_key": "A String",

564

},

565

"defaultVersion": { # Represents a version of the model. # Output only. The default version of the model. This version will be used to

566

# handle prediction requests that do not specify a version.

567

#

568

# You can change the default version by calling

569

# projects.models.versions.setDefault.

570

#

571

# Each version is a trained model deployed in the cloud, ready to handle

572

# prediction requests. A model can have multiple versions. You can get

573

# information about all of the versions of a given model by calling

574

# projects.models.versions.list.

575

"labels": { # Optional. One or more labels that you can add, to organize your model

576

# versions. Each label is a key-value pair, where both the key and the value

577

# are arbitrary strings that you supply.

578

# For more information, see the documentation on

579

# <a href="/ml-engine/docs/tensorflow/resource-labels">using labels</a>.

580

"a_key": "A String",

581

},

582

"machineType": "A String", # Optional. The type of machine on which to serve the model. Currently only

583

# applies to online prediction service. If this field is not specified, it

584

# defaults to `mls1-c1-m2`.

585

#

586

# Online prediction supports the following machine types:

#

# * `mls1-c1-m2`

# * `mls1-c4-m2`

# * `n1-standard-2`

# * `n1-standard-4`

# * `n1-standard-8`

# * `n1-standard-16`

# * `n1-standard-32`

# * `n1-highmem-2`

# * `n1-highmem-4`

# * `n1-highmem-8`

# * `n1-highmem-16`

# * `n1-highmem-32`

# * `n1-highcpu-2`

# * `n1-highcpu-4`

# * `n1-highcpu-8`

# * `n1-highcpu-16`

# * `n1-highcpu-32`

#

# `mls1-c1-m2` is generally available. All other machine types are available

607

# in beta. Learn more about the [differences between machine

608

# types](/ml-engine/docs/machine-types-online-prediction).

609

"packageUris": [ # Optional. Cloud Storage paths (`gs://…`) of packages for [custom

610

# prediction routines](/ml-engine/docs/tensorflow/custom-prediction-routines)

611

# or [scikit-learn pipelines with custom

612

# code](/ml-engine/docs/scikit/exporting-for-prediction#custom-pipeline-code).

613

#

614

# For a custom prediction routine, one of these packages must contain your

615

# Predictor class (see

616

# [`predictionClass`](#Version.FIELDS.prediction_class)). Additionally,

617

# include any dependencies used by your Predictor or scikit-learn pipeline

618

# uses that are not already included in your selected [runtime

619

# version](/ml-engine/docs/tensorflow/runtime-version-list).

620

#

621

# If you specify this field, you must also set

622

# [`runtimeVersion`](#Version.FIELDS.runtime_version) to 1.4 or greater.

623

"A String",

624

],

625

"acceleratorConfig": { # Represents a hardware accelerator request config. # Optional. Accelerator config for using GPUs for online prediction (beta).

626

# Only specify this field if you have specified a Compute Engine (N1) machine

627

# type in the `machineType` field. Learn more about [using GPUs for online

628

# prediction](/ml-engine/docs/machine-types-online-prediction#gpus).

629

# Note that the AcceleratorConfig can be used in both Jobs and Versions.

630

# Learn more about [accelerators for training](/ml-engine/docs/using-gpus) and

631

# [accelerators for online

632

# prediction](/ml-engine/docs/machine-types-online-prediction#gpus).

633

"type": "A String", # The type of accelerator to use.

634

"count": "A String", # The number of accelerators to attach to each machine running the job.

635

},

636

"state": "A String", # Output only. The state of a version.

637

"name": "A String", # Required. The name specified for the version when it was created.

638

#

639

# The version name must be unique within the model it is created in.

640

"autoScaling": { # Options for automatically scaling a model. # Automatically scale the number of nodes used to serve the model in

641

# response to increases and decreases in traffic. Care should be

642

# taken to ramp up traffic according to the model's ability to scale

643

# or you will start seeing increases in latency and 429 response codes.

644

#

645

# Note that you cannot use AutoScaling if your version uses

646

# [GPUs](#Version.FIELDS.accelerator_config). Instead, you must use specify

647

# `manual_scaling`.

648

"minNodes": 42, # Optional. The minimum number of nodes to allocate for this model. These

649

# nodes are always up, starting from the time the model is deployed.

650

# Therefore, the cost of operating this model will be at least

651

# `rate` * `min_nodes` * number of hours since last billing cycle,

652

# where `rate` is the cost per node-hour as documented in the

653

# [pricing guide](/ml-engine/docs/pricing),

654

# even if no predictions are performed. There is additional cost for each

655

# prediction performed.

656

#

657

# Unlike manual scaling, if the load gets too heavy for the nodes

658

# that are up, the service will automatically add nodes to handle the

659

# increased load as well as scale back as traffic drops, always maintaining

660

# at least `min_nodes`. You will be charged for the time in which additional

661

# nodes are used.

662

#

663

# If `min_nodes` is not specified and AutoScaling is used with a [legacy

664

# (MLS1) machine type](/ml-engine/docs/machine-types-online-prediction),

665

# `min_nodes` defaults to 0, in which case, when traffic to a model stops

666

# (and after a cool-down period), nodes will be shut down and no charges will

667

# be incurred until traffic to the model resumes.

668

#

669

# If `min_nodes` is not specified and AutoScaling is used with a [Compute

670

# Engine (N1) machine type](/ml-engine/docs/machine-types-online-prediction),

671

# `min_nodes` defaults to 1. `min_nodes` must be at least 1 for use with a

672

# Compute Engine machine type.

673

#

674

# Note that you cannot use AutoScaling if your version uses

675

# [GPUs](#Version.FIELDS.accelerator_config). Instead, you must use

676

# ManualScaling.

677

#

678

# You can set `min_nodes` when creating the model version, and you can also

679

# update `min_nodes` for an existing version:

# <pre>

# update_body.json:

# {

# 'autoScaling': {

# 'minNodes': 5

# }

# }

# </pre>

# HTTP request:

# <pre style="max-width: 626px;">

690

# PATCH

691

# https://ml.googleapis.com/v1/{name=projects/*/models/*/versions/*}?update_mask=autoScaling.minNodes

692

# -d @./update_body.json

693

# </pre>

694

},

695

"explanationConfig": { # Message holding configuration options for explaining model predictions. # Optional. Configures explainability features on the model's version.

696

# Some explanation features require additional metadata to be loaded

697

# as part of the model payload.

698

# There are two feature attribution methods supported for TensorFlow models:

699

# integrated gradients and sampled Shapley.

700

# [Learn more about feature

701

# attributions.](/ai-platform/prediction/docs/ai-explanations/overview)

702

"integratedGradientsAttribution": { # Attributes credit by computing the Aumann-Shapley value taking advantage # Attributes credit by computing the Aumann-Shapley value taking advantage

703

# of the model's fully differentiable structure. Refer to this paper for

704

# more details: http://proceedings.mlr.press/v70/sundararajan17a.html

705

# of the model's fully differentiable structure. Refer to this paper for

706

# more details: https://arxiv.org/abs/1703.01365

707

"numIntegralSteps": 42, # Number of steps for approximating the path integral.

708

# A good value to start is 50 and gradually increase until the

709

# sum to diff property is met within the desired error range.

710

},

711

"xraiAttribution": { # Attributes credit by computing the XRAI taking advantage # Attributes credit by computing the XRAI taking advantage

712

# of the model's fully differentiable structure. Refer to this paper for

713

# more details: https://arxiv.org/abs/1906.02825

714

# Currently only implemented for models with natural image inputs.

715

# of the model's fully differentiable structure. Refer to this paper for

716

# more details: https://arxiv.org/abs/1906.02825

717

# Currently only implemented for models with natural image inputs.

718

"numIntegralSteps": 42, # Number of steps for approximating the path integral.

719

# A good value to start is 50 and gradually increase until the

720

# sum to diff property is met within the desired error range.

721

},

722

"sampledShapleyAttribution": { # An attribution method that approximates Shapley values for features that # An attribution method that approximates Shapley values for features that

723

# contribute to the label being predicted. A sampling strategy is used to

724

# approximate the value rather than considering all subsets of features.

725

# contribute to the label being predicted. A sampling strategy is used to

726

# approximate the value rather than considering all subsets of features.

727

"numPaths": 42, # The number of feature permutations to consider when approximating the

# Shapley values.

},

},

"pythonVersion": "A String", # Required. The version of Python used in prediction.

732

#

733

# The following Python versions are available:

734

#

735

# * Python '3.7' is available when `runtime_version` is set to '1.15' or

736

# later.

737

# * Python '3.5' is available when `runtime_version` is set to a version

738

# from '1.4' to '1.14'.

739

# * Python '2.7' is available when `runtime_version` is set to '1.15' or

740

# earlier.

741

#

742

# Read more about the Python versions available for [each runtime

743

# version](/ml-engine/docs/runtime-version-list).

744

"requestLoggingConfig": { # Configuration for logging request-response pairs to a BigQuery table. # Optional. *Only* specify this field in a

745

# projects.models.versions.patch

746

# request. Specifying it in a

747

# projects.models.versions.create

748

# request has no effect.

749

#

750

# Configures the request-response pair logging on predictions from this

751

# Version.

752

# Online prediction requests to a model version and the responses to these

753

# requests are converted to raw strings and saved to the specified BigQuery

754

# table. Logging is constrained by [BigQuery quotas and

755

# limits](/bigquery/quotas). If your project exceeds BigQuery quotas or limits,

756

# AI Platform Prediction does not log request-response pairs, but it continues

757

# to serve predictions.

758

#

759

# If you are using [continuous

760

# evaluation](/ml-engine/docs/continuous-evaluation/), you do not need to

761

# specify this configuration manually. Setting up continuous evaluation

762

# automatically enables logging of request-response pairs.

763

"samplingPercentage": 3.14, # Percentage of requests to be logged, expressed as a fraction from 0 to 1.

764

# For example, if you want to log 10% of requests, enter `0.1`. The sampling

765

# window is the lifetime of the model version. Defaults to 0.

766

"bigqueryTableName": "A String", # Required. Fully qualified BigQuery table name in the following format:

767

# "<var>project_id</var>.<var>dataset_name</var>.<var>table_name</var>"

768

#

769

# The specified table must already exist, and the "Cloud ML Service Agent"

770

# for your project must have permission to write to it. The table must have

771

# the following [schema](/bigquery/docs/schemas):

772

#

773

# <table>

774

# <tr><th>Field name</th><th style="display: table-cell">Type</th>

775

# <th style="display: table-cell">Mode</th></tr>

776

# <tr><td>model</td><td>STRING</td><td>REQUIRED</td></tr>

777

# <tr><td>model_version</td><td>STRING</td><td>REQUIRED</td></tr>

778

# <tr><td>time</td><td>TIMESTAMP</td><td>REQUIRED</td></tr>

779

# <tr><td>raw_data</td><td>STRING</td><td>REQUIRED</td></tr>

780

# <tr><td>raw_prediction</td><td>STRING</td><td>NULLABLE</td></tr>

781

# <tr><td>groundtruth</td><td>STRING</td><td>NULLABLE</td></tr>

782

# </table>

783

},

784

"manualScaling": { # Options for manually scaling a model. # Manually select the number of nodes to use for serving the

785

# model. You should generally use `auto_scaling` with an appropriate

786

# `min_nodes` instead, but this option is available if you want more

787

# predictable billing. Beware that latency and error rates will increase

788

# if the traffic exceeds that capability of the system to serve it based

789

# on the selected number of nodes.

790

"nodes": 42, # The number of nodes to allocate for this model. These nodes are always up,

791

# starting from the time the model is deployed, so the cost of operating

792

# this model will be proportional to `nodes` * number of hours since

793

# last billing cycle plus the cost for each prediction performed.

794

},

795

"createTime": "A String", # Output only. The time the version was created.

796

"lastUseTime": "A String", # Output only. The time the version was last used for prediction.

797

"framework": "A String", # Optional. The machine learning framework AI Platform uses to train

798

# this version of the model. Valid values are `TENSORFLOW`, `SCIKIT_LEARN`,

799

# `XGBOOST`. If you do not specify a framework, AI Platform

800

# will analyze files in the deployment_uri to determine a framework. If you

801

# choose `SCIKIT_LEARN` or `XGBOOST`, you must also set the runtime version

802

# of the model to 1.4 or greater.

803

#

804

# Do **not** specify a framework if you're deploying a [custom

805

# prediction routine](/ml-engine/docs/tensorflow/custom-prediction-routines).

806

#

807

# If you specify a [Compute Engine (N1) machine

808

# type](/ml-engine/docs/machine-types-online-prediction) in the

809

# `machineType` field, you must specify `TENSORFLOW`

810

# for the framework.

811

"predictionClass": "A String", # Optional. The fully qualified name

812

# (<var>module_name</var>.<var>class_name</var>) of a class that implements

813

# the Predictor interface described in this reference field. The module

814

# containing this class should be included in a package provided to the

815

# [`packageUris` field](#Version.FIELDS.package_uris).

816

#

817

# Specify this field if and only if you are deploying a [custom prediction

818

# routine (beta)](/ml-engine/docs/tensorflow/custom-prediction-routines).

819

# If you specify this field, you must set

820

# [`runtimeVersion`](#Version.FIELDS.runtime_version) to 1.4 or greater and

821

# you must set `machineType` to a [legacy (MLS1)

822

# machine type](/ml-engine/docs/machine-types-online-prediction).

823

#

824

# The following code sample provides the Predictor interface:

825

#

826

# <pre style="max-width: 626px;">

827

# class Predictor(object):

828

# """Interface for constructing custom predictors."""

829

#

830

# def predict(self, instances, **kwargs):

831

# """Performs custom prediction.

832

#

833

# Instances are the decoded values from the request. They have already

834

# been deserialized from JSON.

835

#

836

# Args:

837

# instances: A list of prediction input instances.

838

# **kwargs: A dictionary of keyword args provided as additional

839

# fields on the predict request body.

840

#

841

# Returns:

842

# A list of outputs containing the prediction results. This list must

843

# be JSON serializable.

844

# """

845

# raise NotImplementedError()

846

#

847

# @classmethod

848

# def from_path(cls, model_dir):

849

# """Creates an instance of Predictor using the given path.

850

#

851

# Loading of the predictor should be done in this method.

852

#

853

# Args:

854

# model_dir: The local directory that contains the exported model

855

# file along with any additional files uploaded when creating the

# version resource.

#

# Returns:

# An instance implementing this Predictor class.

860

# """

861

# raise NotImplementedError()

862

# </pre>

863

#

864

# Learn more about [the Predictor interface and custom prediction

865

# routines](/ml-engine/docs/tensorflow/custom-prediction-routines).

866

"isDefault": True or False, # Output only. If true, this version will be used to handle prediction

867

# requests that do not specify a version.

868

#

869

# You can change the default version by calling

870

# projects.methods.versions.setDefault.

871

"etag": "A String", # `etag` is used for optimistic concurrency control as a way to help

872

# prevent simultaneous updates of a model from overwriting each other.

873

# It is strongly suggested that systems make use of the `etag` in the

874

# read-modify-write cycle to perform model updates in order to avoid race

875

# conditions: An `etag` is returned in the response to `GetVersion`, and

876

# systems are expected to put that etag in the request to `UpdateVersion` to

877

# ensure that their change will be applied to the model as intended.

878

"serviceAccount": "A String", # Optional. Specifies the service account for resource access control.

879

"errorMessage": "A String", # Output only. The details of a failure or a cancellation.

880

"deploymentUri": "A String", # Required. The Cloud Storage location of the trained model used to

881

# create the version. See the

882

# [guide to model

883

# deployment](/ml-engine/docs/tensorflow/deploying-models) for more

884

# information.

885

#

886

# When passing Version to

887

# projects.models.versions.create

888

# the model service uses the specified location as the source of the model.

889

# Once deployed, the model version is hosted by the prediction service, so

890

# this location is useful only as a historical record.

891

# The total number of model files can't exceed 1000.

892

"runtimeVersion": "A String", # Required. The AI Platform runtime version to use for this deployment.

893

#

894

# For more information, see the

895

# [runtime version list](/ml-engine/docs/runtime-version-list) and

896

# [how to manage runtime versions](/ml-engine/docs/versioning).

897

"description": "A String", # Optional. The description specified for the version when it was created.

898

},

899

"onlinePredictionLogging": True or False, # Optional. If true, online prediction access logs are sent to StackDriver

900

# Logging. These logs are like standard server access logs, containing

901

# information like timestamp and latency for each request. Note that

902

# [Stackdriver logs may incur a cost](/stackdriver/pricing), especially if

903

# your project receives prediction requests at a high queries per second rate

904

# (QPS). Estimate your costs before enabling this option.

905

#

906

# Default is false.

907

}</pre>

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

</div>

Thomas Coffee

2017-03-27 10:39:26 -0700

[diff] [blame]

911

<code class="details" id="delete">delete(name, x__xgafv=None)</code>

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

912

<pre>Deletes a model.

913

914

You can only delete a model if there are no versions in it. You can delete

915

versions by calling

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

916

projects.models.versions.delete.

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

917

918

Args:

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

919

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

920

x__xgafv: string, V1 error format.

Allowed values

1 - v1 error format

2 - v2 error format

Returns:

An object of the form:

927

928

{ # This resource represents a long-running operation that is the result of a

929

# network API call.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

930

"error": { # The `Status` type defines a logical error model that is suitable for # The error result of the operation in case of failure or cancellation.

931

# different programming environments, including REST APIs and RPC APIs. It is

932

# used by [gRPC](https://github.com/grpc). Each `Status` message contains

933

# three pieces of data: error code, error message, and error details.

934

#

935

# You can find out more about this error model and how to work with it in the

936

# [API Design Guide](https://cloud.google.com/apis/design/errors).

937

"details": [ # A list of messages that carry the error details. There is a common set of

938

# message types for APIs to use.

939

{

940

"a_key": "", # Properties of the object. Contains field @type with type URL.

941

},

942

],

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

943

"message": "A String", # A developer-facing error message, which should be in English. Any

944

# user-facing error message should be localized and sent in the

945

# google.rpc.Status.details field, or localized by the client.

946

"code": 42, # The status code, which should be an enum value of google.rpc.Code.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

947

},

948

"done": True or False, # If the value is `false`, it means the operation is still in progress.

949

# If `true`, the operation is completed, and either `error` or `response` is

950

# available.

Bu Sun Kim

2020-05-27 12:20:54 -0700

[diff] [blame]

951

"response": { # The normal response of the operation in case of success. If the original

952

# method returns no data on success, such as `Delete`, the response is

953

# `google.protobuf.Empty`. If the original method is standard

954

# `Get`/`Create`/`Update`, the response should be the resource. For other

955

# methods, the response should have the type `XxxResponse`, where `Xxx`

956

# is the original method name. For example, if the original method name

957

# is `TakeSnapshot()`, the inferred response type is

958

# `TakeSnapshotResponse`.

959

"a_key": "", # Properties of the object. Contains field @type with type URL.

960

},

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

961

"metadata": { # Service-specific metadata associated with the operation. It typically

962

# contains progress information and common metadata such as create time.

963

# Some services might not provide such metadata. Any method that returns a

964

# long-running operation should document the metadata type, if any.

965

"a_key": "", # Properties of the object. Contains field @type with type URL.

966

},

967

"name": "A String", # The server-assigned name, which is only unique within the same service that

968

# originally returns it. If you use the default HTTP mapping, the

969

# `name` should be a resource name ending with `operations/{unique_id}`.

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

}</pre>

</div>

Thomas Coffee

2017-03-27 10:39:26 -0700

[diff] [blame]

974

<code class="details" id="get">get(name, x__xgafv=None)</code>

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

975

<pre>Gets information about a model, including its name, the description (if

976

set), and the default version (if at least one version of the model has

977

been deployed).

978

979

Args:

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

980

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

981

x__xgafv: string, V1 error format.

Allowed values

1 - v1 error format

2 - v2 error format

Returns:

An object of the form:

988

989

{ # Represents a machine learning solution.

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

990

#

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

991

# A model can have multiple versions, each of which is a deployed, trained

992

# model ready to receive prediction requests. The model itself is just a

993

# container.

994

"description": "A String", # Optional. The description specified for the model when it was created.

995

"regions": [ # Optional. The list of regions where the model is going to be deployed.

996

# Only one region per model is supported.

997

# Defaults to 'us-central1' if nothing is set.

998

# See the <a href="/ml-engine/docs/tensorflow/regions">available regions</a>

999

# for AI Platform services.

1000

# Note:

1001

# * No matter where a model is deployed, it can always be accessed by

1002

# users from anywhere, both for online and batch prediction.

1003

# * The region for a batch prediction job is set by the region field when

1004

# submitting the batch prediction job and does not take its value from

1005

# this field.

Bu Sun Kim

2020-05-27 12:20:54 -0700

[diff] [blame]

1006

"A String",

1007

],

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1008

"name": "A String", # Required. The name specified for the model when it was created.

Bu Sun Kim

2020-05-27 12:20:54 -0700

[diff] [blame]

1009

#

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1010

# The model name must be unique within the project it is created in.

1011

"onlinePredictionConsoleLogging": True or False, # Optional. If true, online prediction nodes send `stderr` and `stdout`

1012

# streams to Stackdriver Logging. These can be more verbose than the standard

1013

# access logs (see `onlinePredictionLogging`) and can incur higher cost.

1014

# However, they are helpful for debugging. Note that

1015

# [Stackdriver logs may incur a cost](/stackdriver/pricing), especially if

1016

# your project receives prediction requests at a high QPS. Estimate your

1017

# costs before enabling this option.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1018

#

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1019

# Default is false.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1020

"etag": "A String", # `etag` is used for optimistic concurrency control as a way to help

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1021

# prevent simultaneous updates of a model from overwriting each other.

1022

# It is strongly suggested that systems make use of the `etag` in the

1023

# read-modify-write cycle to perform model updates in order to avoid race

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1024

# conditions: An `etag` is returned in the response to `GetModel`, and

1025

# systems are expected to put that etag in the request to `UpdateModel` to

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1026

# ensure that their change will be applied to the model as intended.

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1027

"labels": { # Optional. One or more labels that you can add, to organize your models.

1028

# Each label is a key-value pair, where both the key and the value are

1029

# arbitrary strings that you supply.

1030

# For more information, see the documentation on

1031

# <a href="/ml-engine/docs/tensorflow/resource-labels">using labels</a>.

1032

"a_key": "A String",

1033

},

1034

"defaultVersion": { # Represents a version of the model. # Output only. The default version of the model. This version will be used to

1035

# handle prediction requests that do not specify a version.

1036

#

1037

# You can change the default version by calling

1038

# projects.models.versions.setDefault.

1039

#

1040

# Each version is a trained model deployed in the cloud, ready to handle

1041

# prediction requests. A model can have multiple versions. You can get

1042

# information about all of the versions of a given model by calling

1043

# projects.models.versions.list.

1044

"labels": { # Optional. One or more labels that you can add, to organize your model

1045

# versions. Each label is a key-value pair, where both the key and the value

1046

# are arbitrary strings that you supply.

1047

# For more information, see the documentation on

1048

# <a href="/ml-engine/docs/tensorflow/resource-labels">using labels</a>.

1049

"a_key": "A String",

1050

},

1051

"machineType": "A String", # Optional. The type of machine on which to serve the model. Currently only

1052

# applies to online prediction service. If this field is not specified, it

1053

# defaults to `mls1-c1-m2`.

1054

#

1055

# Online prediction supports the following machine types:

#

# * `mls1-c1-m2`

# * `mls1-c4-m2`

# * `n1-standard-2`

# * `n1-standard-4`

# * `n1-standard-8`

# * `n1-standard-16`

# * `n1-standard-32`

# * `n1-highmem-2`

# * `n1-highmem-4`

# * `n1-highmem-8`

# * `n1-highmem-16`

# * `n1-highmem-32`

# * `n1-highcpu-2`

# * `n1-highcpu-4`

# * `n1-highcpu-8`

# * `n1-highcpu-16`

# * `n1-highcpu-32`

#

# `mls1-c1-m2` is generally available. All other machine types are available

1076

# in beta. Learn more about the [differences between machine

1077

# types](/ml-engine/docs/machine-types-online-prediction).

1078

"packageUris": [ # Optional. Cloud Storage paths (`gs://…`) of packages for [custom

1079

# prediction routines](/ml-engine/docs/tensorflow/custom-prediction-routines)

1080

# or [scikit-learn pipelines with custom

1081

# code](/ml-engine/docs/scikit/exporting-for-prediction#custom-pipeline-code).

1082

#

1083

# For a custom prediction routine, one of these packages must contain your

1084

# Predictor class (see

1085

# [`predictionClass`](#Version.FIELDS.prediction_class)). Additionally,

1086

# include any dependencies used by your Predictor or scikit-learn pipeline

1087

# uses that are not already included in your selected [runtime

1088

# version](/ml-engine/docs/tensorflow/runtime-version-list).

1089

#

1090

# If you specify this field, you must also set

1091

# [`runtimeVersion`](#Version.FIELDS.runtime_version) to 1.4 or greater.

1092

"A String",

1093

],

1094

"acceleratorConfig": { # Represents a hardware accelerator request config. # Optional. Accelerator config for using GPUs for online prediction (beta).

1095

# Only specify this field if you have specified a Compute Engine (N1) machine

1096

# type in the `machineType` field. Learn more about [using GPUs for online

1097

# prediction](/ml-engine/docs/machine-types-online-prediction#gpus).

1098

# Note that the AcceleratorConfig can be used in both Jobs and Versions.

1099

# Learn more about [accelerators for training](/ml-engine/docs/using-gpus) and

1100

# [accelerators for online

1101

# prediction](/ml-engine/docs/machine-types-online-prediction#gpus).

1102

"type": "A String", # The type of accelerator to use.

1103

"count": "A String", # The number of accelerators to attach to each machine running the job.

1104

},

1105

"state": "A String", # Output only. The state of a version.

1106

"name": "A String", # Required. The name specified for the version when it was created.

1107

#

1108

# The version name must be unique within the model it is created in.

1109

"autoScaling": { # Options for automatically scaling a model. # Automatically scale the number of nodes used to serve the model in

1110

# response to increases and decreases in traffic. Care should be

1111

# taken to ramp up traffic according to the model's ability to scale

1112

# or you will start seeing increases in latency and 429 response codes.

1113

#

1114

# Note that you cannot use AutoScaling if your version uses

1115

# [GPUs](#Version.FIELDS.accelerator_config). Instead, you must use specify

1116

# `manual_scaling`.

1117

"minNodes": 42, # Optional. The minimum number of nodes to allocate for this model. These

1118

# nodes are always up, starting from the time the model is deployed.

1119

# Therefore, the cost of operating this model will be at least

1120

# `rate` * `min_nodes` * number of hours since last billing cycle,

1121

# where `rate` is the cost per node-hour as documented in the

1122

# [pricing guide](/ml-engine/docs/pricing),

1123

# even if no predictions are performed. There is additional cost for each

1124

# prediction performed.

1125

#

1126

# Unlike manual scaling, if the load gets too heavy for the nodes

1127

# that are up, the service will automatically add nodes to handle the

1128

# increased load as well as scale back as traffic drops, always maintaining

1129

# at least `min_nodes`. You will be charged for the time in which additional

1130

# nodes are used.

1131

#

1132

# If `min_nodes` is not specified and AutoScaling is used with a [legacy

1133

# (MLS1) machine type](/ml-engine/docs/machine-types-online-prediction),

1134

# `min_nodes` defaults to 0, in which case, when traffic to a model stops

1135

# (and after a cool-down period), nodes will be shut down and no charges will

1136

# be incurred until traffic to the model resumes.

1137

#

1138

# If `min_nodes` is not specified and AutoScaling is used with a [Compute

1139

# Engine (N1) machine type](/ml-engine/docs/machine-types-online-prediction),

1140

# `min_nodes` defaults to 1. `min_nodes` must be at least 1 for use with a

1141

# Compute Engine machine type.

1142

#

1143

# Note that you cannot use AutoScaling if your version uses

1144

# [GPUs](#Version.FIELDS.accelerator_config). Instead, you must use

1145

# ManualScaling.

1146

#

1147

# You can set `min_nodes` when creating the model version, and you can also

1148

# update `min_nodes` for an existing version:

# <pre>

# update_body.json:

# {

# 'autoScaling': {

# 'minNodes': 5

# }

# }

# </pre>

# HTTP request:

# <pre style="max-width: 626px;">

1159

# PATCH

1160

# https://ml.googleapis.com/v1/{name=projects/*/models/*/versions/*}?update_mask=autoScaling.minNodes

1161

# -d @./update_body.json

1162

# </pre>

1163

},

1164

"explanationConfig": { # Message holding configuration options for explaining model predictions. # Optional. Configures explainability features on the model's version.

1165

# Some explanation features require additional metadata to be loaded

1166

# as part of the model payload.

1167

# There are two feature attribution methods supported for TensorFlow models:

1168

# integrated gradients and sampled Shapley.

1169

# [Learn more about feature

1170

# attributions.](/ai-platform/prediction/docs/ai-explanations/overview)

1171

"integratedGradientsAttribution": { # Attributes credit by computing the Aumann-Shapley value taking advantage # Attributes credit by computing the Aumann-Shapley value taking advantage

1172

# of the model's fully differentiable structure. Refer to this paper for

1173

# more details: http://proceedings.mlr.press/v70/sundararajan17a.html

1174

# of the model's fully differentiable structure. Refer to this paper for

1175

# more details: https://arxiv.org/abs/1703.01365

1176

"numIntegralSteps": 42, # Number of steps for approximating the path integral.

1177

# A good value to start is 50 and gradually increase until the

1178

# sum to diff property is met within the desired error range.

1179

},

1180

"xraiAttribution": { # Attributes credit by computing the XRAI taking advantage # Attributes credit by computing the XRAI taking advantage

1181

# of the model's fully differentiable structure. Refer to this paper for

1182

# more details: https://arxiv.org/abs/1906.02825

1183

# Currently only implemented for models with natural image inputs.

1184

# of the model's fully differentiable structure. Refer to this paper for

1185

# more details: https://arxiv.org/abs/1906.02825

1186

# Currently only implemented for models with natural image inputs.

1187

"numIntegralSteps": 42, # Number of steps for approximating the path integral.

1188

# A good value to start is 50 and gradually increase until the

1189

# sum to diff property is met within the desired error range.

1190

},

1191

"sampledShapleyAttribution": { # An attribution method that approximates Shapley values for features that # An attribution method that approximates Shapley values for features that

1192

# contribute to the label being predicted. A sampling strategy is used to

1193

# approximate the value rather than considering all subsets of features.

1194

# contribute to the label being predicted. A sampling strategy is used to

1195

# approximate the value rather than considering all subsets of features.

1196

"numPaths": 42, # The number of feature permutations to consider when approximating the

# Shapley values.

},

},

"pythonVersion": "A String", # Required. The version of Python used in prediction.

1201

#

1202

# The following Python versions are available:

1203

#

1204

# * Python '3.7' is available when `runtime_version` is set to '1.15' or

1205

# later.

1206

# * Python '3.5' is available when `runtime_version` is set to a version

1207

# from '1.4' to '1.14'.

1208

# * Python '2.7' is available when `runtime_version` is set to '1.15' or

1209

# earlier.

1210

#

1211

# Read more about the Python versions available for [each runtime

1212

# version](/ml-engine/docs/runtime-version-list).

1213

"requestLoggingConfig": { # Configuration for logging request-response pairs to a BigQuery table. # Optional. *Only* specify this field in a

1214

# projects.models.versions.patch

1215

# request. Specifying it in a

1216

# projects.models.versions.create

1217

# request has no effect.

1218

#

1219

# Configures the request-response pair logging on predictions from this

1220

# Version.

1221

# Online prediction requests to a model version and the responses to these

1222

# requests are converted to raw strings and saved to the specified BigQuery

1223

# table. Logging is constrained by [BigQuery quotas and

1224

# limits](/bigquery/quotas). If your project exceeds BigQuery quotas or limits,

1225

# AI Platform Prediction does not log request-response pairs, but it continues

1226

# to serve predictions.

1227

#

1228

# If you are using [continuous

1229

# evaluation](/ml-engine/docs/continuous-evaluation/), you do not need to

1230

# specify this configuration manually. Setting up continuous evaluation

1231

# automatically enables logging of request-response pairs.

1232

"samplingPercentage": 3.14, # Percentage of requests to be logged, expressed as a fraction from 0 to 1.

1233

# For example, if you want to log 10% of requests, enter `0.1`. The sampling

1234

# window is the lifetime of the model version. Defaults to 0.

1235

"bigqueryTableName": "A String", # Required. Fully qualified BigQuery table name in the following format:

1236

# "<var>project_id</var>.<var>dataset_name</var>.<var>table_name</var>"

1237

#

1238

# The specified table must already exist, and the "Cloud ML Service Agent"

1239

# for your project must have permission to write to it. The table must have

1240

# the following [schema](/bigquery/docs/schemas):

1241

#

1242

# <table>

1243

# <tr><th>Field name</th><th style="display: table-cell">Type</th>

1244

# <th style="display: table-cell">Mode</th></tr>

1245

# <tr><td>model</td><td>STRING</td><td>REQUIRED</td></tr>

1246

# <tr><td>model_version</td><td>STRING</td><td>REQUIRED</td></tr>

1247

# <tr><td>time</td><td>TIMESTAMP</td><td>REQUIRED</td></tr>

1248

# <tr><td>raw_data</td><td>STRING</td><td>REQUIRED</td></tr>

1249

# <tr><td>raw_prediction</td><td>STRING</td><td>NULLABLE</td></tr>

1250

# <tr><td>groundtruth</td><td>STRING</td><td>NULLABLE</td></tr>

1251

# </table>

1252

},

1253

"manualScaling": { # Options for manually scaling a model. # Manually select the number of nodes to use for serving the

1254

# model. You should generally use `auto_scaling` with an appropriate

1255

# `min_nodes` instead, but this option is available if you want more

1256

# predictable billing. Beware that latency and error rates will increase

1257

# if the traffic exceeds that capability of the system to serve it based

1258

# on the selected number of nodes.

1259

"nodes": 42, # The number of nodes to allocate for this model. These nodes are always up,

1260

# starting from the time the model is deployed, so the cost of operating

1261

# this model will be proportional to `nodes` * number of hours since

1262

# last billing cycle plus the cost for each prediction performed.

1263

},

1264

"createTime": "A String", # Output only. The time the version was created.

1265

"lastUseTime": "A String", # Output only. The time the version was last used for prediction.

1266

"framework": "A String", # Optional. The machine learning framework AI Platform uses to train

1267

# this version of the model. Valid values are `TENSORFLOW`, `SCIKIT_LEARN`,

1268

# `XGBOOST`. If you do not specify a framework, AI Platform

1269

# will analyze files in the deployment_uri to determine a framework. If you

1270

# choose `SCIKIT_LEARN` or `XGBOOST`, you must also set the runtime version

1271

# of the model to 1.4 or greater.

1272

#

1273

# Do **not** specify a framework if you're deploying a [custom

1274

# prediction routine](/ml-engine/docs/tensorflow/custom-prediction-routines).

1275

#

1276

# If you specify a [Compute Engine (N1) machine

1277

# type](/ml-engine/docs/machine-types-online-prediction) in the

1278

# `machineType` field, you must specify `TENSORFLOW`

1279

# for the framework.

1280

"predictionClass": "A String", # Optional. The fully qualified name

1281

# (<var>module_name</var>.<var>class_name</var>) of a class that implements

1282

# the Predictor interface described in this reference field. The module

1283

# containing this class should be included in a package provided to the

1284

# [`packageUris` field](#Version.FIELDS.package_uris).

1285

#

1286

# Specify this field if and only if you are deploying a [custom prediction

1287

# routine (beta)](/ml-engine/docs/tensorflow/custom-prediction-routines).

1288

# If you specify this field, you must set

1289

# [`runtimeVersion`](#Version.FIELDS.runtime_version) to 1.4 or greater and

1290

# you must set `machineType` to a [legacy (MLS1)

1291

# machine type](/ml-engine/docs/machine-types-online-prediction).

1292

#

1293

# The following code sample provides the Predictor interface:

1294

#

1295

# <pre style="max-width: 626px;">

1296

# class Predictor(object):

1297

# """Interface for constructing custom predictors."""

1298

#

1299

# def predict(self, instances, **kwargs):

1300

# """Performs custom prediction.

1301

#

1302

# Instances are the decoded values from the request. They have already

1303

# been deserialized from JSON.

1304

#

1305

# Args:

1306

# instances: A list of prediction input instances.

1307

# **kwargs: A dictionary of keyword args provided as additional

1308

# fields on the predict request body.

1309

#

1310

# Returns:

1311

# A list of outputs containing the prediction results. This list must

1312

# be JSON serializable.

1313

# """

1314

# raise NotImplementedError()

1315

#

1316

# @classmethod

1317

# def from_path(cls, model_dir):

1318

# """Creates an instance of Predictor using the given path.

1319

#

1320

# Loading of the predictor should be done in this method.

1321

#

1322

# Args:

1323

# model_dir: The local directory that contains the exported model

1324

# file along with any additional files uploaded when creating the

# version resource.

#

# Returns:

# An instance implementing this Predictor class.

1329

# """

1330

# raise NotImplementedError()

1331

# </pre>

1332

#

1333

# Learn more about [the Predictor interface and custom prediction

1334

# routines](/ml-engine/docs/tensorflow/custom-prediction-routines).

1335

"isDefault": True or False, # Output only. If true, this version will be used to handle prediction

1336

# requests that do not specify a version.

1337

#

1338

# You can change the default version by calling

1339

# projects.methods.versions.setDefault.

1340

"etag": "A String", # `etag` is used for optimistic concurrency control as a way to help

1341

# prevent simultaneous updates of a model from overwriting each other.

1342

# It is strongly suggested that systems make use of the `etag` in the

1343

# read-modify-write cycle to perform model updates in order to avoid race

1344

# conditions: An `etag` is returned in the response to `GetVersion`, and

1345

# systems are expected to put that etag in the request to `UpdateVersion` to

1346

# ensure that their change will be applied to the model as intended.

1347

"serviceAccount": "A String", # Optional. Specifies the service account for resource access control.

1348

"errorMessage": "A String", # Output only. The details of a failure or a cancellation.

1349

"deploymentUri": "A String", # Required. The Cloud Storage location of the trained model used to

1350

# create the version. See the

1351

# [guide to model

1352

# deployment](/ml-engine/docs/tensorflow/deploying-models) for more

1353

# information.

1354

#

1355

# When passing Version to

1356

# projects.models.versions.create

1357

# the model service uses the specified location as the source of the model.

1358

# Once deployed, the model version is hosted by the prediction service, so

1359

# this location is useful only as a historical record.

1360

# The total number of model files can't exceed 1000.

1361

"runtimeVersion": "A String", # Required. The AI Platform runtime version to use for this deployment.

1362

#

1363

# For more information, see the

1364

# [runtime version list](/ml-engine/docs/runtime-version-list) and

1365

# [how to manage runtime versions](/ml-engine/docs/versioning).

1366

"description": "A String", # Optional. The description specified for the version when it was created.

1367

},

1368

"onlinePredictionLogging": True or False, # Optional. If true, online prediction access logs are sent to StackDriver

1369

# Logging. These logs are like standard server access logs, containing

1370

# information like timestamp and latency for each request. Note that

1371

# [Stackdriver logs may incur a cost](/stackdriver/pricing), especially if

1372

# your project receives prediction requests at a high queries per second rate

1373

# (QPS). Estimate your costs before enabling this option.

1374

#

1375

# Default is false.

1376

}</pre>

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

</div>

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

1380

<code class="details" id="getIamPolicy">getIamPolicy(resource, options_requestedPolicyVersion=None, x__xgafv=None)</code>

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1381

<pre>Gets the access control policy for a resource.

1382

Returns an empty policy if the resource exists and does not have a policy

set.

Args:

resource: string, REQUIRED: The resource for which the policy is being requested.

1387

See the operation documentation for the appropriate value for this field. (required)

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

1388

options_requestedPolicyVersion: integer, Optional. The policy format version to be returned.

1389

1390

Valid values are 0, 1, and 3. Requests specifying an invalid value will be

1391

rejected.

1392

1393

Requests for policies with any conditional bindings must specify version 3.

1394

Policies without any conditional bindings may specify any valid value or

1395

leave the field unset.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1396

1397

To learn which resources support conditions in their IAM policies, see the

1398

[IAM

1399

documentation](https://cloud.google.com/iam/help/conditions/resource-policies).

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1400

x__xgafv: string, V1 error format.

Allowed values

1 - v1 error format

2 - v2 error format

Returns:

An object of the form:

1407

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

1408

{ # An Identity and Access Management (IAM) policy, which specifies access

1409

# controls for Google Cloud resources.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1410

#

1411

#

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

1412

# A `Policy` is a collection of `bindings`. A `binding` binds one or more

1413

# `members` to a single `role`. Members can be user accounts, service accounts,

1414

# Google groups, and domains (such as G Suite). A `role` is a named list of

1415

# permissions; each `role` can be an IAM predefined role or a user-created

1416

# custom role.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1417

#

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1418

# For some types of Google Cloud resources, a `binding` can also specify a

1419

# `condition`, which is a logical expression that allows access to a resource

1420

# only if the expression evaluates to `true`. A condition can add constraints

1421

# based on attributes of the request, the resource, or both. To learn which

1422

# resources support conditions in their IAM policies, see the

1423

# [IAM documentation](https://cloud.google.com/iam/help/conditions/resource-policies).

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

1424

#

1425

# **JSON example:**

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1426

#

1427

# {

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1428

# "bindings": [

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1429

# {

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1430

# "role": "roles/resourcemanager.organizationAdmin",

1431

# "members": [

1432

# "user:mike@example.com",

1433

# "group:admins@example.com",

1434

# "domain:google.com",

1435

# "serviceAccount:my-project-id@appspot.gserviceaccount.com"

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1436

# ]

1437

# },

1438

# {

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1439

# "role": "roles/resourcemanager.organizationViewer",

1440

# "members": [

1441

# "user:eve@example.com"

1442

# ],

1443

# "condition": {

1444

# "title": "expirable access",

1445

# "description": "Does not grant access after Sep 2020",

1446

# "expression": "request.time < timestamp('2020-10-01T00:00:00.000Z')",

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

1447

# }

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1448

# }

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

1449

# ],

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1450

# "etag": "BwWWja0YfJA=",

1451

# "version": 3

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1452

# }

1453

#

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

1454

# **YAML example:**

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

#

# bindings:

# - members:

# - user:mike@example.com

1459

# - group:admins@example.com

1460

# - domain:google.com

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

1461

# - serviceAccount:my-project-id@appspot.gserviceaccount.com

1462

# role: roles/resourcemanager.organizationAdmin

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1463

# - members:

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

1464

# - user:eve@example.com

1465

# role: roles/resourcemanager.organizationViewer

1466

# condition:

1467

# title: expirable access

1468

# description: Does not grant access after Sep 2020

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1469

# expression: request.time < timestamp('2020-10-01T00:00:00.000Z')

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

1470

# - etag: BwWWja0YfJA=

1471

# - version: 3

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1472

#

1473

# For a description of IAM and its features, see the

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

1474

# [IAM documentation](https://cloud.google.com/iam/docs/).

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1475

"etag": "A String", # `etag` is used for optimistic concurrency control as a way to help

1476

# prevent simultaneous updates of a policy from overwriting each other.

1477

# It is strongly suggested that systems make use of the `etag` in the

1478

# read-modify-write cycle to perform policy updates in order to avoid race

1479

# conditions: An `etag` is returned in the response to `getIamPolicy`, and

1480

# systems are expected to put that etag in the request to `setIamPolicy` to

1481

# ensure that their change will be applied to the same version of the policy.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1482

#

1483

# **Important:** If you use IAM Conditions, you must include the `etag` field

1484

# whenever you call `setIamPolicy`. If you omit this field, then IAM allows

1485

# you to overwrite a version `3` policy with a version `1` policy, and all of

1486

# the conditions in the version `3` policy are lost.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1487

"auditConfigs": [ # Specifies cloud audit logging configuration for this policy.

1488

{ # Specifies the audit configuration for a service.

1489

# The configuration determines which permission types are logged, and what

1490

# identities, if any, are exempted from logging.

1491

# An AuditConfig must have one or more AuditLogConfigs.

1492

#

1493

# If there are AuditConfigs for both `allServices` and a specific service,

1494

# the union of the two AuditConfigs is used for that service: the log_types

1495

# specified in each AuditConfig are enabled, and the exempted_members in each

1496

# AuditLogConfig are exempted.

1497

#

1498

# Example Policy with multiple AuditConfigs:

#

# {

# "audit_configs": [

# {

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1503

# "service": "allServices",

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1504

# "audit_log_configs": [

1505

# {

1506

# "log_type": "DATA_READ",

1507

# "exempted_members": [

1508

# "user:jose@example.com"

1509

# ]

1510

# },

1511

# {

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1512

# "log_type": "DATA_WRITE"

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1513

# },

1514

# {

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1515

# "log_type": "ADMIN_READ"

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

# }

# ]

# },

# {

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1520

# "service": "sampleservice.googleapis.com",

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1521

# "audit_log_configs": [

1522

# {

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1523

# "log_type": "DATA_READ"

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1524

# },

1525

# {

1526

# "log_type": "DATA_WRITE",

1527

# "exempted_members": [

1528

# "user:aliya@example.com"

# ]

# }

# ]

# }

# ]

# }

#

# For sampleservice, this policy enables DATA_READ, DATA_WRITE and ADMIN_READ

1537

# logging. It also exempts jose@example.com from DATA_READ logging, and

1538

# aliya@example.com from DATA_WRITE logging.

1539

"service": "A String", # Specifies a service that will be enabled for audit logging.

1540

# For example, `storage.googleapis.com`, `cloudsql.googleapis.com`.

1541

# `allServices` is a special value that covers all services.

1542

"auditLogConfigs": [ # The configuration for logging of each type of permission.

1543

{ # Provides the configuration for logging a type of permissions.

# Example:

#

# {

# "audit_log_configs": [

1548

# {

1549

# "log_type": "DATA_READ",

1550

# "exempted_members": [

1551

# "user:jose@example.com"

1552

# ]

1553

# },

1554

# {

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1555

# "log_type": "DATA_WRITE"

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

# }

# ]

# }

#

# This enables 'DATA_READ' and 'DATA_WRITE' logging, while exempting

1561

# jose@example.com from DATA_READ logging.

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1562

"logType": "A String", # The log type that this config enables.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1563

"exemptedMembers": [ # Specifies the identities that do not cause logging for this type of

1564

# permission.

1565

# Follows the same format of Binding.members.

1566

"A String",

1567

],

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

},

],

},

],

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1572

"version": 42, # Specifies the format of the policy.

1573

#

1574

# Valid values are `0`, `1`, and `3`. Requests that specify an invalid value

1575

# are rejected.

1576

#

1577

# Any operation that affects conditional role bindings must specify version

1578

# `3`. This requirement applies to the following operations:

1579

#

1580

# * Getting a policy that includes a conditional role binding

1581

# * Adding a conditional role binding to a policy

1582

# * Changing a conditional role binding in a policy

1583

# * Removing any role binding, with or without a condition, from a policy

1584

# that includes conditions

1585

#

1586

# **Important:** If you use IAM Conditions, you must include the `etag` field

1587

# whenever you call `setIamPolicy`. If you omit this field, then IAM allows

1588

# you to overwrite a version `3` policy with a version `1` policy, and all of

1589

# the conditions in the version `3` policy are lost.

1590

#

1591

# If a policy does not include any conditions, operations on that policy may

1592

# specify any valid version or leave the field unset.

1593

#

1594

# To learn which resources support conditions in their IAM policies, see the

1595

# [IAM documentation](https://cloud.google.com/iam/help/conditions/resource-policies).

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1596

"bindings": [ # Associates a list of `members` to a `role`. Optionally, may specify a

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

1597

# `condition` that determines how and when the `bindings` are applied. Each

1598

# of the `bindings` must contain at least one member.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1599

{ # Associates `members` with a `role`.

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1600

"role": "A String", # Role that is assigned to `members`.

1601

# For example, `roles/viewer`, `roles/editor`, or `roles/owner`.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1602

"condition": { # Represents a textual expression in the Common Expression Language (CEL) # The condition that is associated with this binding.

1603

#

1604

# If the condition evaluates to `true`, then this binding applies to the

1605

# current request.

1606

#

1607

# If the condition evaluates to `false`, then this binding does not apply to

1608

# the current request. However, a different role binding might grant the same

1609

# role to one or more of the members in this binding.

1610

#

1611

# To learn which resources support conditions in their IAM policies, see the

1612

# [IAM

1613

# documentation](https://cloud.google.com/iam/help/conditions/resource-policies).

1614

# syntax. CEL is a C-like expression language. The syntax and semantics of CEL

1615

# are documented at https://github.com/google/cel-spec.

1616

#

1617

# Example (Comparison):

1618

#

1619

# title: "Summary size limit"

1620

# description: "Determines if a summary is less than 100 chars"

1621

# expression: "document.summary.size() < 100"

1622

#

1623

# Example (Equality):

1624

#

1625

# title: "Requestor is owner"

1626

# description: "Determines if requestor is the document owner"

1627

# expression: "document.owner == request.auth.claims.email"

#

# Example (Logic):

#

# title: "Public documents"

1632

# description: "Determine whether the document should be publicly visible"

1633

# expression: "document.type != 'private' && document.type != 'internal'"

1634

#

1635

# Example (Data Manipulation):

1636

#

1637

# title: "Notification string"

1638

# description: "Create a notification string with a timestamp."

1639

# expression: "'New message received at ' + string(document.create_time)"

1640

#

1641

# The exact variables and functions that may be referenced within an expression

1642

# are determined by the service that evaluates it. See the service

1643

# documentation for additional information.

Bu Sun Kim

2020-05-27 12:20:54 -0700

[diff] [blame]

1644

"expression": "A String", # Textual representation of an expression in Common Expression Language

1645

# syntax.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1646

"title": "A String", # Optional. Title for the expression, i.e. a short string describing

1647

# its purpose. This can be used e.g. in UIs which allow to enter the

1648

# expression.

1649

"location": "A String", # Optional. String indicating the location of the expression for error

1650

# reporting, e.g. a file name and a position in the file.

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1651

"description": "A String", # Optional. Description of the expression. This is a longer text which

1652

# describes the expression, e.g. when hovered over it in a UI.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1653

},

1654

"members": [ # Specifies the identities requesting access for a Cloud Platform resource.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1655

# `members` can have the following values:

1656

#

1657

# * `allUsers`: A special identifier that represents anyone who is

1658

# on the internet; with or without a Google account.

1659

#

1660

# * `allAuthenticatedUsers`: A special identifier that represents anyone

1661

# who is authenticated with a Google account or a service account.

1662

#

1663

# * `user:{emailid}`: An email address that represents a specific Google

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

1664

# account. For example, `alice@example.com` .

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1665

#

1666

#

1667

# * `serviceAccount:{emailid}`: An email address that represents a service

1668

# account. For example, `my-other-app@appspot.gserviceaccount.com`.

1669

#

1670

# * `group:{emailid}`: An email address that represents a Google group.

1671

# For example, `admins@example.com`.

1672

#

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

1673

# * `deleted:user:{emailid}?uid={uniqueid}`: An email address (plus unique

1674

# identifier) representing a user that has been recently deleted. For

1675

# example, `alice@example.com?uid=123456789012345678901`. If the user is

1676

# recovered, this value reverts to `user:{emailid}` and the recovered user

1677

# retains the role in the binding.

1678

#

1679

# * `deleted:serviceAccount:{emailid}?uid={uniqueid}`: An email address (plus

1680

# unique identifier) representing a service account that has been recently

1681

# deleted. For example,

1682

# `my-other-app@appspot.gserviceaccount.com?uid=123456789012345678901`.

1683

# If the service account is undeleted, this value reverts to

1684

# `serviceAccount:{emailid}` and the undeleted service account retains the

1685

# role in the binding.

1686

#

1687

# * `deleted:group:{emailid}?uid={uniqueid}`: An email address (plus unique

1688

# identifier) representing a Google group that has been recently

1689

# deleted. For example, `admins@example.com?uid=123456789012345678901`. If

1690

# the group is recovered, this value reverts to `group:{emailid}` and the

1691

# recovered group retains the role in the binding.

1692

#

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1693

#

1694

# * `domain:{domain}`: The G Suite domain (primary) that represents all the

1695

# users of that domain. For example, `google.com` or `example.com`.

1696

#

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1697

"A String",

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1698

],

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1699

},

1700

],

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

}</pre>

</div>

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1705

<code class="details" id="list">list(parent, pageSize=None, pageToken=None, filter=None, x__xgafv=None)</code>

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

1706

<pre>Lists the models in a project.

1707

1708

Each project can contain multiple models, and each model can have multiple

1709

versions.

1710

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1711

If there are no models that match the request parameters, the list request

1712

returns an empty response body: {}.

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

1713

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1714

Args:

1715

parent: string, Required. The name of the project whose models are to be listed. (required)

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1716

pageSize: integer, Optional. The number of models to retrieve per "page" of results. If there

1717

are more remaining results than this number, the response message will

1718

contain a valid value in the `next_page_token` field.

1719

1720

The default value is 20, and the maximum page size is 100.

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1721

pageToken: string, Optional. A page token to request the next page of results.

1722

1723

You get the token from the `next_page_token` field of the response from

1724

the previous call.

1725

filter: string, Optional. Specifies the subset of models to retrieve.

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

1726

x__xgafv: string, V1 error format.

1727

Allowed values

1728

1 - v1 error format

1729

2 - v2 error format

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

1730

1731

Returns:

1732

An object of the form:

1733

1734

{ # Response message for the ListModels method.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1735

"models": [ # The list of models.

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

1736

{ # Represents a machine learning solution.

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

1737

#

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1738

# A model can have multiple versions, each of which is a deployed, trained

1739

# model ready to receive prediction requests. The model itself is just a

1740

# container.

1741

"description": "A String", # Optional. The description specified for the model when it was created.

1742

"regions": [ # Optional. The list of regions where the model is going to be deployed.

1743

# Only one region per model is supported.

1744

# Defaults to 'us-central1' if nothing is set.

1745

# See the <a href="/ml-engine/docs/tensorflow/regions">available regions</a>

1746

# for AI Platform services.

1747

# Note:

1748

# * No matter where a model is deployed, it can always be accessed by

1749

# users from anywhere, both for online and batch prediction.

1750

# * The region for a batch prediction job is set by the region field when

1751

# submitting the batch prediction job and does not take its value from

1752

# this field.

Bu Sun Kim

2020-05-27 12:20:54 -0700

[diff] [blame]

1753

"A String",

1754

],

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1755

"name": "A String", # Required. The name specified for the model when it was created.

Bu Sun Kim

2020-05-27 12:20:54 -0700

[diff] [blame]

1756

#

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1757

# The model name must be unique within the project it is created in.

1758

"onlinePredictionConsoleLogging": True or False, # Optional. If true, online prediction nodes send `stderr` and `stdout`

1759

# streams to Stackdriver Logging. These can be more verbose than the standard

1760

# access logs (see `onlinePredictionLogging`) and can incur higher cost.

1761

# However, they are helpful for debugging. Note that

1762

# [Stackdriver logs may incur a cost](/stackdriver/pricing), especially if

1763

# your project receives prediction requests at a high QPS. Estimate your

1764

# costs before enabling this option.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1765

#

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1766

# Default is false.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

1767

"etag": "A String", # `etag` is used for optimistic concurrency control as a way to help

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1768

# prevent simultaneous updates of a model from overwriting each other.

1769

# It is strongly suggested that systems make use of the `etag` in the

1770

# read-modify-write cycle to perform model updates in order to avoid race

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1771

# conditions: An `etag` is returned in the response to `GetModel`, and

1772

# systems are expected to put that etag in the request to `UpdateModel` to

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

1773

# ensure that their change will be applied to the model as intended.

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

1774

"labels": { # Optional. One or more labels that you can add, to organize your models.

1775

# Each label is a key-value pair, where both the key and the value are

1776

# arbitrary strings that you supply.

1777

# For more information, see the documentation on

1778

# <a href="/ml-engine/docs/tensorflow/resource-labels">using labels</a>.

1779

"a_key": "A String",

1780

},

1781

"defaultVersion": { # Represents a version of the model. # Output only. The default version of the model. This version will be used to

1782

# handle prediction requests that do not specify a version.

1783

#

1784

# You can change the default version by calling

1785

# projects.models.versions.setDefault.

1786

#

1787

# Each version is a trained model deployed in the cloud, ready to handle

1788

# prediction requests. A model can have multiple versions. You can get

1789

# information about all of the versions of a given model by calling

1790

# projects.models.versions.list.

1791

"labels": { # Optional. One or more labels that you can add, to organize your model

1792

# versions. Each label is a key-value pair, where both the key and the value

1793

# are arbitrary strings that you supply.

1794

# For more information, see the documentation on

1795

# <a href="/ml-engine/docs/tensorflow/resource-labels">using labels</a>.

1796

"a_key": "A String",

1797

},

1798

"machineType": "A String", # Optional. The type of machine on which to serve the model. Currently only

1799

# applies to online prediction service. If this field is not specified, it

1800

# defaults to `mls1-c1-m2`.

1801

#

1802

# Online prediction supports the following machine types:

#

# * `mls1-c1-m2`

# * `mls1-c4-m2`

# * `n1-standard-2`

# * `n1-standard-4`

# * `n1-standard-8`

# * `n1-standard-16`

# * `n1-standard-32`

# * `n1-highmem-2`

# * `n1-highmem-4`

# * `n1-highmem-8`

# * `n1-highmem-16`

# * `n1-highmem-32`

# * `n1-highcpu-2`

# * `n1-highcpu-4`

# * `n1-highcpu-8`

# * `n1-highcpu-16`

# * `n1-highcpu-32`

#

# `mls1-c1-m2` is generally available. All other machine types are available

1823

# in beta. Learn more about the [differences between machine

1824

# types](/ml-engine/docs/machine-types-online-prediction).

1825

"packageUris": [ # Optional. Cloud Storage paths (`gs://…`) of packages for [custom

1826

# prediction routines](/ml-engine/docs/tensorflow/custom-prediction-routines)

1827

# or [scikit-learn pipelines with custom

1828

# code](/ml-engine/docs/scikit/exporting-for-prediction#custom-pipeline-code).

1829

#

1830

# For a custom prediction routine, one of these packages must contain your

1831

# Predictor class (see

1832

# [`predictionClass`](#Version.FIELDS.prediction_class)). Additionally,

1833

# include any dependencies used by your Predictor or scikit-learn pipeline

1834

# uses that are not already included in your selected [runtime

1835

# version](/ml-engine/docs/tensorflow/runtime-version-list).

1836

#

1837

# If you specify this field, you must also set

1838

# [`runtimeVersion`](#Version.FIELDS.runtime_version) to 1.4 or greater.

1839

"A String",

1840

],

1841

"acceleratorConfig": { # Represents a hardware accelerator request config. # Optional. Accelerator config for using GPUs for online prediction (beta).

1842

# Only specify this field if you have specified a Compute Engine (N1) machine

1843

# type in the `machineType` field. Learn more about [using GPUs for online

1844

# prediction](/ml-engine/docs/machine-types-online-prediction#gpus).

1845

# Note that the AcceleratorConfig can be used in both Jobs and Versions.

1846

# Learn more about [accelerators for training](/ml-engine/docs/using-gpus) and

1847

# [accelerators for online

1848

# prediction](/ml-engine/docs/machine-types-online-prediction#gpus).

1849

"type": "A String", # The type of accelerator to use.

1850

"count": "A String", # The number of accelerators to attach to each machine running the job.

1851

},

1852

"state": "A String", # Output only. The state of a version.

1853

"name": "A String", # Required. The name specified for the version when it was created.

1854

#

1855

# The version name must be unique within the model it is created in.

1856

"autoScaling": { # Options for automatically scaling a model. # Automatically scale the number of nodes used to serve the model in

1857

# response to increases and decreases in traffic. Care should be

1858

# taken to ramp up traffic according to the model's ability to scale

1859

# or you will start seeing increases in latency and 429 response codes.

1860

#

1861

# Note that you cannot use AutoScaling if your version uses

1862

# [GPUs](#Version.FIELDS.accelerator_config). Instead, you must use specify

1863

# `manual_scaling`.

1864

"minNodes": 42, # Optional. The minimum number of nodes to allocate for this model. These

1865

# nodes are always up, starting from the time the model is deployed.

1866

# Therefore, the cost of operating this model will be at least

1867

# `rate` * `min_nodes` * number of hours since last billing cycle,

1868

# where `rate` is the cost per node-hour as documented in the

1869

# [pricing guide](/ml-engine/docs/pricing),

1870

# even if no predictions are performed. There is additional cost for each

1871

# prediction performed.

1872

#

1873

# Unlike manual scaling, if the load gets too heavy for the nodes

1874

# that are up, the service will automatically add nodes to handle the

1875

# increased load as well as scale back as traffic drops, always maintaining

1876

# at least `min_nodes`. You will be charged for the time in which additional

1877

# nodes are used.

1878

#

1879

# If `min_nodes` is not specified and AutoScaling is used with a [legacy

1880

# (MLS1) machine type](/ml-engine/docs/machine-types-online-prediction),

1881

# `min_nodes` defaults to 0, in which case, when traffic to a model stops

1882

# (and after a cool-down period), nodes will be shut down and no charges will

1883

# be incurred until traffic to the model resumes.

1884

#

1885

# If `min_nodes` is not specified and AutoScaling is used with a [Compute

1886

# Engine (N1) machine type](/ml-engine/docs/machine-types-online-prediction),

1887

# `min_nodes` defaults to 1. `min_nodes` must be at least 1 for use with a

1888

# Compute Engine machine type.

1889

#

1890

# Note that you cannot use AutoScaling if your version uses

1891

# [GPUs](#Version.FIELDS.accelerator_config). Instead, you must use

1892

# ManualScaling.

1893

#

1894

# You can set `min_nodes` when creating the model version, and you can also

1895

# update `min_nodes` for an existing version:

# <pre>

# update_body.json:

# {

# 'autoScaling': {

# 'minNodes': 5

# }

# }

# </pre>

# HTTP request:

# <pre style="max-width: 626px;">

1906

# PATCH

1907

# https://ml.googleapis.com/v1/{name=projects/*/models/*/versions/*}?update_mask=autoScaling.minNodes

1908

# -d @./update_body.json

1909

# </pre>

1910

},

1911

"explanationConfig": { # Message holding configuration options for explaining model predictions. # Optional. Configures explainability features on the model's version.

1912

# Some explanation features require additional metadata to be loaded

1913

# as part of the model payload.

1914

# There are two feature attribution methods supported for TensorFlow models:

1915

# integrated gradients and sampled Shapley.

1916

# [Learn more about feature

1917

# attributions.](/ai-platform/prediction/docs/ai-explanations/overview)

1918

"integratedGradientsAttribution": { # Attributes credit by computing the Aumann-Shapley value taking advantage # Attributes credit by computing the Aumann-Shapley value taking advantage

1919

# of the model's fully differentiable structure. Refer to this paper for

1920

# more details: http://proceedings.mlr.press/v70/sundararajan17a.html

1921

# of the model's fully differentiable structure. Refer to this paper for

1922

# more details: https://arxiv.org/abs/1703.01365

1923

"numIntegralSteps": 42, # Number of steps for approximating the path integral.

1924

# A good value to start is 50 and gradually increase until the

1925

# sum to diff property is met within the desired error range.

1926

},

1927

"xraiAttribution": { # Attributes credit by computing the XRAI taking advantage # Attributes credit by computing the XRAI taking advantage

1928

# of the model's fully differentiable structure. Refer to this paper for

1929

# more details: https://arxiv.org/abs/1906.02825

1930

# Currently only implemented for models with natural image inputs.

1931

# of the model's fully differentiable structure. Refer to this paper for

1932

# more details: https://arxiv.org/abs/1906.02825

1933

# Currently only implemented for models with natural image inputs.

1934

"numIntegralSteps": 42, # Number of steps for approximating the path integral.

1935

# A good value to start is 50 and gradually increase until the

1936

# sum to diff property is met within the desired error range.

1937

},

1938

"sampledShapleyAttribution": { # An attribution method that approximates Shapley values for features that # An attribution method that approximates Shapley values for features that

1939

# contribute to the label being predicted. A sampling strategy is used to

1940

# approximate the value rather than considering all subsets of features.

1941

# contribute to the label being predicted. A sampling strategy is used to

1942

# approximate the value rather than considering all subsets of features.

1943

"numPaths": 42, # The number of feature permutations to consider when approximating the

# Shapley values.

},

},

"pythonVersion": "A String", # Required. The version of Python used in prediction.

1948

#

1949

# The following Python versions are available:

1950

#

1951

# * Python '3.7' is available when `runtime_version` is set to '1.15' or

1952

# later.

1953

# * Python '3.5' is available when `runtime_version` is set to a version

1954

# from '1.4' to '1.14'.

1955

# * Python '2.7' is available when `runtime_version` is set to '1.15' or

1956

# earlier.

1957

#

1958

# Read more about the Python versions available for [each runtime

1959

# version](/ml-engine/docs/runtime-version-list).

1960

"requestLoggingConfig": { # Configuration for logging request-response pairs to a BigQuery table. # Optional. *Only* specify this field in a

1961

# projects.models.versions.patch

1962

# request. Specifying it in a

1963

# projects.models.versions.create

1964

# request has no effect.

1965

#

1966

# Configures the request-response pair logging on predictions from this

1967

# Version.

1968

# Online prediction requests to a model version and the responses to these

1969

# requests are converted to raw strings and saved to the specified BigQuery

1970

# table. Logging is constrained by [BigQuery quotas and

1971

# limits](/bigquery/quotas). If your project exceeds BigQuery quotas or limits,

1972

# AI Platform Prediction does not log request-response pairs, but it continues

1973

# to serve predictions.

1974

#

1975

# If you are using [continuous

1976

# evaluation](/ml-engine/docs/continuous-evaluation/), you do not need to

1977

# specify this configuration manually. Setting up continuous evaluation

1978

# automatically enables logging of request-response pairs.

1979

"samplingPercentage": 3.14, # Percentage of requests to be logged, expressed as a fraction from 0 to 1.

1980

# For example, if you want to log 10% of requests, enter `0.1`. The sampling

1981

# window is the lifetime of the model version. Defaults to 0.

1982

"bigqueryTableName": "A String", # Required. Fully qualified BigQuery table name in the following format:

1983

# "<var>project_id</var>.<var>dataset_name</var>.<var>table_name</var>"

1984

#

1985

# The specified table must already exist, and the "Cloud ML Service Agent"

1986

# for your project must have permission to write to it. The table must have

1987

# the following [schema](/bigquery/docs/schemas):

1988

#

1989

# <table>

1990

# <tr><th>Field name</th><th style="display: table-cell">Type</th>

1991

# <th style="display: table-cell">Mode</th></tr>

1992

# <tr><td>model</td><td>STRING</td><td>REQUIRED</td></tr>

1993

# <tr><td>model_version</td><td>STRING</td><td>REQUIRED</td></tr>

1994

# <tr><td>time</td><td>TIMESTAMP</td><td>REQUIRED</td></tr>

1995

# <tr><td>raw_data</td><td>STRING</td><td>REQUIRED</td></tr>

1996

# <tr><td>raw_prediction</td><td>STRING</td><td>NULLABLE</td></tr>

1997

# <tr><td>groundtruth</td><td>STRING</td><td>NULLABLE</td></tr>

1998

# </table>

1999

},

2000

"manualScaling": { # Options for manually scaling a model. # Manually select the number of nodes to use for serving the

2001

# model. You should generally use `auto_scaling` with an appropriate

2002

# `min_nodes` instead, but this option is available if you want more

2003

# predictable billing. Beware that latency and error rates will increase

2004

# if the traffic exceeds that capability of the system to serve it based

2005

# on the selected number of nodes.

2006

"nodes": 42, # The number of nodes to allocate for this model. These nodes are always up,

2007

# starting from the time the model is deployed, so the cost of operating

2008

# this model will be proportional to `nodes` * number of hours since

2009

# last billing cycle plus the cost for each prediction performed.

2010

},

2011

"createTime": "A String", # Output only. The time the version was created.

2012

"lastUseTime": "A String", # Output only. The time the version was last used for prediction.

2013

"framework": "A String", # Optional. The machine learning framework AI Platform uses to train

2014

# this version of the model. Valid values are `TENSORFLOW`, `SCIKIT_LEARN`,

2015

# `XGBOOST`. If you do not specify a framework, AI Platform

2016

# will analyze files in the deployment_uri to determine a framework. If you

2017

# choose `SCIKIT_LEARN` or `XGBOOST`, you must also set the runtime version

2018

# of the model to 1.4 or greater.

2019

#

2020

# Do **not** specify a framework if you're deploying a [custom

2021

# prediction routine](/ml-engine/docs/tensorflow/custom-prediction-routines).

2022

#

2023

# If you specify a [Compute Engine (N1) machine

2024

# type](/ml-engine/docs/machine-types-online-prediction) in the

2025

# `machineType` field, you must specify `TENSORFLOW`

2026

# for the framework.

2027

"predictionClass": "A String", # Optional. The fully qualified name

2028

# (<var>module_name</var>.<var>class_name</var>) of a class that implements

2029

# the Predictor interface described in this reference field. The module

2030

# containing this class should be included in a package provided to the

2031

# [`packageUris` field](#Version.FIELDS.package_uris).

2032

#

2033

# Specify this field if and only if you are deploying a [custom prediction

2034

# routine (beta)](/ml-engine/docs/tensorflow/custom-prediction-routines).

2035

# If you specify this field, you must set

2036

# [`runtimeVersion`](#Version.FIELDS.runtime_version) to 1.4 or greater and

2037

# you must set `machineType` to a [legacy (MLS1)

2038

# machine type](/ml-engine/docs/machine-types-online-prediction).

2039

#

2040

# The following code sample provides the Predictor interface:

2041

#

2042

# <pre style="max-width: 626px;">

2043

# class Predictor(object):

2044

# """Interface for constructing custom predictors."""

2045

#

2046

# def predict(self, instances, **kwargs):

2047

# """Performs custom prediction.

2048

#

2049

# Instances are the decoded values from the request. They have already

2050

# been deserialized from JSON.

2051

#

2052

# Args:

2053

# instances: A list of prediction input instances.

2054

# **kwargs: A dictionary of keyword args provided as additional

2055

# fields on the predict request body.

2056

#

2057

# Returns:

2058

# A list of outputs containing the prediction results. This list must

2059

# be JSON serializable.

2060

# """

2061

# raise NotImplementedError()

2062

#

2063

# @classmethod

2064

# def from_path(cls, model_dir):

2065

# """Creates an instance of Predictor using the given path.

2066

#

2067

# Loading of the predictor should be done in this method.

2068

#

2069

# Args:

2070

# model_dir: The local directory that contains the exported model

2071

# file along with any additional files uploaded when creating the

# version resource.

#

# Returns:

# An instance implementing this Predictor class.

2076

# """

2077

# raise NotImplementedError()

2078

# </pre>

2079

#

2080

# Learn more about [the Predictor interface and custom prediction

2081

# routines](/ml-engine/docs/tensorflow/custom-prediction-routines).

2082

"isDefault": True or False, # Output only. If true, this version will be used to handle prediction

2083

# requests that do not specify a version.

2084

#

2085

# You can change the default version by calling

2086

# projects.methods.versions.setDefault.

2087

"etag": "A String", # `etag` is used for optimistic concurrency control as a way to help

2088

# prevent simultaneous updates of a model from overwriting each other.

2089

# It is strongly suggested that systems make use of the `etag` in the

2090

# read-modify-write cycle to perform model updates in order to avoid race

2091

# conditions: An `etag` is returned in the response to `GetVersion`, and

2092

# systems are expected to put that etag in the request to `UpdateVersion` to

2093

# ensure that their change will be applied to the model as intended.

2094

"serviceAccount": "A String", # Optional. Specifies the service account for resource access control.

2095

"errorMessage": "A String", # Output only. The details of a failure or a cancellation.

2096

"deploymentUri": "A String", # Required. The Cloud Storage location of the trained model used to

2097

# create the version. See the

2098

# [guide to model

2099

# deployment](/ml-engine/docs/tensorflow/deploying-models) for more

2100

# information.

2101

#

2102

# When passing Version to

2103

# projects.models.versions.create

2104

# the model service uses the specified location as the source of the model.

2105

# Once deployed, the model version is hosted by the prediction service, so

2106

# this location is useful only as a historical record.

2107

# The total number of model files can't exceed 1000.

2108

"runtimeVersion": "A String", # Required. The AI Platform runtime version to use for this deployment.

2109

#

2110

# For more information, see the

2111

# [runtime version list](/ml-engine/docs/runtime-version-list) and

2112

# [how to manage runtime versions](/ml-engine/docs/versioning).

2113

"description": "A String", # Optional. The description specified for the version when it was created.

2114

},

2115

"onlinePredictionLogging": True or False, # Optional. If true, online prediction access logs are sent to StackDriver

2116

# Logging. These logs are like standard server access logs, containing

2117

# information like timestamp and latency for each request. Note that

2118

# [Stackdriver logs may incur a cost](/stackdriver/pricing), especially if

2119

# your project receives prediction requests at a high queries per second rate

2120

# (QPS). Estimate your costs before enabling this option.

2121

#

2122

# Default is false.

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

2123

},

2124

],

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

2125

"nextPageToken": "A String", # Optional. Pass this token as the `page_token` field of the request for a

2126

# subsequent call.

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

}</pre>

</div>

<code class="details" id="list_next">list_next(previous_request, previous_response)</code>

2132

<pre>Retrieves the next page of results.

2133

2134

Args:

2135

previous_request: The request for the previous page. (required)

2136

previous_response: The response from the request for the previous page. (required)

2137

2138

Returns:

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2139

A request object that you can call 'execute()' on to request the next

Sai Cheemalapati

2017-03-13 12:12:03 -0400

[diff] [blame]

2140

page. Returns None if there are no more items in the collection.

</pre>

</div>

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2144

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2145

<code class="details" id="patch">patch(name, body=None, updateMask=None, x__xgafv=None)</code>

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2146

<pre>Updates a specific model resource.

2147

2148

Currently the only supported fields to update are `description` and

2149

`default_version.name`.

Args:

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2153

body: object, The request body.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2154

The object takes the form of:

2155

2156

{ # Represents a machine learning solution.

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2157

#

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

2158

# A model can have multiple versions, each of which is a deployed, trained

2159

# model ready to receive prediction requests. The model itself is just a

2160

# container.

2161

"description": "A String", # Optional. The description specified for the model when it was created.

2162

"regions": [ # Optional. The list of regions where the model is going to be deployed.

2163

# Only one region per model is supported.

2164

# Defaults to 'us-central1' if nothing is set.

2165

# See the <a href="/ml-engine/docs/tensorflow/regions">available regions</a>

2166

# for AI Platform services.

2167

# Note:

2168

# * No matter where a model is deployed, it can always be accessed by

2169

# users from anywhere, both for online and batch prediction.

2170

# * The region for a batch prediction job is set by the region field when

2171

# submitting the batch prediction job and does not take its value from

2172

# this field.

Bu Sun Kim

2020-05-27 12:20:54 -0700

[diff] [blame]

2173

"A String",

2174

],

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

2175

"name": "A String", # Required. The name specified for the model when it was created.

2176

#

2177

# The model name must be unique within the project it is created in.

2178

"onlinePredictionConsoleLogging": True or False, # Optional. If true, online prediction nodes send `stderr` and `stdout`

2179

# streams to Stackdriver Logging. These can be more verbose than the standard

2180

# access logs (see `onlinePredictionLogging`) and can incur higher cost.

2181

# However, they are helpful for debugging. Note that

2182

# [Stackdriver logs may incur a cost](/stackdriver/pricing), especially if

2183

# your project receives prediction requests at a high QPS. Estimate your

2184

# costs before enabling this option.

2185

#

2186

# Default is false.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2187

"etag": "A String", # `etag` is used for optimistic concurrency control as a way to help

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2188

# prevent simultaneous updates of a model from overwriting each other.

2189

# It is strongly suggested that systems make use of the `etag` in the

2190

# read-modify-write cycle to perform model updates in order to avoid race

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

2191

# conditions: An `etag` is returned in the response to `GetModel`, and

2192

# systems are expected to put that etag in the request to `UpdateModel` to

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2193

# ensure that their change will be applied to the model as intended.

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

2194

"labels": { # Optional. One or more labels that you can add, to organize your models.

2195

# Each label is a key-value pair, where both the key and the value are

2196

# arbitrary strings that you supply.

2197

# For more information, see the documentation on

2198

# <a href="/ml-engine/docs/tensorflow/resource-labels">using labels</a>.

2199

"a_key": "A String",

2200

},

2201

"defaultVersion": { # Represents a version of the model. # Output only. The default version of the model. This version will be used to

2202

# handle prediction requests that do not specify a version.

2203

#

2204

# You can change the default version by calling

2205

# projects.models.versions.setDefault.

2206

#

2207

# Each version is a trained model deployed in the cloud, ready to handle

2208

# prediction requests. A model can have multiple versions. You can get

2209

# information about all of the versions of a given model by calling

2210

# projects.models.versions.list.

2211

"labels": { # Optional. One or more labels that you can add, to organize your model

2212

# versions. Each label is a key-value pair, where both the key and the value

2213

# are arbitrary strings that you supply.

2214

# For more information, see the documentation on

2215

# <a href="/ml-engine/docs/tensorflow/resource-labels">using labels</a>.

2216

"a_key": "A String",

2217

},

2218

"machineType": "A String", # Optional. The type of machine on which to serve the model. Currently only

2219

# applies to online prediction service. If this field is not specified, it

2220

# defaults to `mls1-c1-m2`.

2221

#

2222

# Online prediction supports the following machine types:

#

# * `mls1-c1-m2`

# * `mls1-c4-m2`

# * `n1-standard-2`

# * `n1-standard-4`

# * `n1-standard-8`

# * `n1-standard-16`

# * `n1-standard-32`

# * `n1-highmem-2`

# * `n1-highmem-4`

# * `n1-highmem-8`

# * `n1-highmem-16`

# * `n1-highmem-32`

# * `n1-highcpu-2`

# * `n1-highcpu-4`

# * `n1-highcpu-8`

# * `n1-highcpu-16`

# * `n1-highcpu-32`

#

# `mls1-c1-m2` is generally available. All other machine types are available

2243

# in beta. Learn more about the [differences between machine

2244

# types](/ml-engine/docs/machine-types-online-prediction).

2245

"packageUris": [ # Optional. Cloud Storage paths (`gs://…`) of packages for [custom

2246

# prediction routines](/ml-engine/docs/tensorflow/custom-prediction-routines)

2247

# or [scikit-learn pipelines with custom

2248

# code](/ml-engine/docs/scikit/exporting-for-prediction#custom-pipeline-code).

2249

#

2250

# For a custom prediction routine, one of these packages must contain your

2251

# Predictor class (see

2252

# [`predictionClass`](#Version.FIELDS.prediction_class)). Additionally,

2253

# include any dependencies used by your Predictor or scikit-learn pipeline

2254

# uses that are not already included in your selected [runtime

2255

# version](/ml-engine/docs/tensorflow/runtime-version-list).

2256

#

2257

# If you specify this field, you must also set

2258

# [`runtimeVersion`](#Version.FIELDS.runtime_version) to 1.4 or greater.

2259

"A String",

2260

],

2261

"acceleratorConfig": { # Represents a hardware accelerator request config. # Optional. Accelerator config for using GPUs for online prediction (beta).

2262

# Only specify this field if you have specified a Compute Engine (N1) machine

2263

# type in the `machineType` field. Learn more about [using GPUs for online

2264

# prediction](/ml-engine/docs/machine-types-online-prediction#gpus).

2265

# Note that the AcceleratorConfig can be used in both Jobs and Versions.

2266

# Learn more about [accelerators for training](/ml-engine/docs/using-gpus) and

2267

# [accelerators for online

2268

# prediction](/ml-engine/docs/machine-types-online-prediction#gpus).

2269

"type": "A String", # The type of accelerator to use.

2270

"count": "A String", # The number of accelerators to attach to each machine running the job.

2271

},

2272

"state": "A String", # Output only. The state of a version.

2273

"name": "A String", # Required. The name specified for the version when it was created.

2274

#

2275

# The version name must be unique within the model it is created in.

2276

"autoScaling": { # Options for automatically scaling a model. # Automatically scale the number of nodes used to serve the model in

2277

# response to increases and decreases in traffic. Care should be

2278

# taken to ramp up traffic according to the model's ability to scale

2279

# or you will start seeing increases in latency and 429 response codes.

2280

#

2281

# Note that you cannot use AutoScaling if your version uses

2282

# [GPUs](#Version.FIELDS.accelerator_config). Instead, you must use specify

2283

# `manual_scaling`.

2284

"minNodes": 42, # Optional. The minimum number of nodes to allocate for this model. These

2285

# nodes are always up, starting from the time the model is deployed.

2286

# Therefore, the cost of operating this model will be at least

2287

# `rate` * `min_nodes` * number of hours since last billing cycle,

2288

# where `rate` is the cost per node-hour as documented in the

2289

# [pricing guide](/ml-engine/docs/pricing),

2290

# even if no predictions are performed. There is additional cost for each

2291

# prediction performed.

2292

#

2293

# Unlike manual scaling, if the load gets too heavy for the nodes

2294

# that are up, the service will automatically add nodes to handle the

2295

# increased load as well as scale back as traffic drops, always maintaining

2296

# at least `min_nodes`. You will be charged for the time in which additional

2297

# nodes are used.

2298

#

2299

# If `min_nodes` is not specified and AutoScaling is used with a [legacy

2300

# (MLS1) machine type](/ml-engine/docs/machine-types-online-prediction),

2301

# `min_nodes` defaults to 0, in which case, when traffic to a model stops

2302

# (and after a cool-down period), nodes will be shut down and no charges will

2303

# be incurred until traffic to the model resumes.

2304

#

2305

# If `min_nodes` is not specified and AutoScaling is used with a [Compute

2306

# Engine (N1) machine type](/ml-engine/docs/machine-types-online-prediction),

2307

# `min_nodes` defaults to 1. `min_nodes` must be at least 1 for use with a

2308

# Compute Engine machine type.

2309

#

2310

# Note that you cannot use AutoScaling if your version uses

2311

# [GPUs](#Version.FIELDS.accelerator_config). Instead, you must use

2312

# ManualScaling.

2313

#

2314

# You can set `min_nodes` when creating the model version, and you can also

2315

# update `min_nodes` for an existing version:

# <pre>

# update_body.json:

# {

# 'autoScaling': {

# 'minNodes': 5

# }

# }

# </pre>

# HTTP request:

# <pre style="max-width: 626px;">

2326

# PATCH

2327

# https://ml.googleapis.com/v1/{name=projects/*/models/*/versions/*}?update_mask=autoScaling.minNodes

2328

# -d @./update_body.json

2329

# </pre>

2330

},

2331

"explanationConfig": { # Message holding configuration options for explaining model predictions. # Optional. Configures explainability features on the model's version.

2332

# Some explanation features require additional metadata to be loaded

2333

# as part of the model payload.

2334

# There are two feature attribution methods supported for TensorFlow models:

2335

# integrated gradients and sampled Shapley.

2336

# [Learn more about feature

2337

# attributions.](/ai-platform/prediction/docs/ai-explanations/overview)

2338

"integratedGradientsAttribution": { # Attributes credit by computing the Aumann-Shapley value taking advantage # Attributes credit by computing the Aumann-Shapley value taking advantage

2339

# of the model's fully differentiable structure. Refer to this paper for

2340

# more details: http://proceedings.mlr.press/v70/sundararajan17a.html

2341

# of the model's fully differentiable structure. Refer to this paper for

2342

# more details: https://arxiv.org/abs/1703.01365

2343

"numIntegralSteps": 42, # Number of steps for approximating the path integral.

2344

# A good value to start is 50 and gradually increase until the

2345

# sum to diff property is met within the desired error range.

2346

},

2347

"xraiAttribution": { # Attributes credit by computing the XRAI taking advantage # Attributes credit by computing the XRAI taking advantage

2348

# of the model's fully differentiable structure. Refer to this paper for

2349

# more details: https://arxiv.org/abs/1906.02825

2350

# Currently only implemented for models with natural image inputs.

2351

# of the model's fully differentiable structure. Refer to this paper for

2352

# more details: https://arxiv.org/abs/1906.02825

2353

# Currently only implemented for models with natural image inputs.

2354

"numIntegralSteps": 42, # Number of steps for approximating the path integral.

2355

# A good value to start is 50 and gradually increase until the

2356

# sum to diff property is met within the desired error range.

2357

},

2358

"sampledShapleyAttribution": { # An attribution method that approximates Shapley values for features that # An attribution method that approximates Shapley values for features that

2359

# contribute to the label being predicted. A sampling strategy is used to

2360

# approximate the value rather than considering all subsets of features.

2361

# contribute to the label being predicted. A sampling strategy is used to

2362

# approximate the value rather than considering all subsets of features.

2363

"numPaths": 42, # The number of feature permutations to consider when approximating the

# Shapley values.

},

},

"pythonVersion": "A String", # Required. The version of Python used in prediction.

2368

#

2369

# The following Python versions are available:

2370

#

2371

# * Python '3.7' is available when `runtime_version` is set to '1.15' or

2372

# later.

2373

# * Python '3.5' is available when `runtime_version` is set to a version

2374

# from '1.4' to '1.14'.

2375

# * Python '2.7' is available when `runtime_version` is set to '1.15' or

2376

# earlier.

2377

#

2378

# Read more about the Python versions available for [each runtime

2379

# version](/ml-engine/docs/runtime-version-list).

2380

"requestLoggingConfig": { # Configuration for logging request-response pairs to a BigQuery table. # Optional. *Only* specify this field in a

2381

# projects.models.versions.patch

2382

# request. Specifying it in a

2383

# projects.models.versions.create

2384

# request has no effect.

2385

#

2386

# Configures the request-response pair logging on predictions from this

2387

# Version.

2388

# Online prediction requests to a model version and the responses to these

2389

# requests are converted to raw strings and saved to the specified BigQuery

2390

# table. Logging is constrained by [BigQuery quotas and

2391

# limits](/bigquery/quotas). If your project exceeds BigQuery quotas or limits,

2392

# AI Platform Prediction does not log request-response pairs, but it continues

2393

# to serve predictions.

2394

#

2395

# If you are using [continuous

2396

# evaluation](/ml-engine/docs/continuous-evaluation/), you do not need to

2397

# specify this configuration manually. Setting up continuous evaluation

2398

# automatically enables logging of request-response pairs.

2399

"samplingPercentage": 3.14, # Percentage of requests to be logged, expressed as a fraction from 0 to 1.

2400

# For example, if you want to log 10% of requests, enter `0.1`. The sampling

2401

# window is the lifetime of the model version. Defaults to 0.

2402

"bigqueryTableName": "A String", # Required. Fully qualified BigQuery table name in the following format:

2403

# "<var>project_id</var>.<var>dataset_name</var>.<var>table_name</var>"

2404

#

2405

# The specified table must already exist, and the "Cloud ML Service Agent"

2406

# for your project must have permission to write to it. The table must have

2407

# the following [schema](/bigquery/docs/schemas):

2408

#

2409

# <table>

2410

# <tr><th>Field name</th><th style="display: table-cell">Type</th>

2411

# <th style="display: table-cell">Mode</th></tr>

2412

# <tr><td>model</td><td>STRING</td><td>REQUIRED</td></tr>

2413

# <tr><td>model_version</td><td>STRING</td><td>REQUIRED</td></tr>

2414

# <tr><td>time</td><td>TIMESTAMP</td><td>REQUIRED</td></tr>

2415

# <tr><td>raw_data</td><td>STRING</td><td>REQUIRED</td></tr>

2416

# <tr><td>raw_prediction</td><td>STRING</td><td>NULLABLE</td></tr>

2417

# <tr><td>groundtruth</td><td>STRING</td><td>NULLABLE</td></tr>

2418

# </table>

2419

},

2420

"manualScaling": { # Options for manually scaling a model. # Manually select the number of nodes to use for serving the

2421

# model. You should generally use `auto_scaling` with an appropriate

2422

# `min_nodes` instead, but this option is available if you want more

2423

# predictable billing. Beware that latency and error rates will increase

2424

# if the traffic exceeds that capability of the system to serve it based

2425

# on the selected number of nodes.

2426

"nodes": 42, # The number of nodes to allocate for this model. These nodes are always up,

2427

# starting from the time the model is deployed, so the cost of operating

2428

# this model will be proportional to `nodes` * number of hours since

2429

# last billing cycle plus the cost for each prediction performed.

2430

},

2431

"createTime": "A String", # Output only. The time the version was created.

2432

"lastUseTime": "A String", # Output only. The time the version was last used for prediction.

2433

"framework": "A String", # Optional. The machine learning framework AI Platform uses to train

2434

# this version of the model. Valid values are `TENSORFLOW`, `SCIKIT_LEARN`,

2435

# `XGBOOST`. If you do not specify a framework, AI Platform

2436

# will analyze files in the deployment_uri to determine a framework. If you

2437

# choose `SCIKIT_LEARN` or `XGBOOST`, you must also set the runtime version

2438

# of the model to 1.4 or greater.

2439

#

2440

# Do **not** specify a framework if you're deploying a [custom

2441

# prediction routine](/ml-engine/docs/tensorflow/custom-prediction-routines).

2442

#

2443

# If you specify a [Compute Engine (N1) machine

2444

# type](/ml-engine/docs/machine-types-online-prediction) in the

2445

# `machineType` field, you must specify `TENSORFLOW`

2446

# for the framework.

2447

"predictionClass": "A String", # Optional. The fully qualified name

2448

# (<var>module_name</var>.<var>class_name</var>) of a class that implements

2449

# the Predictor interface described in this reference field. The module

2450

# containing this class should be included in a package provided to the

2451

# [`packageUris` field](#Version.FIELDS.package_uris).

2452

#

2453

# Specify this field if and only if you are deploying a [custom prediction

2454

# routine (beta)](/ml-engine/docs/tensorflow/custom-prediction-routines).

2455

# If you specify this field, you must set

2456

# [`runtimeVersion`](#Version.FIELDS.runtime_version) to 1.4 or greater and

2457

# you must set `machineType` to a [legacy (MLS1)

2458

# machine type](/ml-engine/docs/machine-types-online-prediction).

2459

#

2460

# The following code sample provides the Predictor interface:

2461

#

2462

# <pre style="max-width: 626px;">

2463

# class Predictor(object):

2464

# """Interface for constructing custom predictors."""

2465

#

2466

# def predict(self, instances, **kwargs):

2467

# """Performs custom prediction.

2468

#

2469

# Instances are the decoded values from the request. They have already

2470

# been deserialized from JSON.

2471

#

2472

# Args:

2473

# instances: A list of prediction input instances.

2474

# **kwargs: A dictionary of keyword args provided as additional

2475

# fields on the predict request body.

2476

#

2477

# Returns:

2478

# A list of outputs containing the prediction results. This list must

2479

# be JSON serializable.

2480

# """

2481

# raise NotImplementedError()

2482

#

2483

# @classmethod

2484

# def from_path(cls, model_dir):

2485

# """Creates an instance of Predictor using the given path.

2486

#

2487

# Loading of the predictor should be done in this method.

2488

#

2489

# Args:

2490

# model_dir: The local directory that contains the exported model

2491

# file along with any additional files uploaded when creating the

# version resource.

#

# Returns:

# An instance implementing this Predictor class.

2496

# """

2497

# raise NotImplementedError()

2498

# </pre>

2499

#

2500

# Learn more about [the Predictor interface and custom prediction

2501

# routines](/ml-engine/docs/tensorflow/custom-prediction-routines).

2502

"isDefault": True or False, # Output only. If true, this version will be used to handle prediction

2503

# requests that do not specify a version.

2504

#

2505

# You can change the default version by calling

2506

# projects.methods.versions.setDefault.

2507

"etag": "A String", # `etag` is used for optimistic concurrency control as a way to help

2508

# prevent simultaneous updates of a model from overwriting each other.

2509

# It is strongly suggested that systems make use of the `etag` in the

2510

# read-modify-write cycle to perform model updates in order to avoid race

2511

# conditions: An `etag` is returned in the response to `GetVersion`, and

2512

# systems are expected to put that etag in the request to `UpdateVersion` to

2513

# ensure that their change will be applied to the model as intended.

2514

"serviceAccount": "A String", # Optional. Specifies the service account for resource access control.

2515

"errorMessage": "A String", # Output only. The details of a failure or a cancellation.

2516

"deploymentUri": "A String", # Required. The Cloud Storage location of the trained model used to

2517

# create the version. See the

2518

# [guide to model

2519

# deployment](/ml-engine/docs/tensorflow/deploying-models) for more

2520

# information.

2521

#

2522

# When passing Version to

2523

# projects.models.versions.create

2524

# the model service uses the specified location as the source of the model.

2525

# Once deployed, the model version is hosted by the prediction service, so

2526

# this location is useful only as a historical record.

2527

# The total number of model files can't exceed 1000.

2528

"runtimeVersion": "A String", # Required. The AI Platform runtime version to use for this deployment.

2529

#

2530

# For more information, see the

2531

# [runtime version list](/ml-engine/docs/runtime-version-list) and

2532

# [how to manage runtime versions](/ml-engine/docs/versioning).

2533

"description": "A String", # Optional. The description specified for the version when it was created.

2534

},

2535

"onlinePredictionLogging": True or False, # Optional. If true, online prediction access logs are sent to StackDriver

2536

# Logging. These logs are like standard server access logs, containing

2537

# information like timestamp and latency for each request. Note that

2538

# [Stackdriver logs may incur a cost](/stackdriver/pricing), especially if

2539

# your project receives prediction requests at a high queries per second rate

2540

# (QPS). Estimate your costs before enabling this option.

2541

#

2542

# Default is false.

2543

}

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2544

2545

updateMask: string, Required. Specifies the path, relative to `Model`, of the field to update.

2546

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2547

For example, to change the description of a model to "foo" and set its

2548

default version to "version_1", the `update_mask` parameter would be

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2549

specified as `description`, `default_version.name`, and the `PATCH`

2550

request body would specify the new value, as follows:

2551

{

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2552

"description": "foo",

2553

"defaultVersion": {

2554

"name":"version_1"

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

}

}

Currently the supported update masks are `description` and

2559

`default_version.name`.

2560

x__xgafv: string, V1 error format.

Allowed values

1 - v1 error format

2 - v2 error format

Returns:

An object of the form:

2567

2568

{ # This resource represents a long-running operation that is the result of a

2569

# network API call.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2570

"error": { # The `Status` type defines a logical error model that is suitable for # The error result of the operation in case of failure or cancellation.

2571

# different programming environments, including REST APIs and RPC APIs. It is

2572

# used by [gRPC](https://github.com/grpc). Each `Status` message contains

2573

# three pieces of data: error code, error message, and error details.

2574

#

2575

# You can find out more about this error model and how to work with it in the

2576

# [API Design Guide](https://cloud.google.com/apis/design/errors).

2577

"details": [ # A list of messages that carry the error details. There is a common set of

2578

# message types for APIs to use.

2579

{

2580

"a_key": "", # Properties of the object. Contains field @type with type URL.

2581

},

2582

],

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

2583

"message": "A String", # A developer-facing error message, which should be in English. Any

2584

# user-facing error message should be localized and sent in the

2585

# google.rpc.Status.details field, or localized by the client.

2586

"code": 42, # The status code, which should be an enum value of google.rpc.Code.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2587

},

2588

"done": True or False, # If the value is `false`, it means the operation is still in progress.

2589

# If `true`, the operation is completed, and either `error` or `response` is

2590

# available.

Bu Sun Kim

2020-05-27 12:20:54 -0700

[diff] [blame]

2591

"response": { # The normal response of the operation in case of success. If the original

2592

# method returns no data on success, such as `Delete`, the response is

2593

# `google.protobuf.Empty`. If the original method is standard

2594

# `Get`/`Create`/`Update`, the response should be the resource. For other

2595

# methods, the response should have the type `XxxResponse`, where `Xxx`

2596

# is the original method name. For example, if the original method name

2597

# is `TakeSnapshot()`, the inferred response type is

2598

# `TakeSnapshotResponse`.

2599

"a_key": "", # Properties of the object. Contains field @type with type URL.

2600

},

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

2601

"metadata": { # Service-specific metadata associated with the operation. It typically

2602

# contains progress information and common metadata such as create time.

2603

# Some services might not provide such metadata. Any method that returns a

2604

# long-running operation should document the metadata type, if any.

2605

"a_key": "", # Properties of the object. Contains field @type with type URL.

2606

},

2607

"name": "A String", # The server-assigned name, which is only unique within the same service that

2608

# originally returns it. If you use the default HTTP mapping, the

2609

# `name` should be a resource name ending with `operations/{unique_id}`.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

}</pre>

</div>

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2614

<code class="details" id="setIamPolicy">setIamPolicy(resource, body=None, x__xgafv=None)</code>

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2615

<pre>Sets the access control policy on the specified resource. Replaces any

2616

existing policy.

2617

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2618

Can return `NOT_FOUND`, `INVALID_ARGUMENT`, and `PERMISSION_DENIED` errors.

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2619

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2620

Args:

2621

resource: string, REQUIRED: The resource for which the policy is being specified.

2622

See the operation documentation for the appropriate value for this field. (required)

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2623

body: object, The request body.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2624

The object takes the form of:

2625

2626

{ # Request message for `SetIamPolicy` method.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2627

"policy": { # An Identity and Access Management (IAM) policy, which specifies access # REQUIRED: The complete policy to be applied to the `resource`. The size of

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2628

# the policy is limited to a few 10s of KB. An empty policy is a

2629

# valid policy but certain Cloud Platform services (such as Projects)

2630

# might reject them.

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2631

# controls for Google Cloud resources.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2632

#

2633

#

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2634

# A `Policy` is a collection of `bindings`. A `binding` binds one or more

2635

# `members` to a single `role`. Members can be user accounts, service accounts,

2636

# Google groups, and domains (such as G Suite). A `role` is a named list of

2637

# permissions; each `role` can be an IAM predefined role or a user-created

2638

# custom role.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2639

#

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2640

# For some types of Google Cloud resources, a `binding` can also specify a

2641

# `condition`, which is a logical expression that allows access to a resource

2642

# only if the expression evaluates to `true`. A condition can add constraints

2643

# based on attributes of the request, the resource, or both. To learn which

2644

# resources support conditions in their IAM policies, see the

2645

# [IAM documentation](https://cloud.google.com/iam/help/conditions/resource-policies).

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2646

#

2647

# **JSON example:**

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2648

#

2649

# {

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2650

# "bindings": [

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2651

# {

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2652

# "role": "roles/resourcemanager.organizationAdmin",

2653

# "members": [

2654

# "user:mike@example.com",

2655

# "group:admins@example.com",

2656

# "domain:google.com",

2657

# "serviceAccount:my-project-id@appspot.gserviceaccount.com"

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2658

# ]

2659

# },

2660

# {

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2661

# "role": "roles/resourcemanager.organizationViewer",

2662

# "members": [

2663

# "user:eve@example.com"

2664

# ],

2665

# "condition": {

2666

# "title": "expirable access",

2667

# "description": "Does not grant access after Sep 2020",

2668

# "expression": "request.time < timestamp('2020-10-01T00:00:00.000Z')",

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2669

# }

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2670

# }

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2671

# ],

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2672

# "etag": "BwWWja0YfJA=",

2673

# "version": 3

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2674

# }

2675

#

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2676

# **YAML example:**

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

#

# bindings:

# - members:

# - user:mike@example.com

2681

# - group:admins@example.com

2682

# - domain:google.com

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2683

# - serviceAccount:my-project-id@appspot.gserviceaccount.com

2684

# role: roles/resourcemanager.organizationAdmin

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2685

# - members:

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2686

# - user:eve@example.com

2687

# role: roles/resourcemanager.organizationViewer

2688

# condition:

2689

# title: expirable access

2690

# description: Does not grant access after Sep 2020

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2691

# expression: request.time < timestamp('2020-10-01T00:00:00.000Z')

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2692

# - etag: BwWWja0YfJA=

2693

# - version: 3

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2694

#

2695

# For a description of IAM and its features, see the

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2696

# [IAM documentation](https://cloud.google.com/iam/docs/).

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

2697

"etag": "A String", # `etag` is used for optimistic concurrency control as a way to help

2698

# prevent simultaneous updates of a policy from overwriting each other.

2699

# It is strongly suggested that systems make use of the `etag` in the

2700

# read-modify-write cycle to perform policy updates in order to avoid race

2701

# conditions: An `etag` is returned in the response to `getIamPolicy`, and

2702

# systems are expected to put that etag in the request to `setIamPolicy` to

2703

# ensure that their change will be applied to the same version of the policy.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2704

#

2705

# **Important:** If you use IAM Conditions, you must include the `etag` field

2706

# whenever you call `setIamPolicy`. If you omit this field, then IAM allows

2707

# you to overwrite a version `3` policy with a version `1` policy, and all of

2708

# the conditions in the version `3` policy are lost.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2709

"auditConfigs": [ # Specifies cloud audit logging configuration for this policy.

2710

{ # Specifies the audit configuration for a service.

2711

# The configuration determines which permission types are logged, and what

2712

# identities, if any, are exempted from logging.

2713

# An AuditConfig must have one or more AuditLogConfigs.

2714

#

2715

# If there are AuditConfigs for both `allServices` and a specific service,

2716

# the union of the two AuditConfigs is used for that service: the log_types

2717

# specified in each AuditConfig are enabled, and the exempted_members in each

2718

# AuditLogConfig are exempted.

2719

#

2720

# Example Policy with multiple AuditConfigs:

#

# {

# "audit_configs": [

# {

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

2725

# "service": "allServices",

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2726

# "audit_log_configs": [

2727

# {

2728

# "log_type": "DATA_READ",

2729

# "exempted_members": [

2730

# "user:jose@example.com"

2731

# ]

2732

# },

2733

# {

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

2734

# "log_type": "DATA_WRITE"

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2735

# },

2736

# {

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

2737

# "log_type": "ADMIN_READ"

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

# }

# ]

# },

# {

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

2742

# "service": "sampleservice.googleapis.com",

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2743

# "audit_log_configs": [

2744

# {

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

2745

# "log_type": "DATA_READ"

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2746

# },

2747

# {

2748

# "log_type": "DATA_WRITE",

2749

# "exempted_members": [

2750

# "user:aliya@example.com"

# ]

# }

# ]

# }

# ]

# }

#

# For sampleservice, this policy enables DATA_READ, DATA_WRITE and ADMIN_READ

2759

# logging. It also exempts jose@example.com from DATA_READ logging, and

2760

# aliya@example.com from DATA_WRITE logging.

2761

"service": "A String", # Specifies a service that will be enabled for audit logging.

2762

# For example, `storage.googleapis.com`, `cloudsql.googleapis.com`.

2763

# `allServices` is a special value that covers all services.

2764

"auditLogConfigs": [ # The configuration for logging of each type of permission.

2765

{ # Provides the configuration for logging a type of permissions.

# Example:

#

# {

# "audit_log_configs": [

2770

# {

2771

# "log_type": "DATA_READ",

2772

# "exempted_members": [

2773

# "user:jose@example.com"

2774

# ]

2775

# },

2776

# {

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

2777

# "log_type": "DATA_WRITE"

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

# }

# ]

# }

#

# This enables 'DATA_READ' and 'DATA_WRITE' logging, while exempting

2783

# jose@example.com from DATA_READ logging.

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

2784

"logType": "A String", # The log type that this config enables.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2785

"exemptedMembers": [ # Specifies the identities that do not cause logging for this type of

2786

# permission.

2787

# Follows the same format of Binding.members.

2788

"A String",

2789

],

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

},

],

},

],

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

2794

"version": 42, # Specifies the format of the policy.

2795

#

2796

# Valid values are `0`, `1`, and `3`. Requests that specify an invalid value

2797

# are rejected.

2798

#

2799

# Any operation that affects conditional role bindings must specify version

2800

# `3`. This requirement applies to the following operations:

2801

#

2802

# * Getting a policy that includes a conditional role binding

2803

# * Adding a conditional role binding to a policy

2804

# * Changing a conditional role binding in a policy

2805

# * Removing any role binding, with or without a condition, from a policy

2806

# that includes conditions

2807

#

2808

# **Important:** If you use IAM Conditions, you must include the `etag` field

2809

# whenever you call `setIamPolicy`. If you omit this field, then IAM allows

2810

# you to overwrite a version `3` policy with a version `1` policy, and all of

2811

# the conditions in the version `3` policy are lost.

2812

#

2813

# If a policy does not include any conditions, operations on that policy may

2814

# specify any valid version or leave the field unset.

2815

#

2816

# To learn which resources support conditions in their IAM policies, see the

2817

# [IAM documentation](https://cloud.google.com/iam/help/conditions/resource-policies).

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2818

"bindings": [ # Associates a list of `members` to a `role`. Optionally, may specify a

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2819

# `condition` that determines how and when the `bindings` are applied. Each

2820

# of the `bindings` must contain at least one member.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2821

{ # Associates `members` with a `role`.

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

2822

"role": "A String", # Role that is assigned to `members`.

2823

# For example, `roles/viewer`, `roles/editor`, or `roles/owner`.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2824

"condition": { # Represents a textual expression in the Common Expression Language (CEL) # The condition that is associated with this binding.

2825

#

2826

# If the condition evaluates to `true`, then this binding applies to the

2827

# current request.

2828

#

2829

# If the condition evaluates to `false`, then this binding does not apply to

2830

# the current request. However, a different role binding might grant the same

2831

# role to one or more of the members in this binding.

2832

#

2833

# To learn which resources support conditions in their IAM policies, see the

2834

# [IAM

2835

# documentation](https://cloud.google.com/iam/help/conditions/resource-policies).

2836

# syntax. CEL is a C-like expression language. The syntax and semantics of CEL

2837

# are documented at https://github.com/google/cel-spec.

2838

#

2839

# Example (Comparison):

2840

#

2841

# title: "Summary size limit"

2842

# description: "Determines if a summary is less than 100 chars"

2843

# expression: "document.summary.size() < 100"

2844

#

2845

# Example (Equality):

2846

#

2847

# title: "Requestor is owner"

2848

# description: "Determines if requestor is the document owner"

2849

# expression: "document.owner == request.auth.claims.email"

#

# Example (Logic):

#

# title: "Public documents"

2854

# description: "Determine whether the document should be publicly visible"

2855

# expression: "document.type != 'private' && document.type != 'internal'"

2856

#

2857

# Example (Data Manipulation):

2858

#

2859

# title: "Notification string"

2860

# description: "Create a notification string with a timestamp."

2861

# expression: "'New message received at ' + string(document.create_time)"

2862

#

2863

# The exact variables and functions that may be referenced within an expression

2864

# are determined by the service that evaluates it. See the service

2865

# documentation for additional information.

Bu Sun Kim

2020-05-27 12:20:54 -0700

[diff] [blame]

2866

"expression": "A String", # Textual representation of an expression in Common Expression Language

2867

# syntax.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2868

"title": "A String", # Optional. Title for the expression, i.e. a short string describing

2869

# its purpose. This can be used e.g. in UIs which allow to enter the

2870

# expression.

2871

"location": "A String", # Optional. String indicating the location of the expression for error

2872

# reporting, e.g. a file name and a position in the file.

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

2873

"description": "A String", # Optional. Description of the expression. This is a longer text which

2874

# describes the expression, e.g. when hovered over it in a UI.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2875

},

2876

"members": [ # Specifies the identities requesting access for a Cloud Platform resource.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2877

# `members` can have the following values:

2878

#

2879

# * `allUsers`: A special identifier that represents anyone who is

2880

# on the internet; with or without a Google account.

2881

#

2882

# * `allAuthenticatedUsers`: A special identifier that represents anyone

2883

# who is authenticated with a Google account or a service account.

2884

#

2885

# * `user:{emailid}`: An email address that represents a specific Google

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2886

# account. For example, `alice@example.com` .

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2887

#

2888

#

2889

# * `serviceAccount:{emailid}`: An email address that represents a service

2890

# account. For example, `my-other-app@appspot.gserviceaccount.com`.

2891

#

2892

# * `group:{emailid}`: An email address that represents a Google group.

2893

# For example, `admins@example.com`.

2894

#

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2895

# * `deleted:user:{emailid}?uid={uniqueid}`: An email address (plus unique

2896

# identifier) representing a user that has been recently deleted. For

2897

# example, `alice@example.com?uid=123456789012345678901`. If the user is

2898

# recovered, this value reverts to `user:{emailid}` and the recovered user

2899

# retains the role in the binding.

2900

#

2901

# * `deleted:serviceAccount:{emailid}?uid={uniqueid}`: An email address (plus

2902

# unique identifier) representing a service account that has been recently

2903

# deleted. For example,

2904

# `my-other-app@appspot.gserviceaccount.com?uid=123456789012345678901`.

2905

# If the service account is undeleted, this value reverts to

2906

# `serviceAccount:{emailid}` and the undeleted service account retains the

2907

# role in the binding.

2908

#

2909

# * `deleted:group:{emailid}?uid={uniqueid}`: An email address (plus unique

2910

# identifier) representing a Google group that has been recently

2911

# deleted. For example, `admins@example.com?uid=123456789012345678901`. If

2912

# the group is recovered, this value reverts to `group:{emailid}` and the

2913

# recovered group retains the role in the binding.

2914

#

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2915

#

2916

# * `domain:{domain}`: The G Suite domain (primary) that represents all the

2917

# users of that domain. For example, `google.com` or `example.com`.

2918

#

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2919

"A String",

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2920

],

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2921

},

2922

],

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2923

},

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

2924

"updateMask": "A String", # OPTIONAL: A FieldMask specifying which fields of the policy to modify. Only

2925

# the fields in the mask will be modified. If no mask is provided, the

2926

# following default mask is used:

2927

#

2928

# `paths: "bindings, etag"`

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2929

}

2930

2931

x__xgafv: string, V1 error format.

Allowed values

1 - v1 error format

2 - v2 error format

Returns:

An object of the form:

2938

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2939

{ # An Identity and Access Management (IAM) policy, which specifies access

2940

# controls for Google Cloud resources.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2941

#

2942

#

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2943

# A `Policy` is a collection of `bindings`. A `binding` binds one or more

2944

# `members` to a single `role`. Members can be user accounts, service accounts,

2945

# Google groups, and domains (such as G Suite). A `role` is a named list of

2946

# permissions; each `role` can be an IAM predefined role or a user-created

2947

# custom role.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2948

#

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2949

# For some types of Google Cloud resources, a `binding` can also specify a

2950

# `condition`, which is a logical expression that allows access to a resource

2951

# only if the expression evaluates to `true`. A condition can add constraints

2952

# based on attributes of the request, the resource, or both. To learn which

2953

# resources support conditions in their IAM policies, see the

2954

# [IAM documentation](https://cloud.google.com/iam/help/conditions/resource-policies).

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2955

#

2956

# **JSON example:**

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2957

#

2958

# {

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2959

# "bindings": [

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2960

# {

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2961

# "role": "roles/resourcemanager.organizationAdmin",

2962

# "members": [

2963

# "user:mike@example.com",

2964

# "group:admins@example.com",

2965

# "domain:google.com",

2966

# "serviceAccount:my-project-id@appspot.gserviceaccount.com"

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2967

# ]

2968

# },

2969

# {

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2970

# "role": "roles/resourcemanager.organizationViewer",

2971

# "members": [

2972

# "user:eve@example.com"

2973

# ],

2974

# "condition": {

2975

# "title": "expirable access",

2976

# "description": "Does not grant access after Sep 2020",

2977

# "expression": "request.time < timestamp('2020-10-01T00:00:00.000Z')",

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2978

# }

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2979

# }

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2980

# ],

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

2981

# "etag": "BwWWja0YfJA=",

2982

# "version": 3

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2983

# }

2984

#

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2985

# **YAML example:**

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

#

# bindings:

# - members:

# - user:mike@example.com

2990

# - group:admins@example.com

2991

# - domain:google.com

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2992

# - serviceAccount:my-project-id@appspot.gserviceaccount.com

2993

# role: roles/resourcemanager.organizationAdmin

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

2994

# - members:

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

2995

# - user:eve@example.com

2996

# role: roles/resourcemanager.organizationViewer

2997

# condition:

2998

# title: expirable access

2999

# description: Does not grant access after Sep 2020

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

3000

# expression: request.time < timestamp('2020-10-01T00:00:00.000Z')

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

3001

# - etag: BwWWja0YfJA=

3002

# - version: 3

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

3003

#

3004

# For a description of IAM and its features, see the

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

3005

# [IAM documentation](https://cloud.google.com/iam/docs/).

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

3006

"etag": "A String", # `etag` is used for optimistic concurrency control as a way to help

3007

# prevent simultaneous updates of a policy from overwriting each other.

3008

# It is strongly suggested that systems make use of the `etag` in the

3009

# read-modify-write cycle to perform policy updates in order to avoid race

3010

# conditions: An `etag` is returned in the response to `getIamPolicy`, and

3011

# systems are expected to put that etag in the request to `setIamPolicy` to

3012

# ensure that their change will be applied to the same version of the policy.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

3013

#

3014

# **Important:** If you use IAM Conditions, you must include the `etag` field

3015

# whenever you call `setIamPolicy`. If you omit this field, then IAM allows

3016

# you to overwrite a version `3` policy with a version `1` policy, and all of

3017

# the conditions in the version `3` policy are lost.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

3018

"auditConfigs": [ # Specifies cloud audit logging configuration for this policy.

3019

{ # Specifies the audit configuration for a service.

3020

# The configuration determines which permission types are logged, and what

3021

# identities, if any, are exempted from logging.

3022

# An AuditConfig must have one or more AuditLogConfigs.

3023

#

3024

# If there are AuditConfigs for both `allServices` and a specific service,

3025

# the union of the two AuditConfigs is used for that service: the log_types

3026

# specified in each AuditConfig are enabled, and the exempted_members in each

3027

# AuditLogConfig are exempted.

3028

#

3029

# Example Policy with multiple AuditConfigs:

#

# {

# "audit_configs": [

# {

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

3034

# "service": "allServices",

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

3035

# "audit_log_configs": [

3036

# {

3037

# "log_type": "DATA_READ",

3038

# "exempted_members": [

3039

# "user:jose@example.com"

3040

# ]

3041

# },

3042

# {

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

3043

# "log_type": "DATA_WRITE"

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

3044

# },

3045

# {

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

3046

# "log_type": "ADMIN_READ"

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

# }

# ]

# },

# {

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

3051

# "service": "sampleservice.googleapis.com",

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

3052

# "audit_log_configs": [

3053

# {

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

3054

# "log_type": "DATA_READ"

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

3055

# },

3056

# {

3057

# "log_type": "DATA_WRITE",

3058

# "exempted_members": [

3059

# "user:aliya@example.com"

# ]

# }

# ]

# }

# ]

# }

#

# For sampleservice, this policy enables DATA_READ, DATA_WRITE and ADMIN_READ

3068

# logging. It also exempts jose@example.com from DATA_READ logging, and

3069

# aliya@example.com from DATA_WRITE logging.

3070

"service": "A String", # Specifies a service that will be enabled for audit logging.

3071

# For example, `storage.googleapis.com`, `cloudsql.googleapis.com`.

3072

# `allServices` is a special value that covers all services.

3073

"auditLogConfigs": [ # The configuration for logging of each type of permission.

3074

{ # Provides the configuration for logging a type of permissions.

# Example:

#

# {

# "audit_log_configs": [

3079

# {

3080

# "log_type": "DATA_READ",

3081

# "exempted_members": [

3082

# "user:jose@example.com"

3083

# ]

3084

# },

3085

# {

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

3086

# "log_type": "DATA_WRITE"

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

# }

# ]

# }

#

# This enables 'DATA_READ' and 'DATA_WRITE' logging, while exempting

3092

# jose@example.com from DATA_READ logging.

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

3093

"logType": "A String", # The log type that this config enables.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

3094

"exemptedMembers": [ # Specifies the identities that do not cause logging for this type of

3095

# permission.

3096

# Follows the same format of Binding.members.

3097

"A String",

3098

],

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

},

],

},

],

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

3103

"version": 42, # Specifies the format of the policy.

3104

#

3105

# Valid values are `0`, `1`, and `3`. Requests that specify an invalid value

3106

# are rejected.

3107

#

3108

# Any operation that affects conditional role bindings must specify version

3109

# `3`. This requirement applies to the following operations:

3110

#

3111

# * Getting a policy that includes a conditional role binding

3112

# * Adding a conditional role binding to a policy

3113

# * Changing a conditional role binding in a policy

3114

# * Removing any role binding, with or without a condition, from a policy

3115

# that includes conditions

3116

#

3117

# **Important:** If you use IAM Conditions, you must include the `etag` field

3118

# whenever you call `setIamPolicy`. If you omit this field, then IAM allows

3119

# you to overwrite a version `3` policy with a version `1` policy, and all of

3120

# the conditions in the version `3` policy are lost.

3121

#

3122

# If a policy does not include any conditions, operations on that policy may

3123

# specify any valid version or leave the field unset.

3124

#

3125

# To learn which resources support conditions in their IAM policies, see the

3126

# [IAM documentation](https://cloud.google.com/iam/help/conditions/resource-policies).

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

3127

"bindings": [ # Associates a list of `members` to a `role`. Optionally, may specify a

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

3128

# `condition` that determines how and when the `bindings` are applied. Each

3129

# of the `bindings` must contain at least one member.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

3130

{ # Associates `members` with a `role`.

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

3131

"role": "A String", # Role that is assigned to `members`.

3132

# For example, `roles/viewer`, `roles/editor`, or `roles/owner`.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

3133

"condition": { # Represents a textual expression in the Common Expression Language (CEL) # The condition that is associated with this binding.

3134

#

3135

# If the condition evaluates to `true`, then this binding applies to the

3136

# current request.

3137

#

3138

# If the condition evaluates to `false`, then this binding does not apply to

3139

# the current request. However, a different role binding might grant the same

3140

# role to one or more of the members in this binding.

3141

#

3142

# To learn which resources support conditions in their IAM policies, see the

3143

# [IAM

3144

# documentation](https://cloud.google.com/iam/help/conditions/resource-policies).

3145

# syntax. CEL is a C-like expression language. The syntax and semantics of CEL

3146

# are documented at https://github.com/google/cel-spec.

3147

#

3148

# Example (Comparison):

3149

#

3150

# title: "Summary size limit"

3151

# description: "Determines if a summary is less than 100 chars"

3152

# expression: "document.summary.size() < 100"

3153

#

3154

# Example (Equality):

3155

#

3156

# title: "Requestor is owner"

3157

# description: "Determines if requestor is the document owner"

3158

# expression: "document.owner == request.auth.claims.email"

#

# Example (Logic):

#

# title: "Public documents"

3163

# description: "Determine whether the document should be publicly visible"

3164

# expression: "document.type != 'private' && document.type != 'internal'"

3165

#

3166

# Example (Data Manipulation):

3167

#

3168

# title: "Notification string"

3169

# description: "Create a notification string with a timestamp."

3170

# expression: "'New message received at ' + string(document.create_time)"

3171

#

3172

# The exact variables and functions that may be referenced within an expression

3173

# are determined by the service that evaluates it. See the service

3174

# documentation for additional information.

Bu Sun Kim

2020-05-27 12:20:54 -0700

[diff] [blame]

3175

"expression": "A String", # Textual representation of an expression in Common Expression Language

3176

# syntax.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

3177

"title": "A String", # Optional. Title for the expression, i.e. a short string describing

3178

# its purpose. This can be used e.g. in UIs which allow to enter the

3179

# expression.

3180

"location": "A String", # Optional. String indicating the location of the expression for error

3181

# reporting, e.g. a file name and a position in the file.

Bu Sun Kim

2020-07-22 17:02:09 -0700

[diff] [blame^]

3182

"description": "A String", # Optional. Description of the expression. This is a longer text which

3183

# describes the expression, e.g. when hovered over it in a UI.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

3184

},

3185

"members": [ # Specifies the identities requesting access for a Cloud Platform resource.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

3186

# `members` can have the following values:

3187

#

3188

# * `allUsers`: A special identifier that represents anyone who is

3189

# on the internet; with or without a Google account.

3190

#

3191

# * `allAuthenticatedUsers`: A special identifier that represents anyone

3192

# who is authenticated with a Google account or a service account.

3193

#

3194

# * `user:{emailid}`: An email address that represents a specific Google

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

3195

# account. For example, `alice@example.com` .

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

3196

#

3197

#

3198

# * `serviceAccount:{emailid}`: An email address that represents a service

3199

# account. For example, `my-other-app@appspot.gserviceaccount.com`.

3200

#

3201

# * `group:{emailid}`: An email address that represents a Google group.

3202

# For example, `admins@example.com`.

3203

#

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

3204

# * `deleted:user:{emailid}?uid={uniqueid}`: An email address (plus unique

3205

# identifier) representing a user that has been recently deleted. For

3206

# example, `alice@example.com?uid=123456789012345678901`. If the user is

3207

# recovered, this value reverts to `user:{emailid}` and the recovered user

3208

# retains the role in the binding.

3209

#

3210

# * `deleted:serviceAccount:{emailid}?uid={uniqueid}`: An email address (plus

3211

# unique identifier) representing a service account that has been recently

3212

# deleted. For example,

3213

# `my-other-app@appspot.gserviceaccount.com?uid=123456789012345678901`.

3214

# If the service account is undeleted, this value reverts to

3215

# `serviceAccount:{emailid}` and the undeleted service account retains the

3216

# role in the binding.

3217

#

3218

# * `deleted:group:{emailid}?uid={uniqueid}`: An email address (plus unique

3219

# identifier) representing a Google group that has been recently

3220

# deleted. For example, `admins@example.com?uid=123456789012345678901`. If

3221

# the group is recovered, this value reverts to `group:{emailid}` and the

3222

# recovered group retains the role in the binding.

3223

#

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

3224

#

3225

# * `domain:{domain}`: The G Suite domain (primary) that represents all the

3226

# users of that domain. For example, `google.com` or `example.com`.

3227

#

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

3228

"A String",

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

3229

],

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

3230

},

3231

],

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

}</pre>

</div>

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

3236

<code class="details" id="testIamPermissions">testIamPermissions(resource, body=None, x__xgafv=None)</code>

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

3237

<pre>Returns permissions that a caller has on the specified resource.

3238

If the resource does not exist, this will return an empty set of

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

3239

permissions, not a `NOT_FOUND` error.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

3240

3241

Note: This operation is designed to be used for building permission-aware

3242

UIs and command-line tools, not for authorization checking. This operation

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

3243

may "fail open" without warning.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

3244

3245

Args:

3246

resource: string, REQUIRED: The resource for which the policy detail is being requested.

3247

See the operation documentation for the appropriate value for this field. (required)

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

3248

body: object, The request body.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

3249

The object takes the form of:

3250

3251

{ # Request message for `TestIamPermissions` method.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

3252

"permissions": [ # The set of permissions to check for the `resource`. Permissions with

3253

# wildcards (such as '*' or 'storage.*') are not allowed. For more

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

3254

# information see

3255

# [IAM Overview](https://cloud.google.com/iam/docs/overview#permissions).

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

3256

"A String",

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

],

}

x__xgafv: string, V1 error format.

Allowed values

1 - v1 error format

2 - v2 error format

Returns:

An object of the form:

3267

3268

{ # Response message for `TestIamPermissions` method.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

3269

"permissions": [ # A subset of `TestPermissionsRequest.permissions` that the caller is

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

3270

# allowed.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

3271

"A String",

Bu Sun Kim