MCPcopy Create free account
hub / github.com/DeepRec-AI/DeepRec / Options

Class Options

tensorflow/python/data/ops/dataset_ops.py:2297–2411  ·  view source on GitHub ↗

Represents options for tf.data.Dataset. An `Options` object can be, for instance, used to control which static optimizations to apply or whether to use performance modeling to dynamically tune the parallelism of operations such as `tf.data.Dataset.map` or `tf.data.Dataset.interleave`.

Source from the content-addressed store, hash-verified

2295
2296@tf_export("data.Options")
2297class Options(options_lib.OptionsBase):
2298 """Represents options for tf.data.Dataset.
2299
2300 An `Options` object can be, for instance, used to control which static
2301 optimizations to apply or whether to use performance modeling to dynamically
2302 tune the parallelism of operations such as `tf.data.Dataset.map` or
2303 `tf.data.Dataset.interleave`.
2304 """
2305
2306 experimental_deterministic = options_lib.create_option(
2307 name="experimental_deterministic",
2308 ty=bool,
2309 docstring=
2310 "Whether the outputs need to be produced in deterministic order. If None,"
2311 " defaults to True.")
2312
2313 experimental_distribute = options_lib.create_option(
2314 name="experimental_distribute",
2315 ty=distribute_options.DistributeOptions,
2316 docstring=
2317 "The distribution strategy options associated with the dataset. See "
2318 "`tf.data.experimental.DistributeOptions` for more details.",
2319 default_factory=distribute_options.DistributeOptions)
2320
2321 experimental_optimization = options_lib.create_option(
2322 name="experimental_optimization",
2323 ty=optimization_options.OptimizationOptions,
2324 docstring=
2325 "The optimization options associated with the dataset. See "
2326 "`tf.data.experimental.OptimizationOptions` for more details.",
2327 default_factory=optimization_options.OptimizationOptions)
2328
2329 experimental_slack = options_lib.create_option(
2330 name="experimental_slack",
2331 ty=bool,
2332 docstring="Whether to introduce 'slack' in the last `prefetch` of the "
2333 "input pipeline, if it exists. This may reduce CPU contention with "
2334 "accelerator host-side activity at the start of a step. The slack "
2335 "frequency is determined by the number of devices attached to this "
2336 "input pipeline. If None, defaults to False.")
2337
2338 experimental_stats = options_lib.create_option(
2339 name="experimental_stats",
2340 ty=stats_options.StatsOptions,
2341 docstring=
2342 "The statistics options associated with the dataset. See "
2343 "`tf.data.experimental.StatsOptions` for more details.",
2344 default_factory=stats_options.StatsOptions)
2345
2346 experimental_threading = options_lib.create_option(
2347 name="experimental_threading",
2348 ty=threading_options.ThreadingOptions,
2349 docstring=
2350 "The threading options associated with the dataset. See "
2351 "`tf.data.experimental.ThreadingOptions` for more details.",
2352 default_factory=threading_options.ThreadingOptions)
2353
2354 experimental_stateful_whitelist = options_lib.create_option(

Callers 1

optionsMethod · 0.70

Calls

no outgoing calls

Tested by

no test coverage detected