Learn (contrib)
[TOC]
High level API for learning with TensorFlow.
Estimators
Train and evaluate TensorFlow models.
class tf.contrib.learn.BaseEstimator
Abstract BaseEstimator class to train and evaluate TensorFlow models.
Concrete implementation of this class should provide the following functions:
- _get_train_ops
- _get_eval_ops
- _get_predict_ops
Estimator
implemented below is a good example of how to use this class.
tf.contrib.learn.BaseEstimator.__init__(model_dir=None, config=None)
Initializes a BaseEstimator instance.
Args:
model_dir
: Directory to save model parameters, graph and etc. This can also be used to load checkpoints from the directory into a estimator to continue training a previously saved model.config
: A RunConfig instance.
tf.contrib.learn.BaseEstimator.__repr__()
tf.contrib.learn.BaseEstimator.config
tf.contrib.learn.BaseEstimator.evaluate(x=None, y=None, input_fn=None, feed_fn=None, batch_size=None, steps=None, metrics=None, name=None)
See Evaluable
.
Raises:
ValueError
: If at least one ofx
ory
is provided, and at least one ofinput_fn
orfeed_fn
is provided. Or ifmetrics
is notNone
ordict
.
tf.contrib.learn.BaseEstimator.export(*args, **kwargs)
Exports inference graph into given dir. (deprecated arguments)
SOME ARGUMENTS ARE DEPRECATED. They will be removed after 2016-09-23. Instructions for updating: The signature of the input_fn accepted by export is changing to be consistent with what's used by tf.Learn Estimator's train/evaluate. input_fn (and in most cases, input_feature_key) will become required args, and use_deprecated_input_fn will default to False and be removed altogether.
Args:
export_dir: A string containing a directory to write the exported graph
and checkpoints.
input_fn: If `use_deprecated_input_fn` is true, then a function that given
`Tensor` of `Example` strings, parses it into features that are then
passed to the model. Otherwise, a function that takes no argument and
returns a tuple of (features, targets), where features is a dict of
string key to `Tensor` and targets is a `Tensor` that's currently not
used (and so can be `None`).
input_feature_key: Only used if `use_deprecated_input_fn` is false. String
key into the features dict returned by `input_fn` that corresponds toa
the raw `Example` strings `Tensor` that the exported model will take as
input. Can only be `None` if you're using a custom `signature_fn` that
does not use the first arg (examples).
use_deprecated_input_fn: Determines the signature format of `input_fn`.
signature_fn: Function that returns a default signature and a named
signature map, given `Tensor` of `Example` strings, `dict` of `Tensor`s
for features and `Tensor` or `dict` of `Tensor`s for predictions.
prediction_key: The key for a tensor in the `predictions` dict (output
from the `model_fn`) to use as the `predictions` input to the
`signature_fn`. Optional. If `None`, predictions will pass to
`signature_fn` without filtering.
default_batch_size: Default batch size of the `Example` placeholder.
exports_to_keep: Number of exports to keep.
Returns:
The string path to the exported directory. NB: this functionality was
added ca. 2016/09/25; clients that depend on the return value may need
to handle the case where this function returns None because subclasses
are not returning a value.
tf.contrib.learn.BaseEstimator.fit(x=None, y=None, input_fn=None, steps=None, batch_size=None, monitors=None, max_steps=None)
See Trainable
.
Raises:
ValueError
: Ifx
ory
are notNone
whileinput_fn
is notNone
.ValueError
: If bothsteps
andmax_steps
are notNone
.
tf.contrib.learn.BaseEstimator.get_params(deep=True)
Get parameters for this estimator.
Args:
deep
: boolean, optionalIf
True
, will return the parameters for this estimator and contained subobjects that are estimators.
Returns:
params : mapping of string to any Parameter names mapped to their values.
tf.contrib.learn.BaseEstimator.get_variable_names()
Returns list of all variable names in this model.
Returns:
List of names.
tf.contrib.learn.BaseEstimator.get_variable_value(name)
Returns value of the variable given by name.
Args:
name
: string, name of the tensor.
Returns:
Numpy array - value of the tensor.
tf.contrib.learn.BaseEstimator.model_dir
tf.contrib.learn.BaseEstimator.partial_fit(x=None, y=None, input_fn=None, steps=1, batch_size=None, monitors=None)
Incremental fit on a batch of samples.
This method is expected to be called several times consecutively on different or the same chunks of the dataset. This either can implement iterative training or out-of-core/online training.
This is especially useful when the whole dataset is too big to fit in memory at the same time. Or when model is taking long time to converge, and you want to split up training into subparts.
Args:
x
: Matrix of shape [n_samples, n_features...]. Can be iterator that returns arrays of features. The training input samples for fitting the model. If set,input_fn
must beNone
.y
: Vector or matrix [n_samples] or [n_samples, n_outputs]. Can be iterator that returns array of targets. The training target values (class labels in classification, real numbers in regression). If set,input_fn
must beNone
.input_fn
: Input function. If set,x
,y
, andbatch_size
must beNone
.steps
: Number of steps for which to train model. IfNone
, train forever.batch_size
: minibatch size to use on the input, defaults to first dimension ofx
. Must beNone
ifinput_fn
is provided.monitors
: List ofBaseMonitor
subclass instances. Used for callbacks inside the training loop.
Returns:
self
, for chaining.
Raises:
ValueError
: If at least one ofx
andy
is provided, andinput_fn
is provided.
tf.contrib.learn.BaseEstimator.predict(*args, **kwargs)
Returns predictions for given features. (deprecated arguments)
SOME ARGUMENTS ARE DEPRECATED. They will be removed after 2016-09-15. Instructions for updating: The default behavior of predict() is changing. The default value for as_iterable will change to True, and then the flag will be removed altogether. The behavior of this flag is described below.
Args:
x: Matrix of shape [n_samples, n_features...]. Can be iterator that
returns arrays of features. The training input samples for fitting the
model. If set, `input_fn` must be `None`.
input_fn: Input function. If set, `x` and 'batch_size' must be `None`.
batch_size: Override default batch size. If set, 'input_fn' must be
'None'.
outputs: list of `str`, name of the output to predict.
If `None`, returns all.
as_iterable: If True, return an iterable which keeps yielding predictions
for each example until inputs are exhausted. Note: The inputs must
terminate if you want the iterable to terminate (e.g. be sure to pass
num_epochs=1 if you are using something like read_batch_features).
Returns:
A numpy array of predicted classes or regression values if the
constructor's `model_fn` returns a `Tensor` for `predictions` or a `dict`
of numpy arrays if `model_fn` returns a `dict`. Returns an iterable of
predictions if as_iterable is True.
Raises:
ValueError: If x and input_fn are both provided or both `None`.
tf.contrib.learn.BaseEstimator.set_params(**params)
Set the parameters of this estimator.
The method works on simple estimators as well as on nested objects
(such as pipelines). The former have parameters of the form
<component>__<parameter>
so that it's possible to update each
component of a nested object.
Args:
**params
: Parameters.
Returns:
self
Raises:
ValueError
: If params contain invalid names.
class tf.contrib.learn.Estimator
Estimator class is the basic TensorFlow model trainer/evaluator.
tf.contrib.learn.Estimator.__init__(model_fn=None, model_dir=None, config=None, params=None, feature_engineering_fn=None)
Constructs an Estimator instance.
Args:
model_fn
: Model function, takes features and targets tensors or dicts oftensors and returns predictions and loss tensors. Supports next three signatures for the function:
(features, targets) -> (predictions, loss, train_op)
(features, targets, mode) -> (predictions, loss, train_op)
(features, targets, mode, params) -> (predictions, loss, train_op)
Where
features
are singleTensor
ordict
ofTensor
s(depending on data passed to `fit`),
targets
areTensor
ordict
ofTensor
s (for multi-headmodels). If mode is `ModeKeys.INFER`, `targets=None` will be passed. If the `model_fn`'s signature does not accept `mode`, the `model_fn` must still be able to handle `targets=None`.
mode
represents if this training, evaluation orprediction. See `ModeKeys`.
params
is adict
of hyperparameters. Will receive whatis passed to Estimator in `params` parameter. This allows to configure Estimators from hyper parameter tunning.
model_dir
: Directory to save model parameters, graph and etc. This can also be used to load checkpoints from the directory into a estimator to continue training a previously saved model.config
: Configuration object.params
:dict
of hyper parameters that will be passed intomodel_fn
.Keys are names of parameters, values are basic python types.
feature_engineering_fn
: Feature engineering function. Takes features andtargets which are the output of `input_fn` and returns features and targets which will be fed into `model_fn`. Please check `model_fn` for a definition of features and targets.
Raises:
ValueError
: parameters ofmodel_fn
don't matchparams
.
tf.contrib.learn.Estimator.__repr__()
tf.contrib.learn.Estimator.config
tf.contrib.learn.Estimator.evaluate(x=None, y=None, input_fn=None, feed_fn=None, batch_size=None, steps=None, metrics=None, name=None)
See Evaluable
.
Raises:
ValueError
: If at least one ofx
ory
is provided, and at least one ofinput_fn
orfeed_fn
is provided. Or ifmetrics
is notNone
ordict
.
tf.contrib.learn.Estimator.export(*args, **kwargs)
Exports inference graph into given dir. (deprecated arguments)
SOME ARGUMENTS ARE DEPRECATED. They will be removed after 2016-09-23. Instructions for updating: The signature of the input_fn accepted by export is changing to be consistent with what's used by tf.Learn Estimator's train/evaluate. input_fn (and in most cases, input_feature_key) will become required args, and use_deprecated_input_fn will default to False and be removed altogether.
Args:
export_dir: A string containing a directory to write the exported graph
and checkpoints.
input_fn: If `use_deprecated_input_fn` is true, then a function that given
`Tensor` of `Example` strings, parses it into features that are then
passed to the model. Otherwise, a function that takes no argument and
returns a tuple of (features, targets), where features is a dict of
string key to `Tensor` and targets is a `Tensor` that's currently not
used (and so can be `None`).
input_feature_key: Only used if `use_deprecated_input_fn` is false. String
key into the features dict returned by `input_fn` that corresponds toa
the raw `Example` strings `Tensor` that the exported model will take as
input. Can only be `None` if you're using a custom `signature_fn` that
does not use the first arg (examples).
use_deprecated_input_fn: Determines the signature format of `input_fn`.
signature_fn: Function that returns a default signature and a named
signature map, given `Tensor` of `Example` strings, `dict` of `Tensor`s
for features and `Tensor` or `dict` of `Tensor`s for predictions.
prediction_key: The key for a tensor in the `predictions` dict (output
from the `model_fn`) to use as the `predictions` input to the
`signature_fn`. Optional. If `None`, predictions will pass to
`signature_fn` without filtering.
default_batch_size: Default batch size of the `Example` placeholder.
exports_to_keep: Number of exports to keep.
Returns:
The string path to the exported directory. NB: this functionality was
added ca. 2016/09/25; clients that depend on the return value may need
to handle the case where this function returns None because subclasses
are not returning a value.
tf.contrib.learn.Estimator.fit(x=None, y=None, input_fn=None, steps=None, batch_size=None, monitors=None, max_steps=None)
See Trainable
.
Raises:
ValueError
: Ifx
ory
are notNone
whileinput_fn
is notNone
.ValueError
: If bothsteps
andmax_steps
are notNone
.
tf.contrib.learn.Estimator.get_params(deep=True)
Get parameters for this estimator.
Args:
deep
: boolean, optionalIf
True
, will return the parameters for this estimator and contained subobjects that are estimators.
Returns:
params : mapping of string to any Parameter names mapped to their values.
tf.contrib.learn.Estimator.get_variable_names()
Returns list of all variable names in this model.
Returns:
List of names.
tf.contrib.learn.Estimator.get_variable_value(name)
Returns value of the variable given by name.
Args:
name
: string, name of the tensor.
Returns:
Numpy array - value of the tensor.
tf.contrib.learn.Estimator.model_dir
tf.contrib.learn.Estimator.partial_fit(x=None, y=None, input_fn=None, steps=1, batch_size=None, monitors=None)
Incremental fit on a batch of samples.
This method is expected to be called several times consecutively on different or the same chunks of the dataset. This either can implement iterative training or out-of-core/online training.
This is especially useful when the whole dataset is too big to fit in memory at the same time. Or when model is taking long time to converge, and you want to split up training into subparts.
Args:
x
: Matrix of shape [n_samples, n_features...]. Can be iterator that returns arrays of features. The training input samples for fitting the model. If set,input_fn
must beNone
.y
: Vector or matrix [n_samples] or [n_samples, n_outputs]. Can be iterator that returns array of targets. The training target values (class labels in classification, real numbers in regression). If set,input_fn
must beNone
.input_fn
: Input function. If set,x
,y
, andbatch_size
must beNone
.steps
: Number of steps for which to train model. IfNone
, train forever.batch_size
: minibatch size to use on the input, defaults to first dimension ofx
. Must beNone
ifinput_fn
is provided.monitors
: List ofBaseMonitor
subclass instances. Used for callbacks inside the training loop.
Returns:
self
, for chaining.
Raises:
ValueError
: If at least one ofx
andy
is provided, andinput_fn
is provided.
tf.contrib.learn.Estimator.predict(*args, **kwargs)
Returns predictions for given features. (deprecated arguments)
SOME ARGUMENTS ARE DEPRECATED. They will be removed after 2016-09-15. Instructions for updating: The default behavior of predict() is changing. The default value for as_iterable will change to True, and then the flag will be removed altogether. The behavior of this flag is described below.
Args:
x: Matrix of shape [n_samples, n_features...]. Can be iterator that
returns arrays of features. The training input samples for fitting the
model. If set, `input_fn` must be `None`.
input_fn: Input function. If set, `x` and 'batch_size' must be `None`.
batch_size: Override default batch size. If set, 'input_fn' must be
'None'.
outputs: list of `str`, name of the output to predict.
If `None`, returns all.
as_iterable: If True, return an iterable which keeps yielding predictions
for each example until inputs are exhausted. Note: The inputs must
terminate if you want the iterable to terminate (e.g. be sure to pass
num_epochs=1 if you are using something like read_batch_features).
Returns:
A numpy array of predicted classes or regression values if the
constructor's `model_fn` returns a `Tensor` for `predictions` or a `dict`
of numpy arrays if `model_fn` returns a `dict`. Returns an iterable of
predictions if as_iterable is True.
Raises:
ValueError: If x and input_fn are both provided or both `None`.
tf.contrib.learn.Estimator.set_params(**params)
Set the parameters of this estimator.
The method works on simple estimators as well as on nested objects
(such as pipelines). The former have parameters of the form
<component>__<parameter>
so that it's possible to update each
component of a nested object.
Args:
**params
: Parameters.
Returns:
self
Raises:
ValueError
: If params contain invalid names.
class tf.contrib.learn.ModeKeys
Standard names for model modes.
The following standard keys are defined:
TRAIN
: training mode.EVAL
: evaluation mode.INFER
: inference mode.
class tf.contrib.learn.DNNClassifier
A classifier for TensorFlow DNN models.
Example:
education = sparse_column_with_hash_bucket(column_name="education",
hash_bucket_size=1000)
occupation = sparse_column_with_hash_bucket(column_name="occupation",
hash_bucket_size=1000)
education_emb = embedding_column(sparse_id_column=education, dimension=16,
combiner="sum")
occupation_emb = embedding_column(sparse_id_column=occupation, dimension=16,
combiner="sum")
estimator = DNNClassifier(
feature_columns=[education_emb, occupation_emb],
hidden_units=[1024, 512, 256])
# Or estimator using the ProximalAdagradOptimizer optimizer with
# regularization.
estimator = DNNClassifier(
feature_columns=[education_emb, occupation_emb],
hidden_units=[1024, 512, 256],
optimizer=tf.train.ProximalAdagradOptimizer(
learning_rate=0.1,
l1_regularization_strength=0.001
))
# Input builders
def input_fn_train: # returns x, Y
pass
estimator.fit(input_fn=input_fn_train)
def input_fn_eval: # returns x, Y
pass
estimator.evaluate(input_fn=input_fn_eval)
estimator.predict(x=x)
Input of fit
and evaluate
should have following features,
otherwise there will be a KeyError
:
- if
weight_column_name
is notNone
, a feature withkey=weight_column_name
whose value is aTensor
. - for each
column
infeature_columns
:- if
column
is aSparseColumn
, a feature withkey=column.name
whosevalue
is aSparseTensor
. - if
column
is aWeightedSparseColumn
, two features: the first withkey
the id column name, the second withkey
the weight column name. Both features'value
must be aSparseTensor
. - if
column
is aRealValuedColumn
, a feature withkey=column.name
whosevalue
is aTensor
.
- if
tf.contrib.learn.DNNClassifier.__init__(hidden_units, feature_columns, model_dir=None, n_classes=2, weight_column_name=None, optimizer=None, activation_fn=relu, dropout=None, gradient_clip_norm=None, enable_centered_bias=None, config=None, feature_engineering_fn=None)
Initializes a DNNClassifier instance.
Args:
hidden_units
: List of hidden units per layer. All layers are fully connected. Ex.[64, 32]
means first layer has 64 nodes and second one has 32.feature_columns
: An iterable containing all the feature columns used by the model. All items in the set should be instances of classes derived fromFeatureColumn
.model_dir
: Directory to save model parameters, graph and etc. This can also be used to load checkpoints from the directory into a estimator to continue training a previously saved model.n_classes
: number of target classes. Default is binary classification. It must be greater than 1.weight_column_name
: A string defining feature column name representing weights. It is used to down weight or boost examples during training. It will be multiplied by the loss of the example.optimizer
: An instance oftf.Optimizer
used to train the model. IfNone
, will use an Adagrad optimizer.activation_fn
: Activation function applied to each layer. IfNone
, will usetf.nn.relu
.dropout
: When notNone
, the probability we will drop out a given coordinate.gradient_clip_norm
: A float > 0. If provided, gradients are clipped to their global norm with this clipping ratio. Seetf.clip_by_global_norm
for more details.enable_centered_bias
: A bool. If True, estimator will learn a centered bias variable for each class. Rest of the model structure learns the residual after centered bias.config
:RunConfig
object to configure the runtime settings.feature_engineering_fn
: Feature engineering function. Takes features andtargets which are the output of `input_fn` and returns features and targets which will be fed into the model.
Returns:
A DNNClassifier
estimator.
Raises:
ValueError
: Ifn_classes
< 2.
tf.contrib.learn.DNNClassifier.bias_
DEPRECATED FUNCTION
THIS FUNCTION IS DEPRECATED. It will be removed after 2016-10-30. Instructions for updating: This method will be removed after the deprecation date. To inspect variables, use get_variable_names() and get_variable_value().
tf.contrib.learn.DNNClassifier.config
tf.contrib.learn.DNNClassifier.evaluate(x=None, y=None, input_fn=None, feed_fn=None, batch_size=None, steps=None, metrics=None, name=None)
See evaluable.Evaluable.
tf.contrib.learn.DNNClassifier.export(export_dir, input_fn=None, input_feature_key=None, use_deprecated_input_fn=True, signature_fn=None, default_batch_size=1, exports_to_keep=None)
See BaseEstimator.export.
tf.contrib.learn.DNNClassifier.fit(x=None, y=None, input_fn=None, steps=None, batch_size=None, monitors=None, max_steps=None)
See trainable.Trainable.
tf.contrib.learn.DNNClassifier.get_variable_names()
Returns list of all variable names in this model.
Returns:
List of names.
tf.contrib.learn.DNNClassifier.get_variable_value(name)
Returns value of the variable given by name.
Args:
name
: string, name of the tensor.
Returns:
Tensor
object.
tf.contrib.learn.DNNClassifier.model_dir
tf.contrib.learn.DNNClassifier.predict(*args, **kwargs)
Returns predicted classes for given features. (deprecated arguments)
SOME ARGUMENTS ARE DEPRECATED. They will be removed after 2016-09-15. Instructions for updating: The default behavior of predict() is changing. The default value for as_iterable will change to True, and then the flag will be removed altogether. The behavior of this flag is described below.
Args:
x: features.
input_fn: Input function. If set, x must be None.
batch_size: Override default batch size.
as_iterable: If True, return an iterable which keeps yielding predictions
for each example until inputs are exhausted. Note: The inputs must
terminate if you want the iterable to terminate (e.g. be sure to pass
num_epochs=1 if you are using something like read_batch_features).
Returns:
Numpy array of predicted classes (or an iterable of predicted classes if
as_iterable is True).
tf.contrib.learn.DNNClassifier.predict_proba(*args, **kwargs)
Returns prediction probabilities for given features. (deprecated arguments)
SOME ARGUMENTS ARE DEPRECATED. They will be removed after 2016-09-15. Instructions for updating: The default behavior of predict() is changing. The default value for as_iterable will change to True, and then the flag will be removed altogether. The behavior of this flag is described below.
Args:
x: features.
input_fn: Input function. If set, x and y must be None.
batch_size: Override default batch size.
as_iterable: If True, return an iterable which keeps yielding predictions
for each example until inputs are exhausted. Note: The inputs must
terminate if you want the iterable to terminate (e.g. be sure to pass
num_epochs=1 if you are using something like read_batch_features).
Returns:
Numpy array of predicted probabilities (or an iterable of predicted
probabilities if as_iterable is True).
tf.contrib.learn.DNNClassifier.weights_
DEPRECATED FUNCTION
THIS FUNCTION IS DEPRECATED. It will be removed after 2016-10-30. Instructions for updating: This method will be removed after the deprecation date. To inspect variables, use get_variable_names() and get_variable_value().
class tf.contrib.learn.DNNRegressor
A regressor for TensorFlow DNN models.
Example:
education = sparse_column_with_hash_bucket(column_name="education",
hash_bucket_size=1000)
occupation = sparse_column_with_hash_bucket(column_name="occupation",
hash_bucket_size=1000)
education_emb = embedding_column(sparse_id_column=education, dimension=16,
combiner="sum")
occupation_emb = embedding_column(sparse_id_column=occupation, dimension=16,
combiner="sum")
estimator = DNNRegressor(
feature_columns=[education_emb, occupation_emb],
hidden_units=[1024, 512, 256])
# Or estimator using the ProximalAdagradOptimizer optimizer with
# regularization.
estimator = DNNRegressor(
feature_columns=[education_emb, occupation_emb],
hidden_units=[1024, 512, 256],
optimizer=tf.train.ProximalAdagradOptimizer(
learning_rate=0.1,
l1_regularization_strength=0.001
))
# Input builders
def input_fn_train: # returns x, Y
pass
estimator.fit(input_fn=input_fn_train)
def input_fn_eval: # returns x, Y
pass
estimator.evaluate(input_fn=input_fn_eval)
estimator.predict(x=x)
Input of fit
and evaluate
should have following features,
otherwise there will be a KeyError
:
- if
weight_column_name
is notNone
, a feature withkey=weight_column_name
whose value is aTensor
. - for each
column
infeature_columns
:- if
column
is aSparseColumn
, a feature withkey=column.name
whosevalue
is aSparseTensor
. - if
column
is aWeightedSparseColumn
, two features: the first withkey
the id column name, the second withkey
the weight column name. Both features'value
must be aSparseTensor
. - if
column
is aRealValuedColumn
, a feature withkey=column.name
whosevalue
is aTensor
.
- if
tf.contrib.learn.DNNRegressor.__init__(hidden_units, feature_columns, model_dir=None, weight_column_name=None, optimizer=None, activation_fn=relu, dropout=None, gradient_clip_norm=None, enable_centered_bias=None, config=None, feature_engineering_fn=None)
Initializes a DNNRegressor
instance.
Args:
hidden_units
: List of hidden units per layer. All layers are fully connected. Ex.[64, 32]
means first layer has 64 nodes and second one has 32.feature_columns
: An iterable containing all the feature columns used by the model. All items in the set should be instances of classes derived fromFeatureColumn
.model_dir
: Directory to save model parameters, graph and etc. This can also be used to load checkpoints from the directory into a estimator to continue training a previously saved model.weight_column_name
: A string defining feature column name representing weights. It is used to down weight or boost examples during training. It will be multiplied by the loss of the example.optimizer
: An instance oftf.Optimizer
used to train the model. IfNone
, will use an Adagrad optimizer.activation_fn
: Activation function applied to each layer. IfNone
, will usetf.nn.relu
.dropout
: When notNone
, the probability we will drop out a given coordinate.gradient_clip_norm
: Afloat
> 0. If provided, gradients are clipped to their global norm with this clipping ratio. Seetf.clip_by_global_norm
for more details.enable_centered_bias
: A bool. If True, estimator will learn a centered bias variable for each class. Rest of the model structure learns the residual after centered bias.config
:RunConfig
object to configure the runtime settings.feature_engineering_fn
: Feature engineering function. Takes features andtargets which are the output of `input_fn` and returns features and targets which will be fed into the model.
Returns:
A DNNRegressor
estimator.
tf.contrib.learn.DNNRegressor.__repr__()
tf.contrib.learn.DNNRegressor.bias_
DEPRECATED FUNCTION
THIS FUNCTION IS DEPRECATED. It will be removed after 2016-10-30. Instructions for updating: This method will be removed after the deprecation date. To inspect variables, use get_variable_names() and get_variable_value().
tf.contrib.learn.DNNRegressor.config
tf.contrib.learn.DNNRegressor.dnn_bias_
Returns bias of deep neural network part. (deprecated)
THIS FUNCTION IS DEPRECATED. It will be removed after 2016-10-30. Instructions for updating: This method will be removed after the deprecation date. To inspect variables, use get_variable_names() and get_variable_value().
tf.contrib.learn.DNNRegressor.dnn_weights_
Returns weights of deep neural network part. (deprecated)
THIS FUNCTION IS DEPRECATED. It will be removed after 2016-10-30. Instructions for updating: This method will be removed after the deprecation date. To inspect variables, use get_variable_names() and get_variable_value().
tf.contrib.learn.DNNRegressor.evaluate(x=None, y=None, input_fn=None, feed_fn=None, batch_size=None, steps=None, metrics=None, name=None)
See Evaluable
.
Raises:
ValueError
: If at least one ofx
ory
is provided, and at least one ofinput_fn
orfeed_fn
is provided. Or ifmetrics
is notNone
ordict
.
tf.contrib.learn.DNNRegressor.export(*args, **kwargs)
Exports inference graph into given dir. (deprecated arguments)
SOME ARGUMENTS ARE DEPRECATED. They will be removed after 2016-09-23. Instructions for updating: The signature of the input_fn accepted by export is changing to be consistent with what's used by tf.Learn Estimator's train/evaluate. input_fn (and in most cases, input_feature_key) will become required args, and use_deprecated_input_fn will default to False and be removed altogether.
Args:
export_dir: A string containing a directory to write the exported graph
and checkpoints.
input_fn: If `use_deprecated_input_fn` is true, then a function that given
`Tensor` of `Example` strings, parses it into features that are then
passed to the model. Otherwise, a function that takes no argument and
returns a tuple of (features, targets), where features is a dict of
string key to `Tensor` and targets is a `Tensor` that's currently not
used (and so can be `None`).
input_feature_key: Only used if `use_deprecated_input_fn` is false. String
key into the features dict returned by `input_fn` that corresponds toa
the raw `Example` strings `Tensor` that the exported model will take as
input. Can only be `None` if you're using a custom `signature_fn` that
does not use the first arg (examples).
use_deprecated_input_fn: Determines the signature format of `input_fn`.
signature_fn: Function that returns a default signature and a named
signature map, given `Tensor` of `Example` strings, `dict` of `Tensor`s
for features and `Tensor` or `dict` of `Tensor`s for predictions.
prediction_key: The key for a tensor in the `predictions` dict (output
from the `model_fn`) to use as the `predictions` input to the
`signature_fn`. Optional. If `None`, predictions will pass to
`signature_fn` without filtering.
default_batch_size: Default batch size of the `Example` placeholder.
exports_to_keep: Number of exports to keep.
Returns:
The string path to the exported directory. NB: this functionality was
added ca. 2016/09/25; clients that depend on the return value may need
to handle the case where this function returns None because subclasses
are not returning a value.
tf.contrib.learn.DNNRegressor.fit(x=None, y=None, input_fn=None, steps=None, batch_size=None, monitors=None, max_steps=None)
See Trainable
.
Raises:
ValueError
: Ifx
ory
are notNone
whileinput_fn
is notNone
.ValueError
: If bothsteps
andmax_steps
are notNone
.
tf.contrib.learn.DNNRegressor.get_params(deep=True)
Get parameters for this estimator.
Args:
deep
: boolean, optionalIf
True
, will return the parameters for this estimator and contained subobjects that are estimators.
Returns:
params : mapping of string to any Parameter names mapped to their values.
tf.contrib.learn.DNNRegressor.get_variable_names()
Returns list of all variable names in this model.
Returns:
List of names.
tf.contrib.learn.DNNRegressor.get_variable_value(name)
Returns value of the variable given by name.
Args:
name
: string, name of the tensor.
Returns:
Numpy array - value of the tensor.
tf.contrib.learn.DNNRegressor.linear_bias_
Returns bias of the linear part. (deprecated)
THIS FUNCTION IS DEPRECATED. It will be removed after 2016-10-30. Instructions for updating: This method will be removed after the deprecation date. To inspect variables, use get_variable_names() and get_variable_value().
tf.contrib.learn.DNNRegressor.linear_weights_
Returns weights per feature of the linear part. (deprecated)
THIS FUNCTION IS DEPRECATED. It will be removed after 2016-10-30. Instructions for updating: This method will be removed after the deprecation date. To inspect variables, use get_variable_names() and get_variable_value().
tf.contrib.learn.DNNRegressor.model_dir
tf.contrib.learn.DNNRegressor.partial_fit(x=None, y=None, input_fn=None, steps=1, batch_size=None, monitors=None)
Incremental fit on a batch of samples.
This method is expected to be called several times consecutively on different or the same chunks of the dataset. This either can implement iterative training or out-of-core/online training.
This is especially useful when the whole dataset is too big to fit in memory at the same time. Or when model is taking long time to converge, and you want to split up training into subparts.
Args:
x
: Matrix of shape [n_samples, n_features...]. Can be iterator that returns arrays of features. The training input samples for fitting the model. If set,input_fn
must beNone
.y
: Vector or matrix [n_samples] or [n_samples, n_outputs]. Can be iterator that returns array of targets. The training target values (class labels in classification, real numbers in regression). If set,input_fn
must beNone
.input_fn
: Input function. If set,x
,y
, andbatch_size
must beNone
.steps
: Number of steps for which to train model. IfNone
, train forever.batch_size
: minibatch size to use on the input, defaults to first dimension ofx
. Must beNone
ifinput_fn
is provided.monitors
: List ofBaseMonitor
subclass instances. Used for callbacks inside the training loop.
Returns:
self
, for chaining.
Raises:
ValueError
: If at least one ofx
andy
is provided, andinput_fn
is provided.
tf.contrib.learn.DNNRegressor.predict(*args, **kwargs)
Returns predictions for given features. (deprecated arguments)
SOME ARGUMENTS ARE DEPRECATED. They will be removed after 2016-09-15. Instructions for updating: The default behavior of predict() is changing. The default value for as_iterable will change to True, and then the flag will be removed altogether. The behavior of this flag is described below.
Args:
x: Matrix of shape [n_samples, n_features...]. Can be iterator that
returns arrays of features. The training input samples for fitting the
model. If set, `input_fn` must be `None`.
input_fn: Input function. If set, `x` and 'batch_size' must be `None`.
batch_size: Override default batch size. If set, 'input_fn' must be
'None'.
outputs: list of `str`, name of the output to predict.
If `None`, returns all.
as_iterable: If True, return an iterable which keeps yielding predictions
for each example until inputs are exhausted. Note: The inputs must
terminate if you want the iterable to terminate (e.g. be sure to pass
num_epochs=1 if you are using something like read_batch_features).
Returns:
A numpy array of predicted classes or regression values if the
constructor's `model_fn` returns a `Tensor` for `predictions` or a `dict`
of numpy arrays if `model_fn` returns a `dict`. Returns an iterable of
predictions if as_iterable is True.
Raises:
ValueError: If x and input_fn are both provided or both `None`.
tf.contrib.learn.DNNRegressor.set_params(**params)
Set the parameters of this estimator.
The method works on simple estimators as well as on nested objects
(such as pipelines). The former have parameters of the form
<component>__<parameter>
so that it's possible to update each
component of a nested object.
Args:
**params
: Parameters.
Returns:
self
Raises:
ValueError
: If params contain invalid names.
tf.contrib.learn.DNNRegressor.weights_
DEPRECATED FUNCTION
THIS FUNCTION IS DEPRECATED. It will be removed after 2016-10-30. Instructions for updating: This method will be removed after the deprecation date. To inspect variables, use get_variable_names() and get_variable_value().
class tf.contrib.learn.TensorFlowEstimator
Base class for all TensorFlow estimators.
tf.contrib.learn.TensorFlowEstimator.__init__(model_fn, n_classes, batch_size=32, steps=200, optimizer='Adagrad', learning_rate=0.1, clip_gradients=5.0, class_weight=None, continue_training=False, config=None, verbose=1)
Initializes a TensorFlowEstimator instance.
Args:
model_fn
: Model function, that takes inputx
,y
tensors and outputs prediction and loss tensors.n_classes
: Number of classes in the target.batch_size
: Mini batch size.steps
: Number of steps to run over data.optimizer
: Optimizer name (or class), for example "SGD", "Adam", "Adagrad".learning_rate
: If this is constant float value, no decay function is used. Instead, a customized decay function can be passed that accepts global_step as parameter and returns a Tensor. e.g. exponential decay function:def exp_decay(global_step): return tf.train.exponential_decay( learning_rate=0.1, global_step, decay_steps=2, decay_rate=0.001)
clip_gradients
: Clip norm of the gradients to this value to stop gradient explosion.class_weight
: None or list of n_classes floats. Weight associated with classes for loss computation. If not given, all classes are supposed to have weight one.continue_training
: when continue_training is True, once initialized model will be continually trained on every call of fit.config
: RunConfig object that controls the configurations of the session, e.g. num_cores, gpu_memory_fraction, etc.verbose
: Controls the verbosity, possible values:- 0: the algorithm and debug information is muted.
- 1: trainer prints the progress.
- 2: log device placement is printed.
tf.contrib.learn.TensorFlowEstimator.__repr__()
tf.contrib.learn.TensorFlowEstimator.config
tf.contrib.learn.TensorFlowEstimator.evaluate(x=None, y=None, input_fn=None, feed_fn=None, batch_size=None, steps=None, metrics=None, name=None)
Evaluates given model with provided evaluation data.
See superclass Estimator for more details.
Args:
x
: features.y
: targets.input_fn
: Input function.feed_fn
: Function creating a feed dict every time it is called.batch_size
: minibatch size to use on the input.steps
: Number of steps for which to evaluate model.metrics
: Dict of metric ops to run. If None, the default metrics are used.name
: Name of the evaluation.
Returns:
Returns dict
with evaluation results.
tf.contrib.learn.TensorFlowEstimator.export(*args, **kwargs)
Exports inference graph into given dir. (deprecated arguments)
SOME ARGUMENTS ARE DEPRECATED. They will be removed after 2016-09-23. Instructions for updating: The signature of the input_fn accepted by export is changing to be consistent with what's used by tf.Learn Estimator's train/evaluate. input_fn (and in most cases, input_feature_key) will become required args, and use_deprecated_input_fn will default to False and be removed altogether.
Args:
export_dir: A string containing a directory to write the exported graph
and checkpoints.
input_fn: If `use_deprecated_input_fn` is true, then a function that given
`Tensor` of `Example` strings, parses it into features that are then
passed to the model. Otherwise, a function that takes no argument and
returns a tuple of (features, targets), where features is a dict of
string key to `Tensor` and targets is a `Tensor` that's currently not
used (and so can be `None`).
input_feature_key: Only used if `use_deprecated_input_fn` is false. String
key into the features dict returned by `input_fn` that corresponds toa
the raw `Example` strings `Tensor` that the exported model will take as
input. Can only be `None` if you're using a custom `signature_fn` that
does not use the first arg (examples).
use_deprecated_input_fn: Determines the signature format of `input_fn`.
signature_fn: Function that returns a default signature and a named
signature map, given `Tensor` of `Example` strings, `dict` of `Tensor`s
for features and `Tensor` or `dict` of `Tensor`s for predictions.
prediction_key: The key for a tensor in the `predictions` dict (output
from the `model_fn`) to use as the `predictions` input to the
`signature_fn`. Optional. If `None`, predictions will pass to
`signature_fn` without filtering.
default_batch_size: Default batch size of the `Example` placeholder.
exports_to_keep: Number of exports to keep.
Returns:
The string path to the exported directory. NB: this functionality was
added ca. 2016/09/25; clients that depend on the return value may need
to handle the case where this function returns None because subclasses
are not returning a value.
tf.contrib.learn.TensorFlowEstimator.fit(x, y, steps=None, monitors=None, logdir=None)
Neural network model from provided model_fn
and training data.
Note: called first time constructs the graph and initializers variables. Subsequently, it will continue training the same model. This logic follows partial_fit() interface in scikit-learn. To restart learning, create new estimator.
Args:
x
: matrix or tensor of shape [n_samples, n_features...]. Can be iterator that returns arrays of features. The training input samples for fitting the model.y
: vector or matrix [n_samples] or [n_samples, n_outputs]. Can be iterator that returns array of targets. The training target values (class labels in classification, real numbers in regression).steps
: int, number of steps to train.If None or 0, train for `self.steps`.
monitors
: List ofBaseMonitor
objects to print training progress and invoke early stopping.logdir
: the directory to save the log file that can be used for optional visualization.
Returns:
Returns self.
tf.contrib.learn.TensorFlowEstimator.get_params(deep=True)
Get parameters for this estimator.
Args:
deep
: boolean, optionalIf
True
, will return the parameters for this estimator and contained subobjects that are estimators.
Returns:
params : mapping of string to any Parameter names mapped to their values.
tf.contrib.learn.TensorFlowEstimator.get_tensor(name)
Returns tensor by name.
Args:
name
: string, name of the tensor.
Returns:
Tensor.
tf.contrib.learn.TensorFlowEstimator.get_variable_names()
Returns list of all variable names in this model.
Returns:
List of names.
tf.contrib.learn.TensorFlowEstimator.get_variable_value(name)
Returns value of the variable given by name.
Args:
name
: string, name of the tensor.
Returns:
Numpy array - value of the tensor.
tf.contrib.learn.TensorFlowEstimator.model_dir
tf.contrib.learn.TensorFlowEstimator.partial_fit(x, y)
Incremental fit on a batch of samples.
This method is expected to be called several times consecutively on different or the same chunks of the dataset. This either can implement iterative training or out-of-core/online training. This is especially useful when the whole dataset is too big to fit in memory at the same time. Or when model is taking long time to converge, and you want to split up training into subparts.
Args:
x
: matrix or tensor of shape [n_samples, n_features...]. Can be iterator that returns arrays of features. The training input samples for fitting the model.y
: vector or matrix [n_samples] or [n_samples, n_outputs]. Can be iterator that returns array of targets. The training target values (class label in classification, real numbers in regression).
Returns:
Returns self.
tf.contrib.learn.TensorFlowEstimator.predict(x, axis=1, batch_size=None)
Predict class or regression for x
.
For a classification model, the predicted class for each sample in x
is
returned. For a regression model, the predicted value based on x
is
returned.
Args:
x
: array-like matrix, [n_samples, n_features...] or iterator.axis
: Which axis to argmax for classification. By default axis 1 (next after batch) is used. Use 2 for sequence predictions.batch_size
: If test set is too big, use batch size to split it into mini batches. By default the batch_size member variable is used.
Returns:
y
: array of shape [n_samples]. The predicted classes or predicted value.
tf.contrib.learn.TensorFlowEstimator.predict_proba(x, batch_size=None)
Predict class probability of the input samples x
.
Args:
x
: array-like matrix, [n_samples, n_features...] or iterator.batch_size
: If test set is too big, use batch size to split it into mini batches. By default the batch_size member variable is used.
Returns:
y
: array of shape [n_samples, n_classes]. The predicted probabilities for each class.
tf.contrib.learn.TensorFlowEstimator.restore(cls, path, config=None)
Restores model from give path.
Args:
path
: Path to the checkpoints and other model information.config
: RunConfig object that controls the configurations of the session, e.g. num_cores, gpu_memory_fraction, etc. This is allowed to be reconfigured.
Returns:
Estimator, object of the subclass of TensorFlowEstimator.
Raises:
ValueError
: ifpath
does not contain a model definition.
tf.contrib.learn.TensorFlowEstimator.save(path)
Saves checkpoints and graph to given path.
Args:
path
: Folder to save model to.
tf.contrib.learn.TensorFlowEstimator.set_params(**params)
Set the parameters of this estimator.
The method works on simple estimators as well as on nested objects
(such as pipelines). The former have parameters of the form
<component>__<parameter>
so that it's possible to update each
component of a nested object.
Args:
**params
: Parameters.
Returns:
self
Raises:
ValueError
: If params contain invalid names.
class tf.contrib.learn.LinearClassifier
Linear classifier model.
Train a linear model to classify instances into one of multiple possible classes. When number of possible classes is 2, this is binary classification.
Example:
education = sparse_column_with_hash_bucket(column_name="education",
hash_bucket_size=1000)
occupation = sparse_column_with_hash_bucket(column_name="occupation",
hash_bucket_size=1000)
education_x_occupation = crossed_column(columns=[education, occupation],
hash_bucket_size=10000)
# Estimator using the default optimizer.
estimator = LinearClassifier(
feature_columns=[occupation, education_x_occupation])
# Or estimator using the FTRL optimizer with regularization.
estimator = LinearClassifier(
feature_columns=[occupation, education_x_occupation],
optimizer=tf.train.FtrlOptimizer(
learning_rate=0.1,
l1_regularization_strength=0.001
))
# Or estimator using the SDCAOptimizer.
estimator = LinearClassifier(
feature_columns=[occupation, education_x_occupation],
optimizer=tf.contrib.linear_optimizer.SDCAOptimizer(
example_id_column='example_id',
num_loss_partitions=...,
symmetric_l2_regularization=2.0
))
# Input builders
def input_fn_train: # returns x, y
...
def input_fn_eval: # returns x, y
...
estimator.fit(input_fn=input_fn_train)
estimator.evaluate(input_fn=input_fn_eval)
estimator.predict(x=x)
Input of fit
and evaluate
should have following features,
otherwise there will be a KeyError
:
- if
weight_column_name
is notNone
, a feature withkey=weight_column_name
whose value is aTensor
. - for each
column
infeature_columns
:- if
column
is aSparseColumn
, a feature withkey=column.name
whosevalue
is aSparseTensor
. - if
column
is aWeightedSparseColumn
, two features: the first withkey
the id column name, the second withkey
the weight column name. Both features'value
must be aSparseTensor
. - if
column
is aRealValuedColumn
, a feature withkey=column.name
whosevalue
is aTensor
.
- if
tf.contrib.learn.LinearClassifier.__init__(feature_columns, model_dir=None, n_classes=2, weight_column_name=None, optimizer=None, gradient_clip_norm=None, enable_centered_bias=None, _joint_weight=False, config=None, feature_engineering_fn=None)
Construct a LinearClassifier
estimator object.
Args:
feature_columns
: An iterable containing all the feature columns used by the model. All items in the set should be instances of classes derived fromFeatureColumn
.model_dir
: Directory to save model parameters, graph and etc. This can also be used to load checkpoints from the directory into a estimator to continue training a previously saved model.n_classes
: number of target classes. Default is binary classification.weight_column_name
: A string defining feature column name representing weights. It is used to down weight or boost examples during training. It will be multiplied by the loss of the example.optimizer
: The optimizer used to train the model. If specified, it should be either an instance oftf.Optimizer
or the SDCAOptimizer. IfNone
, the Ftrl optimizer will be used.gradient_clip_norm
: Afloat
> 0. If provided, gradients are clipped to their global norm with this clipping ratio. Seetf.clip_by_global_norm
for more details.enable_centered_bias
: A bool. If True, estimator will learn a centered bias variable for each class. Rest of the model structure learns the residual after centered bias. _joint_weight: If True, the weights for all columns will be stored in a single (possibly partitioned) variable. It's more efficient, but it's incompatible with SDCAOptimizer, and requires all feature columns are sparse and use the 'sum' combiner.config
:RunConfig
object to configure the runtime settings.feature_engineering_fn
: Feature engineering function. Takes features andtargets which are the output of `input_fn` and returns features and targets which will be fed into the model.
Returns:
A LinearClassifier
estimator.
Raises:
ValueError
: if n_classes < 2.
tf.contrib.learn.LinearClassifier.bias_
DEPRECATED FUNCTION
THIS FUNCTION IS DEPRECATED. It will be removed after 2016-10-30. Instructions for updating: This method will be removed after the deprecation date. To inspect variables, use get_variable_names() and get_variable_value().
tf.contrib.learn.LinearClassifier.config
tf.contrib.learn.LinearClassifier.evaluate(x=None, y=None, input_fn=None, feed_fn=None, batch_size=None, steps=None, metrics=None, name=None)
See evaluable.Evaluable.
tf.contrib.learn.LinearClassifier.export(export_dir, input_fn=None, input_feature_key=None, use_deprecated_input_fn=True, signature_fn=None, default_batch_size=1, exports_to_keep=None)
See BaseEstimator.export.
tf.contrib.learn.LinearClassifier.fit(x=None, y=None, input_fn=None, steps=None, batch_size=None, monitors=None, max_steps=None)
See trainable.Trainable.
tf.contrib.learn.LinearClassifier.get_estimator()
tf.contrib.learn.LinearClassifier.get_variable_names()
tf.contrib.learn.LinearClassifier.get_variable_value(name)
tf.contrib.learn.LinearClassifier.model_dir
tf.contrib.learn.LinearClassifier.predict(x=None, input_fn=None, batch_size=None, as_iterable=False)
Runs inference to determine the predicted class.
tf.contrib.learn.LinearClassifier.predict_proba(x=None, input_fn=None, batch_size=None, outputs=None, as_iterable=False)
Runs inference to determine the class probability predictions.
tf.contrib.learn.LinearClassifier.weights_
DEPRECATED FUNCTION
THIS FUNCTION IS DEPRECATED. It will be removed after 2016-10-30. Instructions for updating: This method will be removed after the deprecation date. To inspect variables, use get_variable_names() and get_variable_value().
class tf.contrib.learn.LinearRegressor
Linear regressor model.
Train a linear regression model to predict target variable value given observation of feature values.
Example:
education = sparse_column_with_hash_bucket(column_name="education",
hash_bucket_size=1000)
occupation = sparse_column_with_hash_bucket(column_name="occupation",
hash_bucket_size=1000)
education_x_occupation = crossed_column(columns=[education, occupation],
hash_bucket_size=10000)
estimator = LinearRegressor(
feature_columns=[occupation, education_x_occupation])
# Input builders
def input_fn_train: # returns x, y
...
def input_fn_eval: # returns x, y
...
estimator.fit(input_fn=input_fn_train)
estimator.evaluate(input_fn=input_fn_eval)
estimator.predict(x=x)
Input of fit
and evaluate
should have following features,
otherwise there will be a KeyError:
- if
weight_column_name
is notNone
: key=weight_column_name, value=aTensor
- for column in
feature_columns
:- if isinstance(column,
SparseColumn
): key=column.name, value=aSparseTensor
- if isinstance(column,
WeightedSparseColumn
): {key=id column name, value=aSparseTensor
, key=weight column name, value=aSparseTensor
} - if isinstance(column,
RealValuedColumn
): key=column.name, value=aTensor
- if isinstance(column,
tf.contrib.learn.LinearRegressor.__init__(feature_columns, model_dir=None, weight_column_name=None, optimizer=None, gradient_clip_norm=None, enable_centered_bias=None, target_dimension=1, _joint_weights=False, config=None, feature_engineering_fn=None)
Construct a LinearRegressor
estimator object.
Args:
feature_columns
: An iterable containing all the feature columns used by the model. All items in the set should be instances of classes derived fromFeatureColumn
.model_dir
: Directory to save model parameters, graph, etc. This can also be used to load checkpoints from the directory into a estimator to continue training a previously saved model.weight_column_name
: A string defining feature column name representing weights. It is used to down weight or boost examples during training. It will be multiplied by the loss of the example.optimizer
: An instance oftf.Optimizer
used to train the model. IfNone
, will use an Ftrl optimizer.gradient_clip_norm
: Afloat
> 0. If provided, gradients are clipped to their global norm with this clipping ratio. Seetf.clip_by_global_norm
for more details.enable_centered_bias
: A bool. If True, estimator will learn a centered bias variable for each class. Rest of the model structure learns the residual after centered bias.target_dimension
: dimension of the target for multilabels. _joint_weights: If True use a single (possibly partitioned) variable to store the weights. It's faster, but requires all feature columns are sparse and have the 'sum' combiner. Incompatible with SDCAOptimizer.config
:RunConfig
object to configure the runtime settings.feature_engineering_fn
: Feature engineering function. Takes features andtargets which are the output of `input_fn` and returns features and targets which will be fed into the model.
Returns:
A LinearRegressor
estimator.
tf.contrib.learn.LinearRegressor.__repr__()
tf.contrib.learn.LinearRegressor.bias_
DEPRECATED FUNCTION
THIS FUNCTION IS DEPRECATED. It will be removed after 2016-10-30. Instructions for updating: This method will be removed after the deprecation date. To inspect variables, use get_variable_names() and get_variable_value().
tf.contrib.learn.LinearRegressor.config
tf.contrib.learn.LinearRegressor.dnn_bias_
Returns bias of deep neural network part. (deprecated)
THIS FUNCTION IS DEPRECATED. It will be removed after 2016-10-30. Instructions for updating: This method will be removed after the deprecation date. To inspect variables, use get_variable_names() and get_variable_value().
tf.contrib.learn.LinearRegressor.dnn_weights_
Returns weights of deep neural network part. (deprecated)
THIS FUNCTION IS DEPRECATED. It will be removed after 2016-10-30. Instructions for updating: This method will be removed after the deprecation date. To inspect variables, use get_variable_names() and get_variable_value().
tf.contrib.learn.LinearRegressor.evaluate(x=None, y=None, input_fn=None, feed_fn=None, batch_size=None, steps=None, metrics=None, name=None)
See Evaluable
.
Raises:
ValueError
: If at least one ofx
ory
is provided, and at least one ofinput_fn
orfeed_fn
is provided. Or ifmetrics
is notNone
ordict
.
tf.contrib.learn.LinearRegressor.export(*args, **kwargs)
Exports inference graph into given dir. (deprecated arguments)
SOME ARGUMENTS ARE DEPRECATED. They will be removed after 2016-09-23. Instructions for updating: The signature of the input_fn accepted by export is changing to be consistent with what's used by tf.Learn Estimator's train/evaluate. input_fn (and in most cases, input_feature_key) will become required args, and use_deprecated_input_fn will default to False and be removed altogether.
Args:
export_dir: A string containing a directory to write the exported graph
and checkpoints.
input_fn: If `use_deprecated_input_fn` is true, then a function that given
`Tensor` of `Example` strings, parses it into features that are then
passed to the model. Otherwise, a function that takes no argument and
returns a tuple of (features, targets), where features is a dict of
string key to `Tensor` and targets is a `Tensor` that's currently not
used (and so can be `None`).
input_feature_key: Only used if `use_deprecated_input_fn` is false. String
key into the features dict returned by `input_fn` that corresponds toa
the raw `Example` strings `Tensor` that the exported model will take as
input. Can only be `None` if you're using a custom `signature_fn` that
does not use the first arg (examples).
use_deprecated_input_fn: Determines the signature format of `input_fn`.
signature_fn: Function that returns a default signature and a named
signature map, given `Tensor` of `Example` strings, `dict` of `Tensor`s
for features and `Tensor` or `dict` of `Tensor`s for predictions.
prediction_key: The key for a tensor in the `predictions` dict (output
from the `model_fn`) to use as the `predictions` input to the
`signature_fn`. Optional. If `None`, predictions will pass to
`signature_fn` without filtering.
default_batch_size: Default batch size of the `Example` placeholder.
exports_to_keep: Number of exports to keep.
Returns:
The string path to the exported directory. NB: this functionality was
added ca. 2016/09/25; clients that depend on the return value may need
to handle the case where this function returns None because subclasses
are not returning a value.
tf.contrib.learn.LinearRegressor.fit(x=None, y=None, input_fn=None, steps=None, batch_size=None, monitors=None, max_steps=None)
See Trainable
.
Raises:
ValueError
: Ifx
ory
are notNone
whileinput_fn
is notNone
.ValueError
: If bothsteps
andmax_steps
are notNone
.
tf.contrib.learn.LinearRegressor.get_params(deep=True)
Get parameters for this estimator.
Args:
deep
: boolean, optionalIf
True
, will return the parameters for this estimator and contained subobjects that are estimators.
Returns:
params : mapping of string to any Parameter names mapped to their values.
tf.contrib.learn.LinearRegressor.get_variable_names()
Returns list of all variable names in this model.
Returns:
List of names.
tf.contrib.learn.LinearRegressor.get_variable_value(name)
Returns value of the variable given by name.
Args:
name
: string, name of the tensor.
Returns:
Numpy array - value of the tensor.
tf.contrib.learn.LinearRegressor.linear_bias_
Returns bias of the linear part. (deprecated)
THIS FUNCTION IS DEPRECATED. It will be removed after 2016-10-30. Instructions for updating: This method will be removed after the deprecation date. To inspect variables, use get_variable_names() and get_variable_value().
tf.contrib.learn.LinearRegressor.linear_weights_
Returns weights per feature of the linear part. (deprecated)
THIS FUNCTION IS DEPRECATED. It will be removed after 2016-10-30. Instructions for updating: This method will be removed after the deprecation date. To inspect variables, use get_variable_names() and get_variable_value().
tf.contrib.learn.LinearRegressor.model_dir
tf.contrib.learn.LinearRegressor.partial_fit(x=None, y=None, input_fn=None, steps=1, batch_size=None, monitors=None)
Incremental fit on a batch of samples.
This method is expected to be called several times consecutively on different or the same chunks of the dataset. This either can implement iterative training or out-of-core/online training.
This is especially useful when the whole dataset is too big to fit in memory at the same time. Or when model is taking long time to converge, and you want to split up training into subparts.
Args:
x
: Matrix of shape [n_samples, n_features...]. Can be iterator that returns arrays of features. The training input samples for fitting the model. If set,input_fn
must beNone
.y
: Vector or matrix [n_samples] or [n_samples, n_outputs]. Can be iterator that returns array of targets. The training target values (class labels in classification, real numbers in regression). If set,input_fn
must beNone
.input_fn
: Input function. If set,x
,y
, andbatch_size
must beNone
.steps
: Number of steps for which to train model. IfNone
, train forever.batch_size
: minibatch size to use on the input, defaults to first dimension ofx
. Must beNone
ifinput_fn
is provided.monitors
: List ofBaseMonitor
subclass instances. Used for callbacks inside the training loop.
Returns:
self
, for chaining.
Raises:
ValueError
: If at least one ofx
andy
is provided, andinput_fn
is provided.
tf.contrib.learn.LinearRegressor.predict(*args, **kwargs)
Returns predictions for given features. (deprecated arguments)
SOME ARGUMENTS ARE DEPRECATED. They will be removed after 2016-09-15. Instructions for updating: The default behavior of predict() is changing. The default value for as_iterable will change to True, and then the flag will be removed altogether. The behavior of this flag is described below.
Args:
x: Matrix of shape [n_samples, n_features...]. Can be iterator that
returns arrays of features. The training input samples for fitting the
model. If set, `input_fn` must be `None`.
input_fn: Input function. If set, `x` and 'batch_size' must be `None`.
batch_size: Override default batch size. If set, 'input_fn' must be
'None'.
outputs: list of `str`, name of the output to predict.
If `None`, returns all.
as_iterable: If True, return an iterable which keeps yielding predictions
for each example until inputs are exhausted. Note: The inputs must
terminate if you want the iterable to terminate (e.g. be sure to pass
num_epochs=1 if you are using something like read_batch_features).
Returns:
A numpy array of predicted classes or regression values if the
constructor's `model_fn` returns a `Tensor` for `predictions` or a `dict`
of numpy arrays if `model_fn` returns a `dict`. Returns an iterable of
predictions if as_iterable is True.
Raises:
ValueError: If x and input_fn are both provided or both `None`.
tf.contrib.learn.LinearRegressor.set_params(**params)
Set the parameters of this estimator.
The method works on simple estimators as well as on nested objects
(such as pipelines). The former have parameters of the form
<component>__<parameter>
so that it's possible to update each
component of a nested object.
Args:
**params
: Parameters.
Returns:
self
Raises:
ValueError
: If params contain invalid names.
tf.contrib.learn.LinearRegressor.weights_
DEPRECATED FUNCTION
THIS FUNCTION IS DEPRECATED. It will be removed after 2016-10-30. Instructions for updating: This method will be removed after the deprecation date. To inspect variables, use get_variable_names() and get_variable_value().
class tf.contrib.learn.TensorFlowRNNClassifier
TensorFlow RNN Classifier model.
tf.contrib.learn.TensorFlowRNNClassifier.__init__(rnn_size, n_classes, cell_type='gru', num_layers=1, input_op_fn=null_input_op_fn, initial_state=None, bidirectional=False, sequence_length=None, attn_length=None, attn_size=None, attn_vec_size=None, batch_size=32, steps=50, optimizer='Adagrad', learning_rate=0.1, class_weight=None, clip_gradients=5.0, continue_training=False, config=None, verbose=1)
Initializes a TensorFlowRNNClassifier instance.
Args:
rnn_size
: The size for rnn cell, e.g. size of your word embeddings.cell_type
: The type of rnn cell, including rnn, gru, and lstm.num_layers
: The number of layers of the rnn model.input_op_fn
: Function that will transform the input tensor, such as creating word embeddings, byte list, etc. This takes an argument x for input and returns transformed x.bidirectional
: boolean, Whether this is a bidirectional rnn.sequence_length
: If sequence_length is provided, dynamic calculation is performed. This saves computational time when unrolling past max sequence length.initial_state
: An initial state for the RNN. This must be a tensor of appropriate type and shape [batch_size x cell.state_size].attn_length
: integer, the size of attention vector attached to rnn cells.attn_size
: integer, the size of an attention window attached to rnn cells.attn_vec_size
: integer, the number of convolutional features calculated on attention state and the size of the hidden layer built from base cell state.n_classes
: Number of classes in the target.batch_size
: Mini batch size.steps
: Number of steps to run over data.optimizer
: Optimizer name (or class), for example "SGD", "Adam", "Adagrad".learning_rate
: If this is constant float value, no decay function is used. Instead, a customized decay function can be passed that accepts global_step as parameter and returns a Tensor. e.g. exponential decay function:def exp_decay(global_step): return tf.train.exponential_decay( learning_rate=0.1, global_step, decay_steps=2, decay_rate=0.001)
class_weight
: None or list of n_classes floats. Weight associated with classes for loss computation. If not given, all classes are supposed to have weight one.continue_training
: when continue_training is True, once initialized model will be continually trained on every call of fit.config
: RunConfig object that controls the configurations of the session, e.g. num_cores, gpu_memory_fraction, etc.
tf.contrib.learn.TensorFlowRNNClassifier.__repr__()
tf.contrib.learn.TensorFlowRNNClassifier.bias_
Returns bias of the rnn layer.
tf.contrib.learn.TensorFlowRNNClassifier.config
tf.contrib.learn.TensorFlowRNNClassifier.evaluate(x=None, y=None, input_fn=None, feed_fn=None, batch_size=None, steps=None, metrics=None, name=None)
Evaluates given model with provided evaluation data.
See superclass Estimator for more details.
Args:
x
: features.y
: targets.input_fn
: Input function.feed_fn
: Function creating a feed dict every time it is called.batch_size
: minibatch size to use on the input.steps
: Number of steps for which to evaluate model.metrics
: Dict of metric ops to run. If None, the default metrics are used.name
: Name of the evaluation.
Returns:
Returns dict
with evaluation results.
tf.contrib.learn.TensorFlowRNNClassifier.export(*args, **kwargs)
Exports inference graph into given dir. (deprecated arguments)
SOME ARGUMENTS ARE DEPRECATED. They will be removed after 2016-09-23. Instructions for updating: The signature of the input_fn accepted by export is changing to be consistent with what's used by tf.Learn Estimator's train/evaluate. input_fn (and in most cases, input_feature_key) will become required args, and use_deprecated_input_fn will default to False and be removed altogether.
Args:
export_dir: A string containing a directory to write the exported graph
and checkpoints.
input_fn: If `use_deprecated_input_fn` is true, then a function that given
`Tensor` of `Example` strings, parses it into features that are then
passed to the model. Otherwise, a function that takes no argument and
returns a tuple of (features, targets), where features is a dict of
string key to `Tensor` and targets is a `Tensor` that's currently not
used (and so can be `None`).
input_feature_key: Only used if `use_deprecated_input_fn` is false. String
key into the features dict returned by `input_fn` that corresponds toa
the raw `Example` strings `Tensor` that the exported model will take as
input. Can only be `None` if you're using a custom `signature_fn` that
does not use the first arg (examples).
use_deprecated_input_fn: Determines the signature format of `input_fn`.
signature_fn: Function that returns a default signature and a named
signature map, given `Tensor` of `Example` strings, `dict` of `Tensor`s
for features and `Tensor` or `dict` of `Tensor`s for predictions.
prediction_key: The key for a tensor in the `predictions` dict (output
from the `model_fn`) to use as the `predictions` input to the
`signature_fn`. Optional. If `None`, predictions will pass to
`signature_fn` without filtering.
default_batch_size: Default batch size of the `Example` placeholder.
exports_to_keep: Number of exports to keep.
Returns:
The string path to the exported directory. NB: this functionality was
added ca. 2016/09/25; clients that depend on the return value may need
to handle the case where this function returns None because subclasses
are not returning a value.
tf.contrib.learn.TensorFlowRNNClassifier.fit(x, y, steps=None, monitors=None, logdir=None)
Neural network model from provided model_fn
and training data.
Note: called first time constructs the graph and initializers variables. Subsequently, it will continue training the same model. This logic follows partial_fit() interface in scikit-learn. To restart learning, create new estimator.
Args:
x
: matrix or tensor of shape [n_samples, n_features...]. Can be iterator that returns arrays of features. The training input samples for fitting the model.y
: vector or matrix [n_samples] or [n_samples, n_outputs]. Can be iterator that returns array of targets. The training target values (class labels in classification, real numbers in regression).steps
: int, number of steps to train.If None or 0, train for `self.steps`.
monitors
: List ofBaseMonitor
objects to print training progress and invoke early stopping.logdir
: the directory to save the log file that can be used for optional visualization.
Returns:
Returns self.
tf.contrib.learn.TensorFlowRNNClassifier.get_params(deep=True)
Get parameters for this estimator.
Args:
deep
: boolean, optionalIf
True
, will return the parameters for this estimator and contained subobjects that are estimators.
Returns:
params : mapping of string to any Parameter names mapped to their values.
tf.contrib.learn.TensorFlowRNNClassifier.get_tensor(name)
Returns tensor by name.
Args:
name
: string, name of the tensor.
Returns:
Tensor.
tf.contrib.learn.TensorFlowRNNClassifier.get_variable_names()
Returns list of all variable names in this model.
Returns:
List of names.
tf.contrib.learn.TensorFlowRNNClassifier.get_variable_value(name)
Returns value of the variable given by name.
Args:
name
: string, name of the tensor.
Returns:
Numpy array - value of the tensor.
tf.contrib.learn.TensorFlowRNNClassifier.model_dir
tf.contrib.learn.TensorFlowRNNClassifier.partial_fit(x, y)
Incremental fit on a batch of samples.
This method is expected to be called several times consecutively on different or the same chunks of the dataset. This either can implement iterative training or out-of-core/online training. This is especially useful when the whole dataset is too big to fit in memory at the same time. Or when model is taking long time to converge, and you want to split up training into subparts.
Args:
x
: matrix or tensor of shape [n_samples, n_features...]. Can be iterator that returns arrays of features. The training input samples for fitting the model.y
: vector or matrix [n_samples] or [n_samples, n_outputs]. Can be iterator that returns array of targets. The training target values (class label in classification, real numbers in regression).
Returns:
Returns self.
tf.contrib.learn.TensorFlowRNNClassifier.predict(x, axis=1, batch_size=None)
Predict class or regression for x
.
For a classification model, the predicted class for each sample in x
is
returned. For a regression model, the predicted value based on x
is
returned.
Args:
x
: array-like matrix, [n_samples, n_features...] or iterator.axis
: Which axis to argmax for classification. By default axis 1 (next after batch) is used. Use 2 for sequence predictions.batch_size
: If test set is too big, use batch size to split it into mini batches. By default the batch_size member variable is used.
Returns:
y
: array of shape [n_samples]. The predicted classes or predicted value.
tf.contrib.learn.TensorFlowRNNClassifier.predict_proba(x, batch_size=None)
Predict class probability of the input samples x
.
Args:
x
: array-like matrix, [n_samples, n_features...] or iterator.batch_size
: If test set is too big, use batch size to split it into mini batches. By default the batch_size member variable is used.
Returns:
y
: array of shape [n_samples, n_classes]. The predicted probabilities for each class.
tf.contrib.learn.TensorFlowRNNClassifier.restore(cls, path, config=None)
Restores model from give path.
Args:
path
: Path to the checkpoints and other model information.config
: RunConfig object that controls the configurations of the session, e.g. num_cores, gpu_memory_fraction, etc. This is allowed to be reconfigured.
Returns:
Estimator, object of the subclass of TensorFlowEstimator.
Raises:
ValueError
: ifpath
does not contain a model definition.
tf.contrib.learn.TensorFlowRNNClassifier.save(path)
Saves checkpoints and graph to given path.
Args:
path
: Folder to save model to.
tf.contrib.learn.TensorFlowRNNClassifier.set_params(**params)
Set the parameters of this estimator.
The method works on simple estimators as well as on nested objects
(such as pipelines). The former have parameters of the form
<component>__<parameter>
so that it's possible to update each
component of a nested object.
Args:
**params
: Parameters.
Returns:
self
Raises:
ValueError
: If params contain invalid names.
tf.contrib.learn.TensorFlowRNNClassifier.weights_
Returns weights of the rnn layer.
class tf.contrib.learn.TensorFlowRNNRegressor
TensorFlow RNN Regressor model.
tf.contrib.learn.TensorFlowRNNRegressor.__init__(rnn_size, cell_type='gru', num_layers=1, input_op_fn=null_input_op_fn, initial_state=None, bidirectional=False, sequence_length=None, attn_length=None, attn_size=None, attn_vec_size=None, n_classes=0, batch_size=32, steps=50, optimizer='Adagrad', learning_rate=0.1, clip_gradients=5.0, continue_training=False, config=None, verbose=1)
Initializes a TensorFlowRNNRegressor instance.
Args:
rnn_size
: The size for rnn cell, e.g. size of your word embeddings.cell_type
: The type of rnn cell, including rnn, gru, and lstm.num_layers
: The number of layers of the rnn model.input_op_fn
: Function that will transform the input tensor, such as creating word embeddings, byte list, etc. This takes an argument x for input and returns transformed x.bidirectional
: boolean, Whether this is a bidirectional rnn.sequence_length
: If sequence_length is provided, dynamic calculation is performed. This saves computational time when unrolling past max sequence length.attn_length
: integer, the size of attention vector attached to rnn cells.attn_size
: integer, the size of an attention window attached to rnn cells.attn_vec_size
: integer, the number of convolutional features calculated on attention state and the size of the hidden layer built from base cell state.initial_state
: An initial state for the RNN. This must be a tensor of appropriate type and shape [batch_size x cell.state_size].batch_size
: Mini batch size.steps
: Number of steps to run over data.optimizer
: Optimizer name (or class), for example "SGD", "Adam", "Adagrad".learning_rate
: If this is constant float value, no decay function is used. Instead, a customized decay function can be passed that accepts global_step as parameter and returns a Tensor. e.g. exponential decay function:def exp_decay(global_step): return tf.train.exponential_decay( learning_rate=0.1, global_step, decay_steps=2, decay_rate=0.001)
continue_training
: when continue_training is True, once initialized model will be continually trained on every call of fit.config
: RunConfig object that controls the configurations of the session, e.g. num_cores, gpu_memory_fraction, etc.verbose
: Controls the verbosity, possible values:- 0: the algorithm and debug information is muted.
- 1: trainer prints the progress.
- 2: log device placement is printed.
tf.contrib.learn.TensorFlowRNNRegressor.__repr__()
tf.contrib.learn.TensorFlowRNNRegressor.bias_
Returns bias of the rnn layer.
tf.contrib.learn.TensorFlowRNNRegressor.config
tf.contrib.learn.TensorFlowRNNRegressor.evaluate(x=None, y=None, input_fn=None, feed_fn=None, batch_size=None, steps=None, metrics=None, name=None)
Evaluates given model with provided evaluation data.
See superclass Estimator for more details.
Args:
x
: features.y
: targets.input_fn
: Input function.feed_fn
: Function creating a feed dict every time it is called.batch_size
: minibatch size to use on the input.steps
: Number of steps for which to evaluate model.metrics
: Dict of metric ops to run. If None, the default metrics are used.name
: Name of the evaluation.
Returns:
Returns dict
with evaluation results.
tf.contrib.learn.TensorFlowRNNRegressor.export(*args, **kwargs)
Exports inference graph into given dir. (deprecated arguments)
SOME ARGUMENTS ARE DEPRECATED. They will be removed after 2016-09-23. Instructions for updating: The signature of the input_fn accepted by export is changing to be consistent with what's used by tf.Learn Estimator's train/evaluate. input_fn (and in most cases, input_feature_key) will become required args, and use_deprecated_input_fn will default to False and be removed altogether.
Args:
export_dir: A string containing a directory to write the exported graph
and checkpoints.
input_fn: If `use_deprecated_input_fn` is true, then a function that given
`Tensor` of `Example` strings, parses it into features that are then
passed to the model. Otherwise, a function that takes no argument and
returns a tuple of (features, targets), where features is a dict of
string key to `Tensor` and targets is a `Tensor` that's currently not
used (and so can be `None`).
input_feature_key: Only used if `use_deprecated_input_fn` is false. String
key into the features dict returned by `input_fn` that corresponds toa
the raw `Example` strings `Tensor` that the exported model will take as
input. Can only be `None` if you're using a custom `signature_fn` that
does not use the first arg (examples).
use_deprecated_input_fn: Determines the signature format of `input_fn`.
signature_fn: Function that returns a default signature and a named
signature map, given `Tensor` of `Example` strings, `dict` of `Tensor`s
for features and `Tensor` or `dict` of `Tensor`s for predictions.
prediction_key: The key for a tensor in the `predictions` dict (output
from the `model_fn`) to use as the `predictions` input to the
`signature_fn`. Optional. If `None`, predictions will pass to
`signature_fn` without filtering.
default_batch_size: Default batch size of the `Example` placeholder.
exports_to_keep: Number of exports to keep.
Returns:
The string path to the exported directory. NB: this functionality was
added ca. 2016/09/25; clients that depend on the return value may need
to handle the case where this function returns None because subclasses
are not returning a value.
tf.contrib.learn.TensorFlowRNNRegressor.fit(x, y, steps=None, monitors=None, logdir=None)
Neural network model from provided model_fn
and training data.
Note: called first time constructs the graph and initializers variables. Subsequently, it will continue training the same model. This logic follows partial_fit() interface in scikit-learn. To restart learning, create new estimator.
Args:
x
: matrix or tensor of shape [n_samples, n_features...]. Can be iterator that returns arrays of features. The training input samples for fitting the model.y
: vector or matrix [n_samples] or [n_samples, n_outputs]. Can be iterator that returns array of targets. The training target values (class labels in classification, real numbers in regression).steps
: int, number of steps to train.If None or 0, train for `self.steps`.
monitors
: List ofBaseMonitor
objects to print training progress and invoke early stopping.logdir
: the directory to save the log file that can be used for optional visualization.
Returns:
Returns self.
tf.contrib.learn.TensorFlowRNNRegressor.get_params(deep=True)
Get parameters for this estimator.
Args:
deep
: boolean, optionalIf
True
, will return the parameters for this estimator and contained subobjects that are estimators.
Returns:
params : mapping of string to any Parameter names mapped to their values.
tf.contrib.learn.TensorFlowRNNRegressor.get_tensor(name)
Returns tensor by name.
Args:
name
: string, name of the tensor.
Returns:
Tensor.
tf.contrib.learn.TensorFlowRNNRegressor.get_variable_names()
Returns list of all variable names in this model.
Returns:
List of names.
tf.contrib.learn.TensorFlowRNNRegressor.get_variable_value(name)
Returns value of the variable given by name.
Args:
name
: string, name of the tensor.
Returns:
Numpy array - value of the tensor.
tf.contrib.learn.TensorFlowRNNRegressor.model_dir
tf.contrib.learn.TensorFlowRNNRegressor.partial_fit(x, y)
Incremental fit on a batch of samples.
This method is expected to be called several times consecutively on different or the same chunks of the dataset. This either can implement iterative training or out-of-core/online training. This is especially useful when the whole dataset is too big to fit in memory at the same time. Or when model is taking long time to converge, and you want to split up training into subparts.
Args:
x
: matrix or tensor of shape [n_samples, n_features...]. Can be iterator that returns arrays of features. The training input samples for fitting the model.y
: vector or matrix [n_samples] or [n_samples, n_outputs]. Can be iterator that returns array of targets. The training target values (class label in classification, real numbers in regression).
Returns:
Returns self.
tf.contrib.learn.TensorFlowRNNRegressor.predict(x, axis=1, batch_size=None)
Predict class or regression for x
.
For a classification model, the predicted class for each sample in x
is
returned. For a regression model, the predicted value based on x
is
returned.
Args:
x
: array-like matrix, [n_samples, n_features...] or iterator.axis
: Which axis to argmax for classification. By default axis 1 (next after batch) is used. Use 2 for sequence predictions.batch_size
: If test set is too big, use batch size to split it into mini batches. By default the batch_size member variable is used.
Returns:
y
: array of shape [n_samples]. The predicted classes or predicted value.
tf.contrib.learn.TensorFlowRNNRegressor.predict_proba(x, batch_size=None)
Predict class probability of the input samples x
.
Args:
x
: array-like matrix, [n_samples, n_features...] or iterator.batch_size
: If test set is too big, use batch size to split it into mini batches. By default the batch_size member variable is used.
Returns:
y
: array of shape [n_samples, n_classes]. The predicted probabilities for each class.
tf.contrib.learn.TensorFlowRNNRegressor.restore(cls, path, config=None)
Restores model from give path.
Args:
path
: Path to the checkpoints and other model information.config
: RunConfig object that controls the configurations of the session, e.g. num_cores, gpu_memory_fraction, etc. This is allowed to be reconfigured.
Returns:
Estimator, object of the subclass of TensorFlowEstimator.
Raises:
ValueError
: ifpath
does not contain a model definition.
tf.contrib.learn.TensorFlowRNNRegressor.save(path)
Saves checkpoints and graph to given path.
Args:
path
: Folder to save model to.
tf.contrib.learn.TensorFlowRNNRegressor.set_params(**params)
Set the parameters of this estimator.
The method works on simple estimators as well as on nested objects
(such as pipelines). The former have parameters of the form
<component>__<parameter>
so that it's possible to update each
component of a nested object.
Args:
**params
: Parameters.
Returns:
self
Raises:
ValueError
: If params contain invalid names.
tf.contrib.learn.TensorFlowRNNRegressor.weights_
Returns weights of the rnn layer.
Graph actions
Perform various training, evaluation, and inference actions on a graph.
class tf.train.NanLossDuringTrainingError
tf.train.NanLossDuringTrainingError.__str__()
class tf.contrib.learn.RunConfig
This class specifies the specific configurations for the run.
If you're a Google-internal user using command line flags with learn_runner.py (for instance, to do distributed training or to use parameter servers), you probably want to use learn_runner.EstimatorConfig instead.
tf.contrib.learn.RunConfig.__init__(master=None, task=None, num_ps_replicas=None, num_cores=0, log_device_placement=False, gpu_memory_fraction=1, cluster_spec=None, tf_random_seed=None, save_summary_steps=100, save_checkpoints_secs=600, save_checkpoints_steps=None, keep_checkpoint_max=5, keep_checkpoint_every_n_hours=10000, job_name=None, is_chief=None, evaluation_master='')
Constructor.
If set to None, master
, task
, num_ps_replicas
, cluster_spec
,
job_name
, and is_chief
are set based on the TF_CONFIG environment
variable, if the pertinent information is present; otherwise, the defaults
listed in the Args section apply.
The TF_CONFIG environment variable is a JSON object with two relevant
attributes: task
and cluster_spec
. cluster_spec
is a JSON serialized
version of the Python dict described in server_lib.py. task
has two
attributes: type
and index
, where type
can be any of the task types
in the cluster_spec. When TF_CONFIG contains said information, the
following properties are set on this class:
job_name
is set to [task
][type
]task
is set to [task
][index
]cluster_spec
is parsed from [cluster
]- 'master' is determined by looking up
job_name
andtask
in the cluster_spec. num_ps_replicas
is set by counting the number of nodes listed in theps
job ofcluster_spec
.is_chief
: true whenjob_name
== "master" andtask
== 0.
Example:
cluster = {'ps': ['host1:2222', 'host2:2222'],
'worker': ['host3:2222', 'host4:2222', 'host5:2222']}
os.environ['TF_CONFIG'] = json.dumps({
{'cluster': cluster,
'task': {'type': 'worker', 'index': 1}}})
config = RunConfig()
assert config.master == 'host4:2222'
assert config.task == 1
assert config.num_ps_replicas == 2
assert config.cluster_spec == server_lib.ClusterSpec(cluster)
assert config.job_name == 'worker'
assert not config.is_chief
Args:
master
: TensorFlow master. Defaults to empty string for local.task
: Task id of the replica running the training (default: 0).num_ps_replicas
: Number of parameter server tasks to use (default: 0).num_cores
: Number of cores to be used. If 0, the system picks an appropriate number (default: 0).log_device_placement
: Log the op placement to devices (default: False).gpu_memory_fraction
: Fraction of GPU memory used by the process on each GPU uniformly on the same machine.cluster_spec
: atf.train.ClusterSpec
object that describes the cluster in the case of distributed computation. If missing, reasonable assumptions are made for the addresses of jobs.tf_random_seed
: Random seed for TensorFlow initializers. Setting this value allows consistency between reruns.save_summary_steps
: Save summaries every this many steps.save_checkpoints_secs
: Save checkpoints every this many seconds. Can not be specified withsave_checkpoints_steps
.save_checkpoints_steps
: Save checkpoints every this many steps. Can not be specified withsave_checkpoints_secs
.keep_checkpoint_max
: The maximum number of recent checkpoint files to keep. As new files are created, older files are deleted. If None or 0, all checkpoint files are kept. Defaults to 5 (that is, the 5 most recent checkpoint files are kept.)keep_checkpoint_every_n_hours
: Number of hours between each checkpoint to be saved. The default value of 10,000 hours effectively disables the feature.job_name
: the type of task, e.g., 'ps', 'worker', etc. Thejob_name
must exist in thecluster_spec.jobs
.is_chief
: whether or not this task (as identified by the other parameters) should be the chief task.evaluation_master
: the master on which to perform evaluation.
Raises:
ValueError
: if num_ps_replicas and cluster_spec are set (cluster_spec may come from the TF_CONFIG environment variable).
tf.contrib.learn.RunConfig.is_chief
tf.contrib.learn.RunConfig.job_name
tf.contrib.learn.evaluate(graph, output_dir, checkpoint_path, eval_dict, update_op=None, global_step_tensor=None, supervisor_master='', log_every_steps=10, feed_fn=None, max_steps=None)
Evaluate a model loaded from a checkpoint.
Given graph
, a directory to write summaries to (output_dir
), a checkpoint
to restore variables from, and a dict
of Tensor
s to evaluate, run an eval
loop for max_steps
steps, or until an exception (generally, an
end-of-input signal from a reader operation) is raised from running
eval_dict
.
In each step of evaluation, all tensors in the eval_dict
are evaluated, and
every log_every_steps
steps, they are logged. At the very end of evaluation,
a summary is evaluated (finding the summary ops using Supervisor
's logic)
and written to output_dir
.
Args:
graph
: AGraph
to train. It is expected that this graph is not in use elsewhere.output_dir
: A string containing the directory to write a summary to.checkpoint_path
: A string containing the path to a checkpoint to restore. Can beNone
if the graph doesn't require loading any variables.eval_dict
: Adict
mapping string names to tensors to evaluate. It is evaluated in every logging step. The result of the final evaluation is returned. Ifupdate_op
is None, then it's evaluated in every step. Ifmax_steps
isNone
, this should depend on a reader that will raise an end-of-input exception when the inputs are exhausted.update_op
: ATensor
which is run in every step.global_step_tensor
: AVariable
containing the global step. IfNone
, one is extracted from the graph using the same logic as inSupervisor
. Used to place eval summaries on training curves.supervisor_master
: The master string to use when preparing the session.log_every_steps
: Integer. Output logs everylog_every_steps
evaluation steps. The logs contain theeval_dict
and timing information.feed_fn
: A function that is called every iteration to produce afeed_dict
passed tosession.run
calls. Optional.max_steps
: Integer. Evaluateeval_dict
this many times.
Returns:
A tuple (eval_results, global_step)
:
eval_results
: Adict
mappingstring
to numeric values (int
,float
) that are the result of running eval_dict in the last step.None
if no eval steps were run.global_step
: The global step this evaluation corresponds to.
Raises:
ValueError
: ifoutput_dir
is empty.
tf.contrib.learn.infer(restore_checkpoint_path, output_dict, feed_dict=None)
Restore graph from restore_checkpoint_path
and run output_dict
tensors.
If restore_checkpoint_path
is supplied, restore from checkpoint. Otherwise,
init all variables.
Args:
restore_checkpoint_path
: A string containing the path to a checkpoint to restore.output_dict
: Adict
mapping string names toTensor
objects to run. Tensors must all be from the same graph.feed_dict
:dict
object mappingTensor
objects to input values to feed.
Returns:
Dict of values read from output_dict
tensors. Keys are the same as
output_dict
, values are the results read from the corresponding Tensor
in output_dict
.
Raises:
ValueError
: ifoutput_dict
orfeed_dicts
is None or empty.
tf.contrib.learn.run_feeds(*args, **kwargs)
See run_feeds_iter(). Returns a list
instead of an iterator.
tf.contrib.learn.run_n(output_dict, feed_dict=None, restore_checkpoint_path=None, n=1)
Run output_dict
tensors n
times, with the same feed_dict
each run.
Args:
output_dict
: Adict
mapping string names to tensors to run. Must all be from the same graph.feed_dict
:dict
of input values to feed each run.restore_checkpoint_path
: A string containing the path to a checkpoint to restore.n
: Number of times to repeat.
Returns:
A list of n
dict
objects, each containing values read from output_dict
tensors.
tf.contrib.learn.train(graph, output_dir, train_op, loss_op, global_step_tensor=None, init_op=None, init_feed_dict=None, init_fn=None, log_every_steps=10, supervisor_is_chief=True, supervisor_master='', supervisor_save_model_secs=600, keep_checkpoint_max=5, supervisor_save_summaries_steps=100, feed_fn=None, steps=None, fail_on_nan_loss=True, monitors=None, max_steps=None)
Train a model.
Given graph
, a directory to write outputs to (output_dir
), and some ops,
run a training loop. The given train_op
performs one step of training on the
model. The loss_op
represents the objective function of the training. It is
expected to increment the global_step_tensor
, a scalar integer tensor
counting training steps. This function uses Supervisor
to initialize the
graph (from a checkpoint if one is available in output_dir
), write summaries
defined in the graph, and write regular checkpoints as defined by
supervisor_save_model_secs
.
Training continues until global_step_tensor
evaluates to max_steps
, or, if
fail_on_nan_loss
, until loss_op
evaluates to NaN
. In that case the
program is terminated with exit code 1.
Args:
graph
: A graph to train. It is expected that this graph is not in use elsewhere.output_dir
: A directory to write outputs to.train_op
: An op that performs one training step when run.loss_op
: A scalar loss tensor.global_step_tensor
: A tensor representing the global step. If none is given, one is extracted from the graph using the same logic as inSupervisor
.init_op
: An op that initializes the graph. IfNone
, useSupervisor
's default.init_feed_dict
: A dictionary that mapsTensor
objects to feed values. This feed dictionary will be used wheninit_op
is evaluated.init_fn
: Optional callable passed to Supervisor to initialize the model.log_every_steps
: Output logs regularly. The logs contain timing data and the current loss.supervisor_is_chief
: Whether the current process is the chief supervisor in charge of restoring the model and running standard services.supervisor_master
: The master string to use when preparing the session.supervisor_save_model_secs
: Save a checkpoint everysupervisor_save_model_secs
seconds when training.keep_checkpoint_max
: The maximum number of recent checkpoint files to keep. As new files are created, older files are deleted. If None or 0, all checkpoint files are kept. This is simply passed as the max_to_keep arg to tf.Saver constructor.supervisor_save_summaries_steps
: Save summaries everysupervisor_save_summaries_steps
seconds when training.feed_fn
: A function that is called every iteration to produce afeed_dict
passed tosession.run
calls. Optional.steps
: Trains for this many steps (e.g. current global step +steps
).fail_on_nan_loss
: If true, raiseNanLossDuringTrainingError
ifloss_op
evaluates toNaN
. If false, continue training as if nothing happened.monitors
: List ofBaseMonitor
subclass instances. Used for callbacks inside the training loop.max_steps
: Number of total steps for which to train model. IfNone
, train forever. Two calls fit(steps=100) means 200 training iterations. On the other hand two calls of fit(max_steps=100) means, second call will not do any iteration since first call did all 100 steps.
Returns:
The final loss value.
Raises:
ValueError
: Ifoutput_dir
,train_op
,loss_op
, orglobal_step_tensor
is not provided. Seetf.contrib.framework.get_global_step
for how we look up the latter if not provided explicitly.NanLossDuringTrainingError
: Iffail_on_nan_loss
isTrue
, and loss ever evaluates toNaN
.ValueError
: If bothsteps
andmax_steps
are notNone
.
Input processing
Queue and read batched input data.
tf.contrib.learn.extract_dask_data(data)
Extract data from dask.Series or dask.DataFrame for predictors.
tf.contrib.learn.extract_dask_labels(labels)
Extract data from dask.Series for labels.
tf.contrib.learn.extract_pandas_data(data)
Extract data from pandas.DataFrame for predictors.
Given a DataFrame, will extract the values and cast them to float. The DataFrame is expected to contain values of type int, float or bool.
Args:
data
:pandas.DataFrame
containing the data to be extracted.
Returns:
A numpy ndarray
of the DataFrame's values as floats.
Raises:
ValueError
: if data contains types other than int, float or bool.
tf.contrib.learn.extract_pandas_labels(labels)
Extract data from pandas.DataFrame for labels.
Args:
labels
:pandas.DataFrame
orpandas.Series
containing one column of labels to be extracted.
Returns:
A numpy ndarray
of labels from the DataFrame.
Raises:
ValueError
: if more than one column is found or type is not int, float or bool.
tf.contrib.learn.extract_pandas_matrix(data)
Extracts numpy matrix from pandas DataFrame.
Args:
data
:pandas.DataFrame
containing the data to be extracted.
Returns:
A numpy ndarray
of the DataFrame's values.
tf.contrib.learn.read_batch_examples(file_pattern, batch_size, reader, randomize_input=True, num_epochs=None, queue_capacity=10000, num_threads=1, read_batch_size=1, parse_fn=None, name=None)
Adds operations to read, queue, batch Example
protos.
Given file pattern (or list of files), will setup a queue for file names,
read Example
proto using provided reader
, use batch queue to create
batches of examples of size batch_size
.
All queue runners are added to the queue runners collection, and may be
started via start_queue_runners
.
All ops are added to the default graph.
Use parse_fn
if you need to do parsing / processing on single examples.
Args:
file_pattern
: List of files or pattern of file paths containingExample
records. Seetf.gfile.Glob
for pattern rules.batch_size
: An int or scalarTensor
specifying the batch size to use.reader
: A function or class that returns an object withread
method, (filename tensor) -> (example tensor).randomize_input
: Whether the input should be randomized.num_epochs
: Integer specifying the number of times to read through the dataset. IfNone
, cycles through the dataset forever. NOTE - If specified, creates a variable that must be initialized, so calltf.initialize_all_variables()
as shown in the tests.queue_capacity
: Capacity for input queue.num_threads
: The number of threads enqueuing examples.read_batch_size
: An int or scalarTensor
specifying the number of records to read at onceparse_fn
: Parsing function, takesExample
Tensor returns parsed representation. IfNone
, no parsing is done.name
: Name of resulting op.
Returns:
String Tensor
of batched Example
proto.
Raises:
ValueError
: for invalid inputs.
tf.contrib.learn.read_batch_features(file_pattern, batch_size, features, reader, randomize_input=True, num_epochs=None, queue_capacity=10000, feature_queue_capacity=100, reader_num_threads=1, parser_num_threads=1, parse_fn=None, name=None)
Adds operations to read, queue, batch and parse Example
protos.
Given file pattern (or list of files), will setup a queue for file names,
read Example
proto using provided reader
, use batch queue to create
batches of examples of size batch_size
and parse example given features
specification.
All queue runners are added to the queue runners collection, and may be
started via start_queue_runners
.
All ops are added to the default graph.
Args:
file_pattern
: List of files or pattern of file paths containingExample
records. Seetf.gfile.Glob
for pattern rules.batch_size
: An int or scalarTensor
specifying the batch size to use.features
: Adict
mapping feature keys toFixedLenFeature
orVarLenFeature
values.reader
: A function or class that returns an object withread
method, (filename tensor) -> (example tensor).randomize_input
: Whether the input should be randomized.num_epochs
: Integer specifying the number of times to read through the dataset. If None, cycles through the dataset forever. NOTE - If specified, creates a variable that must be initialized, so call tf.initialize_local_variables() as shown in the tests.queue_capacity
: Capacity for input queue.feature_queue_capacity
: Capacity of the parsed features queue. Set this value to a small number, for example 5 if the parsed features are large.reader_num_threads
: The number of threads to read examples.parser_num_threads
: The number of threads to parse examples. records to read at onceparse_fn
: Parsing function, takesExample
Tensor returns parsed representation. IfNone
, no parsing is done.name
: Name of resulting op.
Returns:
A dict of Tensor
or SparseTensor
objects for each in features
.
Raises:
ValueError
: for invalid inputs.
tf.contrib.learn.read_batch_record_features(file_pattern, batch_size, features, randomize_input=True, num_epochs=None, queue_capacity=10000, reader_num_threads=1, parser_num_threads=1, name='dequeue_record_examples')
Reads TFRecord, queues, batches and parses Example
proto.
See more detailed description in read_examples
.
Args:
file_pattern
: List of files or pattern of file paths containingExample
records. Seetf.gfile.Glob
for pattern rules.batch_size
: An int or scalarTensor
specifying the batch size to use.features
: Adict
mapping feature keys toFixedLenFeature
orVarLenFeature
values.randomize_input
: Whether the input should be randomized.num_epochs
: Integer specifying the number of times to read through the dataset. If None, cycles through the dataset forever. NOTE - If specified, creates a variable that must be initialized, so call tf.initialize_local_variables() as shown in the tests.queue_capacity
: Capacity for input queue.reader_num_threads
: The number of threads to read examples.parser_num_threads
: The number of threads to parse examples.name
: Name of resulting op.
Returns:
A dict of Tensor
or SparseTensor
objects for each in features
.
Raises:
ValueError
: for invalid inputs.