| /external/tensorflow/tensorflow/python/training/ |
| learning_rate_decay.py | 32 def exponential_decay(learning_rate,
49 decayed_learning_rate = learning_rate *
62 learning_rate = tf.train.exponential_decay(starter_learning_rate, global_step,
66 tf.train.GradientDescentOptimizer(learning_rate)
72 learning_rate: A scalar `float32` or `float64` `Tensor` or a
85 A scalar `Tensor` of the same type as `learning_rate`. The decayed
95 [learning_rate, global_step, decay_steps, decay_rate]) as name:
96 learning_rate = ops.convert_to_tensor(learning_rate, name="learning_rate")
[all...] |
| momentum.py | 36 variable -= learning_rate * accumulation
46 def __init__(self, learning_rate, momentum,
51 learning_rate: A `Tensor` or a floating point value. The learning rate.
64 When eager execution is enabled, learning_rate and momentum can each be a
71 self._learning_rate = learning_rate
80 learning_rate = self._learning_rate
81 if callable(learning_rate):
82 learning_rate = learning_rate()
83 self._learning_rate_tensor = ops.convert_to_tensor(learning_rate,
[all...] |
| gradient_descent.py | 34 def __init__(self, learning_rate, use_locking=False, name="GradientDescent"):
38 learning_rate: A Tensor or a floating point value. The learning
45 self._learning_rate = learning_rate
73 name="learning_rate")
|
| proximal_gradient_descent.py | 38 def __init__(self, learning_rate, l1_regularization_strength=0.0,
44 learning_rate: A Tensor or a floating point value. The learning
55 self._learning_rate = learning_rate
101 name="learning_rate")
|
| /external/tensorflow/tensorflow/python/estimator/canned/ |
| optimizers.py | 41 def get_optimizer_instance(opt, learning_rate=None):
46 * A string: Creates an `Optimizer` subclass with the given `learning_rate`.
56 learning_rate: A float. Only used if `opt` is a string.
63 ValueError: If `opt` is a supported string but `learning_rate` was not
69 if not learning_rate:
70 raise ValueError('learning_rate must be specified when opt is string.')
71 return _OPTIMIZER_CLS_NAMES[opt](learning_rate=learning_rate)
|
| optimizers_test.py | 36 optimizers.get_optimizer_instance('unsupported_name', learning_rate=0.1)
40 ValueError, 'learning_rate must be specified when opt is string'):
41 optimizers.get_optimizer_instance('Adagrad', learning_rate=None)
44 opt = optimizers.get_optimizer_instance('Adagrad', learning_rate=0.1)
49 opt = optimizers.get_optimizer_instance('Adam', learning_rate=0.1)
54 opt = optimizers.get_optimizer_instance('Ftrl', learning_rate=0.1)
59 opt = optimizers.get_optimizer_instance('RMSProp', learning_rate=0.1)
64 opt = optimizers.get_optimizer_instance('SGD', learning_rate=0.1)
|
| linear.py | 46 learning_rate = min(_LEARNING_RATE, 1.0 / math.sqrt(len(feature_columns)))
47 return ftrl.FtrlOptimizer(learning_rate=learning_rate)
143 learning_rate=_LEARNING_RATE)
199 learning_rate=0.1,
|
| /external/tensorflow/tensorflow/contrib/opt/python/training/ |
| addsign_test.py | 63 learning_rate=0.1,
91 learning_rate=learning_rate,
127 learning_rate,
137 learning_rate,
153 self._testDense(use_resource=False, learning_rate=0.01, alpha=0.1, beta=0.8)
159 self._testDense(use_resource=True, learning_rate=0.01, alpha=0.1, beta=0.8)
166 learning_rate=0.1,
199 learning_rate=learning_rate,
[all...] |
| powersign_test.py | 64 learning_rate=0.1,
92 learning_rate=learning_rate,
129 learning_rate,
139 learning_rate,
156 learning_rate=0.1,
164 self._testDense(use_resource=True, learning_rate=0.1, base=10.0, beta=0.8)
171 learning_rate=0.1,
204 learning_rate=learning_rate,
[all...] |
| /external/webrtc/webrtc/base/ |
| rollingaccumulator.h | 124 // Weights nth sample with weight (learning_rate)^n. Learning_rate should be
126 double ComputeWeightedMean(double learning_rate) const {
127 if (count_ < 1 || learning_rate <= 0.0 || learning_rate >= 1.0) {
135 current_weight *= learning_rate;
|
| /external/tensorflow/tensorflow/contrib/layers/python/layers/ |
| optimizers.py | 50 "learning_rate",
60 learning_rate,
81 Alternatively, if `learning_rate` is `None`, the function takes no
82 arguments. E.g. `optimize_loss(..., learning_rate=None,
99 learning_rate: float or `Tensor`, magnitude of update per each training
119 learning_rate_decay_fn: function, takes `learning_rate` and `global_step`
124 Ignored if `learning_rate` is not supplied.
149 * `learning_rate` is an invalid type or value.
152 * `learning_rate` and `learning_rate_decay_fn` are supplied, but no
172 if learning_rate is not None
[all...] |
| optimizers_test.py | 63 gradient_descent.GradientDescentOptimizer(learning_rate=0.1),
64 lambda lr: gradient_descent.GradientDescentOptimizer(learning_rate=lr)
71 loss, global_step, learning_rate=0.1, optimizer=optimizer)
81 return gradient_descent.GradientDescentOptimizer(learning_rate=0.1)
87 loss, global_step, learning_rate=None, optimizer=optimizer_fn)
102 loss, global_step, learning_rate=0.1, optimizer=optimizer)
109 loss, global_step, learning_rate=0.1, optimizer="SGD",
117 None, global_step, learning_rate=0.1, optimizer="SGD")
120 [[1.0]], global_step, learning_rate=0.1, optimizer="SGD")
133 learning_rate=0.1
[all...] |
| /external/tensorflow/tensorflow/examples/learn/ |
| iris_custom_decay_dnn.py | 55 learning_rate = tf.train.exponential_decay(
56 learning_rate=0.1, global_step=global_step,
58 optimizer = tf.train.AdagradOptimizer(learning_rate=learning_rate)
|
| /external/tensorflow/tensorflow/contrib/timeseries/python/timeseries/ |
| test_utils.py | 104 learning_rate=0.1, ignore_params_fn=lambda _: (),
156 optimizer=adam.AdamOptimizer(learning_rate))
172 learning_rate=0.1, rtol=0.2, atol=0.1, train_loss_tolerance_coeff=0.99,
188 learning_rate: Step size for optimization.
214 train_iterations=train_iterations, seed=seed, learning_rate=learning_rate,
257 learning_rate=0.1,
269 learning_rate: Step size for optimization.
279 seed=seed, learning_rate=learning_rate,
[all...] |
| /external/tensorflow/tensorflow/contrib/training/python/training/ |
| training_test.py | 99 optimizer = gradient_descent.GradientDescentOptimizer(learning_rate=1.0)
116 optimizer = gradient_descent.GradientDescentOptimizer(learning_rate=1.0)
150 optimizer = gradient_descent.GradientDescentOptimizer(learning_rate=1.0)
183 optimizer = gradient_descent.GradientDescentOptimizer(learning_rate=1.0)
206 optimizer = gradient_descent.GradientDescentOptimizer(learning_rate=1.0)
245 optimizer = gradient_descent.GradientDescentOptimizer(learning_rate=1.0)
280 optimizer = gradient_descent.GradientDescentOptimizer(learning_rate=1.0)
303 optimizer = gradient_descent.GradientDescentOptimizer(learning_rate=1.0)
329 optimizer = gradient_descent.GradientDescentOptimizer(learning_rate=1.0)
349 def create_train_op(self, learning_rate=1.0, gradient_multiplier=1.0)
[all...] |
| /external/tensorflow/tensorflow/contrib/learn/python/learn/estimators/ |
| dynamic_rnn_estimator_test.py | 260 learning_rate=0.1)
361 learning_rate = 0.1
398 learning_rate=learning_rate,
417 learning_rate = 0.1
454 learning_rate=learning_rate,
513 learning_rate = 0.1
548 learning_rate=learning_rate,
[all...] |
| dynamic_rnn_estimator.py | 389 learning_rate=None,
428 learning_rate: Learning rate used for optimization. This argument has no
524 learning_rate=learning_rate,
553 learning_rate=0.1,
619 learning_rate: Learning rate. This argument has no effect if `optimizer`
671 optimizer = momentum_opt.MomentumOptimizer(learning_rate, momentum)
682 learning_rate=learning_rate,
|
| state_saving_rnn_estimator.py | 393 learning_rate=None,
429 learning_rate: Learning rate used for optimization. This argument has no
517 learning_rate=learning_rate,
542 learning_rate=0.1,
578 learning_rate: Learning rate. This argument has no effect if `optimizer`
628 optimizer_type = momentum_opt.MomentumOptimizer(learning_rate, momentum)
643 learning_rate=learning_rate,
|
| state_saving_rnn_estimator_test.py | 349 learning_rate=0.1)
472 learning_rate = 0.3
502 learning_rate=learning_rate,
531 learning_rate = 0.5
559 learning_rate=learning_rate,
603 learning_rate = 0.4
642 learning_rate=learning_rate,
[all...] |
| /external/tensorflow/tensorflow/examples/tutorials/mnist/ |
| mnist.py | 100 def training(loss, learning_rate):
112 learning_rate: The learning rate to use for gradient descent.
120 optimizer = tf.train.GradientDescentOptimizer(learning_rate)
|
| /external/tensorflow/tensorflow/contrib/slim/python/slim/ |
| learning_test.py | 252 optimizer = gradient_descent.GradientDescentOptimizer(learning_rate=1.0)
287 optimizer = gradient_descent.GradientDescentOptimizer(learning_rate=1.0)
321 optimizer = gradient_descent.GradientDescentOptimizer(learning_rate=1.0)
354 optimizer = gradient_descent.GradientDescentOptimizer(learning_rate=1.0)
379 optimizer = gradient_descent.GradientDescentOptimizer(learning_rate=1.0)
406 optimizer = gradient_descent.GradientDescentOptimizer(learning_rate=1.0)
439 optimizer = gradient_descent.GradientDescentOptimizer(learning_rate=1.0)
458 optimizer = gradient_descent.GradientDescentOptimizer(learning_rate=1.0)
477 optimizer = gradient_descent.GradientDescentOptimizer(learning_rate=1.0)
513 optimizer = gradient_descent.GradientDescentOptimizer(learning_rate=1.0
[all...] |
| /external/tensorflow/tensorflow/contrib/eager/python/examples/mnist/ |
| mnist_graph_test.py | 46 optimizer = tf.train.GradientDescentOptimizer(learning_rate=1.0)
|
| /external/tensorflow/tensorflow/python/keras/_impl/keras/ |
| model_subclassing_test.py | 182 optimizer=RMSPropOptimizer(learning_rate=0.001),
202 optimizer=RMSPropOptimizer(learning_rate=0.001),
223 model.compile(loss='mse', optimizer=RMSPropOptimizer(learning_rate=0.001))
241 model.compile(loss='mse', optimizer=RMSPropOptimizer(learning_rate=0.001))
261 model.compile(loss='mse', optimizer=RMSPropOptimizer(learning_rate=0.001))
294 model.compile(loss='mse', optimizer=RMSPropOptimizer(learning_rate=0.001))
326 model.compile(loss='mse', optimizer=RMSPropOptimizer(learning_rate=0.001))
358 model.compile(loss='mse', optimizer=RMSPropOptimizer(learning_rate=0.001))
378 model.compile(loss='mse', optimizer=RMSPropOptimizer(learning_rate=0.001))
387 model.compile(loss='mse', optimizer=RMSPropOptimizer(learning_rate=0.001)
[all...] |
| /external/tensorflow/tensorflow/contrib/eager/python/examples/rnn_ptb/ |
| rnn_ptb.py | 307 # Make learning_rate a Variable so it can be included in the checkpoint
308 # and we can resume training with the last saved learning_rate.
309 learning_rate = tfe.Variable(20.0, name="learning_rate")
310 sys.stderr.write("learning_rate=%f\n" % learning_rate.numpy())
314 optimizer = tf.train.GradientDescentOptimizer(learning_rate)
322 tfe.Saver(model.trainable_weights + [learning_rate]).save(
326 learning_rate.assign(learning_rate / 4.0
[all...] |
| /external/tensorflow/tensorflow/python/keras/_impl/keras/engine/ |
| training_eager_test.py | 45 optimizer = RMSPropOptimizer(learning_rate=0.001)
196 model.compile(optimizer=RMSPropOptimizer(learning_rate=0.001), loss='mse')
213 optimizer = RMSPropOptimizer(learning_rate=0.001)
291 optimizer=RMSPropOptimizer(learning_rate=0.001))
305 optimizer=RMSPropOptimizer(learning_rate=0.001),
330 optimizer=RMSPropOptimizer(learning_rate=0.001))
397 optimizer=RMSPropOptimizer(learning_rate=0.001))
475 optimizer=RMSPropOptimizer(learning_rate=0.001),
493 optimizer=RMSPropOptimizer(learning_rate=0.001))
512 optimizer=RMSPropOptimizer(learning_rate=0.001)
[all...] |
