a Sic9@sddZddlZddlmmZddlmZddl m Z ddl m Z e dGddde j ZeZdS) z!RMSprop optimizer implementation.N)backend_config) optimizer_v2) keras_exportzkeras.optimizers.RMSpropcsfeZdZdZdZdfdd Zd d Zfd d ZdddZdddZ fddZ fddZ Z S)RMSpropa Optimizer that implements the RMSprop algorithm. The gist of RMSprop is to: - Maintain a moving (discounted) average of the square of gradients - Divide the gradient by the root of this average This implementation of RMSprop uses plain momentum, not Nesterov momentum. The centered version additionally maintains a moving average of the gradients, and uses that average to estimate the variance. Args: learning_rate: A `Tensor`, floating point value, or a schedule that is a `tf.keras.optimizers.schedules.LearningRateSchedule`, or a callable that takes no arguments and returns the actual value to use. The learning rate. Defaults to 0.001. rho: Discounting factor for the history/coming gradient. Defaults to 0.9. momentum: A scalar or a scalar `Tensor`. Defaults to 0.0. epsilon: A small constant for numerical stability. This epsilon is "epsilon hat" in the Kingma and Ba paper (in the formula just before Section 2.1), not the epsilon in Algorithm 1 of the paper. Defaults to 1e-7. centered: Boolean. If `True`, gradients are normalized by the estimated variance of the gradient; if False, by the uncentered second moment. Setting this to `True` may help with training, but is slightly more expensive in terms of computation and memory. Defaults to `False`. name: Optional name prefix for the operations created when applying gradients. Defaults to `"RMSprop"`. **kwargs: keyword arguments. Allowed arguments are `clipvalue`, `clipnorm`, `global_clipnorm`. If `clipvalue` (float) is set, the gradient of each weight is clipped to be no higher than this value. If `clipnorm` (float) is set, the gradient of each weight is individually clipped so that its norm is no higher than this value. If `global_clipnorm` (float) is set the gradient of all weights is clipped so that their global norm is no higher than this value. Note that in the dense implementation of this algorithm, variables and their corresponding accumulators (momentum, gradient moving average, square gradient moving average) will be updated even if the gradient is zero (i.e. accumulators will decay, momentum will be applied). The sparse implementation (used when the gradient is an `IndexedSlices` object, typically because of `tf.gather` or an embedding lookup in the forward pass) will not update variable slices or their accumulators unless those slices were used in the forward pass (nor is there an "eventual" correction to account for these omitted updates). This leads to more efficient updates for large embedding lookup tables (where most of the slices are not accessed in a particular graph execution), but differs from the published algorithm. Usage: >>> opt = tf.keras.optimizers.RMSprop(learning_rate=0.1) >>> var1 = tf.Variable(10.0) >>> loss = lambda: (var1 ** 2) / 2.0 # d(loss) / d(var1) = var1 >>> step_count = opt.minimize(loss, [var1]).numpy() >>> var1.numpy() 9.683772 Reference: - [Hinton, 2012]( http://www.cs.toronto.edu/~tijmen/csc321/slides/lecture_slides_lec6.pdf) TMbP??Hz>Fc stj|fi||d|d||d|j|d|d|_t|tjsdt |sd|dkrjd|_t|t t fr|dks|dkrt d |d t |d |d ||pt|_||_d S)aConstruct a new RMSprop optimizer. Args: learning_rate: A `Tensor`, floating point value, or a schedule that is a `tf.keras.optimizers.schedules.LearningRateSchedule`, or a callable that takes no arguments and returns the actual value to use. The learning rate. Defaults to 0.001. rho: Discounting factor for the history/coming gradient. Defaults to 0.9. momentum: A scalar or a scalar `Tensor`. Defaults to 0.0. epsilon: A small constant for numerical stability. This epsilon is "epsilon hat" in the Kingma and Ba paper (in the formula just before Section 2.1), not the epsilon in Algorithm 1 of the paper. Defaults to 1e-7. centered: Boolean. If `True`, gradients are normalized by the estimated variance of the gradient; if False, by the uncentered second moment. Setting this to `True` may help with training, but is slightly more expensive in terms of computation and memory. Defaults to `False`. name: Optional name prefix for the operations created when applying gradients. Defaults to "RMSprop". **kwargs: keyword arguments. Allowed to be {`clipnorm`, `clipvalue`, `lr`, `decay`}. `clipnorm` is clip gradients by norm; `clipvalue` is clip gradients by value, `decay` is included for backward compatibility to allow time inverse decay of learning rate. `lr` is included for backward compatibility, recommended to use `learning_rate` instead. @compatibility(eager) When eager execution is enabled, `learning_rate`, `decay`, `momentum`, and `epsilon` can each be a callable that takes no arguments and returns the actual value to use. This can be useful for changing these values across different invocations of optimizer functions. @end_compatibility learning_ratelrdecayrhoFrTz6`momentum` must be between [0, 1]. Received: momentum=z (of type z).momentumN)super__init__ _set_hyperget_initial_decay _momentum isinstancetfTensorcallableintfloat ValueErrortyperepsiloncentered)selfr r rrrnamekwargs __class__a/var/www/html/django/DPS/env/lib/python3.9/site-packages/keras/optimizers/optimizer_v2/rmsprop.pyr_s6-   zRMSprop.__init__cCsR|D]}||dq|jr2|D]}||dq |jrN|D]}||dqr?r@rrArBrrrrC"ResourceSparseApplyCenteredRMSProprErFResourceSparseApplyRMSProprIrJrKcontrol_dependencies_resource_scatter_addgatherrHrLgroup)r r;r+rTr8r6r7rNr'r:r(rms_scaled_g_valuesrO rms_slice denom_slicemg_scaled_g_valuesrQmg_slice var_updater%r%r&_resource_apply_sparses     ,   .zRMSprop._resource_apply_sparsecs:|j}t|t|dkr*tdg|}t|dS)Nrr)weightslennparrayr set_weights)r rbparamsr#r%r&rf?szRMSprop.set_weightsc s>t}||d|j|d|d|j|jd|S)Nr r r)r r r rrr)r get_configr3_serialize_hyperparameterrrr)r configr#r%r&rhHs  zRMSprop.get_config)rrrr Fr)N)N) __name__ __module__ __qualname____doc___HAS_AGGREGATE_GRADrr,r0rSrarfrh __classcell__r%r%r#r&rs@E  < G r)rnnumpyrdtensorflow.compat.v2rIv2rkerasrkeras.optimizers.optimizer_v2r tensorflow.python.util.tf_exportr OptimizerV2rRMSPropr%r%r%r&s   ?