def testStaircase(self): initial_lr = 0.1 k = 10 decay_rate = 0.96 step = resource_variable_ops.ResourceVariable(0) decayed_lr = learning_rate_decay_v2.inverse_time_decay( initial_lr, step, k, decay_rate, staircase=True) self.evaluate(variables.global_variables_initializer()) for i in range(k + 1): expected = initial_lr / (1 + decay_rate * (i // k)) self.assertAllClose(self.evaluate(decayed_lr()), expected, 1e-6) self.evaluate(step.assign_add(1))
def testStaircase(self): initial_lr = 0.1 k = 10 decay_rate = 0.96 step = resource_variable_ops.ResourceVariable(0) decayed_lr = learning_rate_decay_v2.inverse_time_decay( initial_lr, step, k, decay_rate, staircase=True) self.evaluate(variables.global_variables_initializer()) for i in range(k + 1): expected = initial_lr / (1 + decay_rate * (i // k)) self.assertAllClose(self.evaluate(decayed_lr()), expected, 1e-6) self.evaluate(step.assign_add(1))
def inverse_time_decay(learning_rate, global_step, decay_steps, decay_rate, staircase=False, name=None): """Applies inverse time decay to the initial learning rate. When training a model, it is often recommended to lower the learning rate as the training progresses. This function applies an inverse decay function to a provided initial learning rate. It requires an `global_step` value to compute the decayed learning rate. You can just pass a TensorFlow variable that you increment at each training step. The function returns the decayed learning rate. It is computed as: ```python decayed_learning_rate = learning_rate / (1 + decay_rate * global_step / decay_step) ``` or, if `staircase` is `True`, as: ```python decayed_learning_rate = learning_rate / (1 + decay_rate * floor(global_step / decay_step)) ``` Example: decay 1/t with a rate of 0.5: ```python ... global_step = tf.Variable(0, trainable=False) learning_rate = 0.1 decay_steps = 1.0 decay_rate = 0.5 learning_rate = tf.train.inverse_time_decay(learning_rate, global_step, decay_steps, decay_rate) # Passing global_step to minimize() will increment it at each step. learning_step = ( tf.train.GradientDescentOptimizer(learning_rate) .minimize(...my loss..., global_step=global_step) ) ``` Args: learning_rate: A scalar `float32` or `float64` `Tensor` or a Python number. The initial learning rate. global_step: A Python number. Global step to use for the decay computation. Must not be negative. decay_steps: How often to apply decay. decay_rate: A Python number. The decay rate. staircase: Whether to apply decay in a discrete staircase, as opposed to continuous, fashion. name: String. Optional name of the operation. Defaults to 'InverseTimeDecay'. Returns: A scalar `Tensor` of the same type as `learning_rate`. The decayed learning rate. Raises: ValueError: if `global_step` is not supplied. @compatibility(eager) When eager execution is enabled, this function returns a function which in turn returns the decayed learning rate Tensor. This can be useful for changing the learning rate value across different invocations of optimizer functions. @end_compatibility """ decayed_lr = learning_rate_decay_v2.inverse_time_decay(learning_rate, global_step, decay_steps, decay_rate, staircase=staircase, name=name) if not context.executing_eagerly(): decayed_lr = decayed_lr() return decayed_lr
def inverse_time_decay(learning_rate, global_step, decay_steps, decay_rate, staircase=False, name=None): """Applies inverse time decay to the initial learning rate. When training a model, it is often recommended to lower the learning rate as the training progresses. This function applies an inverse decay function to a provided initial learning rate. It requires an `global_step` value to compute the decayed learning rate. You can just pass a TensorFlow variable that you increment at each training step. The function returns the decayed learning rate. It is computed as: ```python decayed_learning_rate = learning_rate / (1 + decay_rate * global_step / decay_step) ``` or, if `staircase` is `True`, as: ```python decayed_learning_rate = learning_rate / (1 + decay_rate * floor(global_step / decay_step)) ``` Example: decay 1/t with a rate of 0.5: ```python ... global_step = tf.Variable(0, trainable=False) learning_rate = 0.1 decay_steps = 1.0 decay_rate = 0.5 learning_rate = tf.train.inverse_time_decay(learning_rate, global_step, decay_steps, decay_rate) # Passing global_step to minimize() will increment it at each step. learning_step = ( tf.train.GradientDescentOptimizer(learning_rate) .minimize(...my loss..., global_step=global_step) ) ``` Args: learning_rate: A scalar `float32` or `float64` `Tensor` or a Python number. The initial learning rate. global_step: A Python number. Global step to use for the decay computation. Must not be negative. decay_steps: How often to apply decay. decay_rate: A Python number. The decay rate. staircase: Whether to apply decay in a discrete staircase, as opposed to continuous, fashion. name: String. Optional name of the operation. Defaults to 'InverseTimeDecay'. Returns: A scalar `Tensor` of the same type as `learning_rate`. The decayed learning rate. Raises: ValueError: if `global_step` is not supplied. @compatibility(eager) When eager execution is enabled, this function returns a function which in turn returns the decayed learning rate Tensor. This can be useful for changing the learning rate value across different invocations of optimizer functions. @end_compatibility """ decayed_lr = learning_rate_decay_v2.inverse_time_decay( learning_rate, global_step, decay_steps, decay_rate, staircase=staircase, name=name) if not context.executing_eagerly(): decayed_lr = decayed_lr() return decayed_lr