ApplyAdam

tensorflow C++ API

tensorflow::ops::ApplyAdam

Update '*var' according to the Adam algorithm.

Summary

lr_t <- learning_rate * sqrt(1 - beta2^t) / (1 - beta1^t) m_t <- beta1 * m_{t-1} + (1 - beta1) * g_t v_t <- beta2 * v_{t-1} + (1 - beta2) * g_t * g_t variable <- variable - lr_t * m_t / (sqrt(v_t) + epsilon)

Arguments:

• scope: A Scope object
• var: Should be from a Variable().
• m: Should be from a Variable().
• v: Should be from a Variable().
• beta1_power: Must be a scalar.
• beta2_power: Must be a scalar.
• lr: Scaling factor. Must be a scalar.
• beta1: Momentum factor. Must be a scalar.
• beta2: Momentum factor. Must be a scalar.
• epsilon: Ridge term. Must be a scalar.
• grad: The gradient.

Optional attributes (seeAttrs):

• use_locking: If True, updating of the var, m, and v tensors will be protected by a lock; otherwise the behavior is undefined, but may exhibit less contention.
• use_nesterov: If True, uses the nesterov update.

Returns:

• Output : Same as "var".

ApplyAdam block

Source link : https://github.com/EXPNUNI/enuSpaceTensorflow/blob/master/enuSpaceTensorflow/tf_training.cpp

Argument:

• Scope scope : A Scope object (A scope is generated automatically each page. A scope is not connected.)
• Input var: connect Input node.
• Input m: connect Input node.
• Input v: connect Input node.
• Input beta1_power: connect Input node.
• Input beta2_power: connect Input node.
• Input lr: connect Input node.
• Input beta1: connect Input node.
• Input beta2: connect Input node.
• Input epsilon: connect Input node.
• Input grad: connect Input node.
• ApplyAdam ::Attrs attrs : Input attrs in value. ex) use_locking_ = false;usenesterov_=false;

Return:

• Output output : Output object of ApplyAdam class object.

Result:

• std::vector(Tensor) result_output : Returned object of executed result by calling session.

Using Method