mpcrl.optim.GradientDescent

class mpcrl.optim.GradientDescent(learning_rate, weight_decay=0.0, momentum=0.0, dampening=0.0, nesterov=False, hook='on_update', max_percentage_update=inf, bound_consistency=False)

Bases: GradientBasedOptimizer[LrType]

First-order Gradient descent optimizer, based on [14] and torch.optim.SGD.

In its basic formulation, this optimizer updates the parameters as

\[\theta \gets \theta - \alpha g,\]

where \(\theta\) are the learnable parameters, \(\alpha\) is the learning rate (could be extended to the case this is a vector of rates), and \(g\) is the gradient of the loss function w.r.t. the parameters. If momentum or weight decay are used, the gradient \(g\) is modified before using it, but the update rule remains the same. However, when considering a constrained parameter space, we need to solve a Quadratic Programming (QP) problem to ensure the parameters stay within their bounds. For gradient descent, the QP problem is

\[\begin{split}\begin{aligned} \min_{\Delta\theta} & \quad \frac{1}{2} \lVert \Delta\theta \rVert_2^2 + \alpha g^\top \Delta\theta \\ \text{s.t.} & \quad \theta_{\text{lower}} \leq \theta + \Delta\theta \leq \theta_{\text{upper}} \end{aligned}\end{split}\]

followed by the update \(\theta \gets \theta + \Delta\theta\).

Parameters:

learning_ratefloat or array or mpcrl.core.schedulers.Scheduler

The learning rate of the optimizer. It can be:

• a float, in case the learning rate must stay constant and is the same for all learnable parameters

• an array, in case the learning rate must stay constant but is different for each parameter (should have the same size as the number of learnable parameters)

• a mpcrl.core.schedulers.Scheduler, in case the learning rate can vary during the learning process (usually, it is set to decay). See the hook argument for more details on when this scheduler is stepped.

weight_decayfloat, optional

A positive float that specifies the decay of the learnable parameters in the form of an L2 regularization term. By default, it is set to 0.0, so no decay/regularization takes place.

momentumfloat, optional

A positive float that specifies the momentum factor. By default, it is set to 0.0, so no momentum is used.

dampeningfloat, optional

A positive float that specifies the dampening factor for the momentum. By default, it is set to 0.0, so no dampening is used.

nesterovbool, optional

A boolean that specifies whether to use Nesterov momentum. By default, it is set to False.

hook{“on_update”, “on_episode_end”, “on_timestep_end”}, optional

Specifies when to step the optimizer’s learning rate’s scheduler to decay its value. This allows to vary the rate over the learning iterations. The options are:

• "on_update" steps the learning rate after each agent’s update

• "on_episode_end" steps the learning rate after each episode’s end

• "on_timestep_end" steps the learning rate after each env’s timestep.

By default, "on_update" is selected.

max_percentage_updatefloat, optional

A positive float that specifies the maximum percentage change the learnable parameters can experience in each update. For example, max_percentage_update=0.5 means that the parameters can be updated by up to 50% of their current value. By default, it is set to +inf. If specified, the update becomes constrained and has to be solved as a QP, which is inevitably slower than its unconstrained counterpart (a linear system).

bound_consistencybool, optional

A boolean that, if True, forces the learnable parameters to lie in their bounds when updated. This is done via numpy.clip. Only beneficial if numerical issues arise during updates, e.g., due to the QP solver not being able to guarantee bounds.

Methods

set_learnable_parameters(pars)	Makes the optimization class aware of the dictionary of the learnable parameters whose values are to be updated.
step(*_, **__)	Steps/decays the learning rate according to its scheduler.
update(gradient[, hessian])	Computes the gradient-based update of the learnable parameters dictated by the current RL algorithm.

Attributes

hook	Gets the hook to which the scheduler is attached to, i.e., when to step the learning rate's scheduler to decay its value.
order	Gets the order of the optimizer: 1 for first-order, 2 for second-order.

property hook: str | None

Gets the hook to which the scheduler is attached to, i.e., when to step the learning rate’s scheduler to decay its value.

Returns:

optional str

The hook to which the scheduler is attached to. Can be None in case no hook is needed (e.g., a scheduler was not passed as learning_rate).

property order: Literal[1, 2]

Gets the order of the optimizer: 1 for first-order, 2 for second-order.

Returns:

1 or 2

The order of the optimizer.

set_learnable_parameters(pars)

Makes the optimization class aware of the dictionary of the learnable parameters whose values are to be updated.

Parameters:

pars:class`mpcrl.LearnableParametersDict`

The dictionary of the learnable parameters.

Return type:

None

step(*_, **__)

Steps/decays the learning rate according to its scheduler.

Return type:

None

update(gradient, hessian=None)

Computes the gradient-based update of the learnable parameters dictated by the current RL algorithm.

Parameters:

gradient1D array

The gradient of the learnable parameters.

hessian2D array, optional

The hessian of the learnable parameters. When the optimizer is firt-order, it is expected to be None since it is unused. When the optimizer is second-order, it is expected to be a 2D array.

Returns:

statusstr, optional

An optional string containing the status of the update, e.g., the status of the QP solver, if used.

Return type:

Optional[str]

Examples using mpcrl.optim.GradientDescent

On-policy Deterministic Policy Gradient

On-policy Deterministic Policy Gradient