SGD(params, lr=<required parameter>, momentum=0, dampening=0, weight_decay=0, nesterov=False)¶
Implements stochastic gradient descent (optionally with momentum).
Nesterov momentum is based on the formula from On the importance of initialization and momentum in deep learning.
params (iterable) – iterable of parameters to optimize or dicts defining parameter groups
lr (float) – learning rate
momentum (float, optional) – momentum factor (default: 0)
weight_decay (float, optional) – weight decay (L2 penalty) (default: 0)
dampening (float, optional) – dampening for momentum (default: 0)
nesterov (bool, optional) – enables Nesterov momentum (default: False)
>>> optimizer = torch.optim.SGD(model.parameters(), lr=0.1, momentum=0.9) >>> optimizer.zero_grad() >>> loss_fn(model(input), target).backward() >>> optimizer.step()
The implementation of SGD with Momentum/Nesterov subtly differs from Sutskever et. al. and implementations in some other frameworks.
Considering the specific case of Momentum, the update can be written as
where , , and denote the parameters, gradient, velocity, and momentum respectively.
This is in contrast to Sutskever et. al. and other frameworks which employ an update of the form
The Nesterov version is analogously modified.
Add a param group to the
This can be useful when fine tuning a pre-trained network as frozen layers can be made trainable and added to the
Optimizeras training progresses.
param_group (dict) – Specifies what Tensors should be optimized along with group
optimization options. (specific) –
Loads the optimizer state.
Returns the state of the optimizer as a
It contains two entries:
- state - a dict holding current optimization state. Its content
differs between optimizer classes.
param_groups - a dict containing all parameter groups
Performs a single optimization step.
closure (callable, optional) – A closure that reevaluates the model and returns the loss.
Sets the gradients of all optimized
torch.Tensors to zero.
set_to_none (bool) – instead of setting to zero, set the grads to None. This will in general have lower memory footprint, and can modestly improve performance. However, it changes certain behaviors. For example: 1. When the user tries to access a gradient and perform manual ops on it, a None attribute or a Tensor full of 0s will behave differently. 2. If the user requests
zero_grad(set_to_none=True)followed by a backward pass,
.grads are guaranteed to be None for params that did not receive a gradient. 3.
torch.optimoptimizers have a different behavior if the gradient is 0 or None (in one case it does the step with a gradient of 0 and in the other it skips the step altogether).