from typing import Any, Callable, Dict, Optional, overload
import torch
from torch.optim import Optimizer
[docs]
class LARS(Optimizer):
"""Implements the Layer-wise Adaptive Rate Scaling (LARS) optimizer.
Implementation borrowed from lightly SSL library.
"""
def __init__(
self,
params: Any,
lr: float,
momentum: float = 0.9,
dampening: float = 0,
weight_decay: float = 0.9,
nesterov: bool = False,
trust_coefficient: float = 0.001,
eps: float = 1e-8,
):
"""Constructs a new LARS optimizer.
Parameters
----------
params : Any
Parameters to optimize.
lr : float
Learning rate.
momentum : float, optional
Momentum factor, by default 0.9
dampening : float, optional
Dampening for momentum, by default 0
weight_decay : float, optional
Weight decay (L2 penalty), by default 0.9
nesterov : bool, optional
Enables Nesterov momentum, by default False
trust_coefficient : float, optional
Trust coefficient for computing learning rate, by default 0.001
eps : float, optional
Eps for division denominator, by default 1e-8
"""
if lr <= 0.0:
raise ValueError(f"Invalid learning rate: {lr}")
if momentum < 0.0:
raise ValueError(f"Invalid momentum value: {momentum}")
if weight_decay < 0.0:
raise ValueError(f"Invalid weight_decay value: {weight_decay}")
defaults = dict(
lr=lr,
momentum=momentum,
dampening=dampening,
weight_decay=weight_decay,
nesterov=nesterov,
trust_coefficient=trust_coefficient,
eps=eps,
)
if nesterov and (momentum <= 0 or dampening != 0):
raise ValueError("Nesterov momentum requires a momentum and zero dampening")
super().__init__(params, defaults)
[docs]
def __setstate__(self, state: Dict[str, Any]) -> None:
super().__setstate__(state)
for group in self.param_groups:
group.setdefault("nesterov", False)
# Type ignore for overloads is required for Python 3.7.
@overload # type: ignore[override]
def step(self, closure: None = None) -> None: ...
@overload
def step(self, closure: Callable[[], float]) -> float: ...
[docs]
@torch.no_grad()
def step(self, closure: Optional[Callable[[], float]] = None) -> Optional[float]:
"""Performs a single optimization step.
Args:
closure (callable, optional): A closure that reevaluates the model
and returns the loss.
"""
loss = None
if closure is not None:
with torch.enable_grad():
loss = closure()
# Exclude scaling for params with 0 weight decay.
for group in self.param_groups:
weight_decay = group["weight_decay"]
momentum = group["momentum"]
dampening = group["dampening"]
nesterov = group["nesterov"]
for p in group["params"]:
if p.grad is None:
continue
d_p = p.grad
p_norm = torch.norm(p.data)
g_norm = torch.norm(p.grad.data)
# Apply Lars scaling and weight decay.
if weight_decay != 0:
if p_norm != 0 and g_norm != 0:
lars_lr = p_norm / (
g_norm + p_norm * weight_decay + group["eps"]
)
lars_lr *= group["trust_coefficient"]
d_p = d_p.add(p, alpha=weight_decay)
d_p *= lars_lr
# Apply momentum.
if momentum != 0:
param_state = self.state[p]
if "momentum_buffer" not in param_state:
buf = param_state["momentum_buffer"] = torch.clone(d_p).detach()
else:
buf = param_state["momentum_buffer"]
buf.mul_(momentum).add_(d_p, alpha=1 - dampening)
if nesterov:
d_p = d_p.add(buf, alpha=momentum)
else:
d_p = buf
p.add_(d_p, alpha=-group["lr"])
return loss