# type: ignore
# from https://github.com/nanowell/AdEMAMix-Optimizer-Pytorch. Thanks a lot!
import math
import torch
from torch.optim import Optimizer
[docs]
class AdEMAMix(Optimizer):
"""Implements the AdEMAMix optimizer.
This optimizer is based on the paper "AdEMAMix: A Novel Adaptive Optimizer for
Deep Learning". It combines the advantages of Adam and EMA, and introduces a
mixing term to further improve performance.
Args:
params (iterable): Iterable of parameters to optimize or dicts defining
parameter groups.
lr (float, optional): Learning rate (default: 1e-3).
betas (Tuple[float, float, float], optional): Coefficients used for
computing running averages of gradient and its square
(default: (0.9, 0.999, 0.9999)).
eps (float, optional): Term added to the denominator to improve
numerical stability (default: 1e-8).
weight_decay (float, optional): Weight decay (L2 penalty) (default: 0).
alpha (float, optional): Mixing coefficient (default: 5.0).
T_alpha_beta3 (int, optional): Time period for alpha and beta3 scheduling
(default: None).
"""
def __init__(
self,
params={},
lr=1e-3,
betas=(0.9, 0.999, 0.9999),
eps=1e-8,
weight_decay=0,
alpha=5.0,
T_alpha_beta3=None,
):
if not 0.0 <= lr:
raise ValueError(f"Invalid learning rate: {lr}")
if not 0.0 <= eps:
raise ValueError(f"Invalid epsilon value: {eps}")
assert len(betas) == 3, f"Invalid beta parameters: {betas}, expected 3"
assert all(
0.0 <= beta < 1.0 for beta in betas
), f"Invalid beta parameters: {betas}"
if not 0.0 <= weight_decay:
raise ValueError(f"Invalid weight_decay value: {weight_decay}")
defaults = dict(
lr=lr,
betas=betas,
eps=eps,
weight_decay=weight_decay,
alpha=alpha,
T_alpha_beta3=T_alpha_beta3,
)
super(AdEMAMix, self).__init__(params, defaults)
[docs]
def __setstate__(self, state):
"""Set the state of the optimizer.
Args:
state (dict): The state of the optimizer.
"""
super(AdEMAMix, self).__setstate__(state)
[docs]
@torch.no_grad()
def step(self, closure=None):
"""Perform a single optimization step.
Args:
closure (callable, optional): A closure that reevaluates the model
and returns the loss. (default: None)
Returns:
The loss, if the closure is provided. Otherwise, returns None.
"""
loss = None
if closure is not None:
with torch.enable_grad():
loss = closure()
for group in self.param_groups:
params_with_grad = []
grads = []
exp_avgs = []
exp_avg_sqs = []
exp_avg_slow = []
state_steps = []
for p in group["params"]:
if p.grad is not None:
params_with_grad.append(p)
if p.grad.is_sparse:
raise RuntimeError("AdEMAMix does not support sparse gradients")
grads.append(p.grad)
state = self.state[p]
# Lazy state initialization
if len(state) == 0:
state["step"] = 0
# Exponential moving average of gradient values
state["exp_avg"] = torch.zeros_like(
p, memory_format=torch.preserve_format
)
# Exponential moving average of squared gradient values
state["exp_avg_sq"] = torch.zeros_like(
p, memory_format=torch.preserve_format
)
# Slow exponential moving average
state["exp_avg_slow"] = torch.zeros_like(
p, memory_format=torch.preserve_format
)
exp_avgs.append(state["exp_avg"])
exp_avg_sqs.append(state["exp_avg_sq"])
exp_avg_slow.append(state["exp_avg_slow"])
state["step"] += 1
state_steps.append(state["step"])
beta1, beta2, beta3 = group["betas"]
alpha = group["alpha"]
T_alpha_beta3 = group["T_alpha_beta3"]
self._update_adamemix(
params_with_grad,
grads,
exp_avgs,
exp_avg_sqs,
exp_avg_slow,
state_steps,
beta1=beta1,
beta2=beta2,
beta3=beta3,
alpha=alpha,
T_alpha_beta3=T_alpha_beta3,
lr=group["lr"],
weight_decay=group["weight_decay"],
eps=group["eps"],
)
return loss
def _update_adamemix(
self,
params,
grads,
exp_avgs,
exp_avg_sqs,
exp_avg_slow,
state_steps,
beta1,
beta2,
beta3,
alpha,
T_alpha_beta3,
lr,
weight_decay,
eps,
):
"""Perform the AdEMAMix update for a single parameter group.
Args:
params (list[torch.Tensor]): List of parameters to update.
grads (list[torch.Tensor]): List of gradients for each parameter.
exp_avgs (list[torch.Tensor]): List of exponential moving averages of
gradients.
exp_avg_sqs (list[torch.Tensor]): List of exponential moving averages
of squared gradients.
exp_avg_slow (list[torch.Tensor]): List of slow exponential moving
averages of gradients.
state_steps (list[int]): List of steps for each parameter.
beta1 (float): Coefficient for the first moment estimate.
beta2 (float): Coefficient for the second moment estimate.
beta3 (float): Coefficient for the slow moment estimate.
alpha (float): Mixing coefficient.
T_alpha_beta3 (int): Time period for alpha and beta3 scheduling.
lr (float): Learning rate.
weight_decay (float): Weight decay.
eps (float): Epsilon term for numerical stability.
"""
for i, param in enumerate(params):
grad = grads[i]
exp_avg = exp_avgs[i]
exp_avg_sq = exp_avg_sqs[i]
exp_avg_slow_i = exp_avg_slow[i]
step = state_steps[i]
bias_correction1 = 1 - beta1**step
bias_correction2 = 1 - beta2**step
if T_alpha_beta3 is not None:
alpha_t = min(step * alpha / T_alpha_beta3, alpha)
beta3_t = min(
math.exp(
math.log(beta1)
* math.log(beta3)
/ (
(1 - step / T_alpha_beta3) * math.log(beta3)
+ (step / T_alpha_beta3) * math.log(beta1)
)
),
beta3,
)
else:
alpha_t = alpha
beta3_t = beta3
# Decay the first and second moment running average coefficient
exp_avg.mul_(beta1).add_(grad, alpha=1 - beta1)
exp_avg_sq.mul_(beta2).addcmul_(grad, grad, value=1 - beta2)
exp_avg_slow_i.mul_(beta3_t).add_(grad, alpha=1 - beta3_t)
denom = (exp_avg_sq.sqrt() / math.sqrt(bias_correction2)).add_(eps)
step_size = lr / bias_correction1
if weight_decay != 0:
param.add_(param, alpha=-weight_decay * lr)
param.addcdiv_(exp_avg + alpha_t * exp_avg_slow_i, denom, value=-step_size)