```# Copyright (c) MONAI Consortium
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
# Unless required by applicable law or agreed to in writing, software
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and

from typing import Callable, Iterable, Optional, Tuple

import torch
from torch.optim import Optimizer

"""
<https://arxiv.org/pdf/1905.11286.pdf>`_.
The code is adapted from the implementations in `Jasper for PyTorch
<https://github.com/NVIDIA/DeepLearningExamples/blob/master/PyTorch/SpeechRecognition/Jasper/common/optimizers.py>`_,

Args:
params: iterable of parameters to optimize or dicts defining parameter groups.
lr: learning rate. Defaults to 1e-3.
betas: coefficients used for computing running averages of gradient and its square. Defaults to (0.9, 0.98).
eps: term added to the denominator to improve numerical stability. Defaults to 1e-8.
weight_decay: weight decay (L2 penalty). Defaults to 0.
amsgrad: whether to use the AMSGrad variant of this algorithm from the paper
`On the Convergence of Adam and Beyond <https://arxiv.org/pdf/1904.09237.pdf>`_. Defaults to ``False``.
"""

def __init__(
self,
params: Iterable,
lr: float = 1e-3,
betas: Tuple[float, float] = (0.9, 0.98),
eps: float = 1e-8,
weight_decay: float = 0,
):
if 0.0 > lr:
raise ValueError(f"Invalid learning rate: {lr}")
if 0.0 > eps:
raise ValueError(f"Invalid epsilon value: {eps}")
if not 0.0 <= betas[0] < 1.0:
raise ValueError(f"Invalid beta parameter at index 0: {betas[0]}")
if not 0.0 <= betas[1] < 1.0:
raise ValueError(f"Invalid beta parameter at index 1: {betas[1]}")
if 0.0 > weight_decay:
raise ValueError(f"Invalid weight_decay value: {weight_decay}")
defaults = dict(
)

super().__init__(params, defaults)

def __setstate__(self, state):
super().__setstate__(state)
for group in self.param_groups:

[docs]    def step(self, closure: Optional[Callable] = None):
"""Performs a single optimization step.

Arguments:
closure: A closure that reevaluates the model and returns the loss. Defaults to ``None``.
"""
loss = None
if closure is not None:
loss = closure()

for group in self.param_groups:
for p in group["params"]:
continue
raise RuntimeError("Sparse gradients are not supported.")

state = self.state[p]

# State initialization
if len(state) == 0:
state["step"] = 0
# Exponential moving average of gradient values
state["exp_avg"] = torch.zeros_like(p.data)
# Exponential moving average of squared gradient values
state["exp_avg_sq"] = torch.zeros([]).to(state["exp_avg"].device)
# Maintains max of all exp. moving avg. of sq. grad. values
state["max_exp_avg_sq"] = torch.zeros([]).to(state["exp_avg"].device)

exp_avg, exp_avg_sq = state["exp_avg"], state["exp_avg_sq"]
max_exp_avg_sq = state["max_exp_avg_sq"]
beta1, beta2 = group["betas"]

state["step"] += 1

if exp_avg_sq == 0:
exp_avg_sq.copy_(norm)
else:

# Maintains the maximum of all 2nd moment running avg. till now
torch.max(max_exp_avg_sq, exp_avg_sq, out=max_exp_avg_sq)
# Use the max. for normalizing running avg. of gradient
else: