Source code for common.pytorch.optim.Lion

# Copyright 2022 Cerebras Systems.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

# Copyright 2023 Google Research. All Rights Reserved.
# Modifications Copyright 2023 Cerebras.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================

from typing import Iterable, Tuple

import torch
from torch import nn

from modelzoo.common.pytorch.optim.CSOptimizer import CSOptimizer


[docs]class Lion(CSOptimizer): r"""Implements Lion algorithm. As proposed in `Symbolic Discovery of Optimization Algorithms`_. Args: params (iterable): iterable of parameters to optimize or dicts defining parameter groups lr (float, optional): learning rate (default: 1e-4) betas (Tuple[float, float], optional): coefficients used for computing running averages of gradient and its square (default: (0.9, 0.99)) weight_decay (float, optional): weight decay coefficient (default: 0) .. _Symbolic Discovery of Optimization Algorithms: https://arxiv.org/pdf/2302.06675.pdf """
[docs] def __init__( self, params: Iterable[nn.parameter.Parameter], lr: float = 1e-4, betas: Tuple[float, float] = (0.9, 0.99), weight_decay: float = 0.0, ): if not 0.0 <= lr: raise ValueError(f"Invalid learning rate: {lr}") 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 not 0.0 <= weight_decay: raise ValueError(f"Invalid weight decay value: {weight_decay}") defaults = dict(lr=lr, betas=betas, weight_decay=weight_decay) super().__init__(params, defaults)
[docs] def state_names_to_sparsify(self): return ["exp_avg"]
[docs] def preinitialize(self): """ Allocates tensors for the optimizer state to allow direct compilation of the model before the first step. """ for group in self.param_groups: for p in group["params"]: self.state[p]["exp_avg"] = torch.zeros_like(p, device="cpu").to( p.device )
@torch.no_grad() def step(self, closure=None): """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() for group in self.param_groups: lr = group["lr"] beta1, beta2 = group["betas"] weight_decay = group["weight_decay"] for p in group["params"]: if p.grad is not None: grad = p.grad state = self.state[p] exp_avg = state['exp_avg'] # Perform weight decay if weight_decay != 0: p.data.mul_(1 - lr * weight_decay) # Perform weight update update = ( exp_avg.clone() .mul_(beta1) .add(grad, alpha=1 - beta1) .sign_() ) p.add_(-lr * update) # Update exponential moving average exp_avg.mul_(beta2).add_(grad, alpha=1 - beta2) return loss