| """RAdam Optimizer. |
| Implementation lifted from: https://github.com/LiyuanLucasLiu/RAdam |
| Paper: `On the Variance of the Adaptive Learning Rate and Beyond` - https://arxiv.org/abs/1908.03265 |
| |
| NOTE: This impl has been deprecated in favour of torch.optim.RAdam and remains as a reference |
| """ |
| import math |
| import torch |
| from torch.optim.optimizer import Optimizer |
|
|
|
|
| class RAdamLegacy(Optimizer): |
| """ PyTorch RAdam optimizer |
| |
| NOTE: This impl has been deprecated in favour of torch.optim.AdamW and remains as a reference |
| """ |
| def __init__( |
| self, |
| params, |
| lr=1e-3, |
| betas=(0.9, 0.999), |
| eps=1e-8, |
| weight_decay=0, |
| ): |
| defaults = dict( |
| lr=lr, |
| betas=betas, |
| eps=eps, |
| weight_decay=weight_decay, |
| buffer=[[None, None, None] for _ in range(10)] |
| ) |
| super(RAdamLegacy, self).__init__(params, defaults) |
|
|
| def __setstate__(self, state): |
| super(RAdamLegacy, self).__setstate__(state) |
|
|
| @torch.no_grad() |
| def step(self, closure=None): |
| loss = None |
| if closure is not None: |
| with torch.enable_grad(): |
| loss = closure() |
|
|
| for group in self.param_groups: |
|
|
| for p in group['params']: |
| if p.grad is None: |
| continue |
| grad = p.grad.float() |
| if grad.is_sparse: |
| raise RuntimeError('RAdam does not support sparse gradients') |
|
|
| p_fp32 = p.float() |
|
|
| state = self.state[p] |
|
|
| if len(state) == 0: |
| state['step'] = 0 |
| state['exp_avg'] = torch.zeros_like(p_fp32) |
| state['exp_avg_sq'] = torch.zeros_like(p_fp32) |
| else: |
| state['exp_avg'] = state['exp_avg'].type_as(p_fp32) |
| state['exp_avg_sq'] = state['exp_avg_sq'].type_as(p_fp32) |
|
|
| exp_avg, exp_avg_sq = state['exp_avg'], state['exp_avg_sq'] |
| beta1, beta2 = group['betas'] |
|
|
| exp_avg_sq.mul_(beta2).addcmul_(grad, grad, value=1 - beta2) |
| exp_avg.mul_(beta1).add_(grad, alpha=1 - beta1) |
|
|
| state['step'] += 1 |
| buffered = group['buffer'][int(state['step'] % 10)] |
| if state['step'] == buffered[0]: |
| num_sma, step_size = buffered[1], buffered[2] |
| else: |
| buffered[0] = state['step'] |
| beta2_t = beta2 ** state['step'] |
| num_sma_max = 2 / (1 - beta2) - 1 |
| num_sma = num_sma_max - 2 * state['step'] * beta2_t / (1 - beta2_t) |
| buffered[1] = num_sma |
|
|
| |
| if num_sma >= 5: |
| step_size = group['lr'] * math.sqrt( |
| (1 - beta2_t) * |
| (num_sma - 4) / (num_sma_max - 4) * |
| (num_sma - 2) / num_sma * |
| num_sma_max / (num_sma_max - 2)) / (1 - beta1 ** state['step']) |
| else: |
| step_size = group['lr'] / (1 - beta1 ** state['step']) |
| buffered[2] = step_size |
|
|
| if group['weight_decay'] != 0: |
| p_fp32.add_(p_fp32, alpha=-group['weight_decay'] * group['lr']) |
|
|
| |
| if num_sma >= 5: |
| denom = exp_avg_sq.sqrt().add_(group['eps']) |
| p_fp32.addcdiv_(exp_avg, denom, value=-step_size) |
| else: |
| p_fp32.add_(exp_avg, alpha=-step_size) |
|
|
| p.copy_(p_fp32) |
|
|
| return loss |
|
|
