Add LAMB optimizer

Summary: Pull Request resolved: https://github.com/fairinternal/fairseq-py/pull/572

Differential Revision: D15317928

Pulled By: myleott

fbshipit-source-id: b3f0e9229737a63b49937e7c5b918470f18ddc45

fairseq/optim/lamb.py

+# Copyright (c) 2017-present, Facebook, Inc.
+# All rights reserved.
+#
+# This source code is licensed under the license found in the LICENSE file in
+# the root directory of this source tree. An additional grant of patent rights
+# can be found in the PATENTS file in the same directory.
+"""
+LAMB optimizer from github.com/cybertronai/pytorch-lamb.
+"""
+
+import math
+
+import torch
+import torch.optim
+
+from . import FairseqOptimizer, register_optimizer
+
+
+@register_optimizer('lamb')
+class FairseqLamb(FairseqOptimizer):
+
+    def __init__(self, args, params):
+        super().__init__(args, params)
+        self._optimizer = Lamb(params, **self.optimizer_config)
+
+    @staticmethod
+    def add_args(parser):
+        """Add optimizer-specific arguments to the parser."""
+        # fmt: off
+        parser.add_argument('--lamb-betas', default='(0.9, 0.999)', metavar='B',
+                            help='betas for Adam optimizer')
+        parser.add_argument('--lamb-eps', type=float, default=1e-8, metavar='D',
+                            help='epsilon for Adam optimizer')
+        parser.add_argument('--weight-decay', '--wd', default=0.0, type=float, metavar='WD',
+                            help='weight decay')
+        # fmt: on
+
+    @property
+    def optimizer_config(self):
+        """
+        Return a kwarg dictionary that will be used to override optimizer
+        args stored in checkpoints. This allows us to load a checkpoint and
+        resume training using a different set of optimizer args, e.g., with a
+        different learning rate.
+        """
+        return {
+            'lr': self.args.lr[0],
+            'betas': eval(self.args.lamb_betas),
+            'eps': self.args.lamb_eps,
+            'weight_decay': self.args.weight_decay,
+        }
+
+
+class Lamb(torch.optim.Optimizer):
+    r"""Implements Lamb algorithm.
+    It has been proposed in `Reducing BERT Pre-Training Time from 3 Days to 76 Minutes`_.
+    Arguments:
+        params (iterable): iterable of parameters to optimize or dicts defining
+            parameter groups
+        lr (float, optional): learning rate (default: 1e-3)
+        betas (Tuple[float, float], optional): coefficients used for computing
+            running averages of gradient and its square (default: (0.9, 0.999))
+        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)
+        adam (bool, optional): always use trust ratio = 1, which turns this into
+            Adam. Useful for comparison purposes.
+    .. _Reducing BERT Pre-Training Time from 3 Days to 76 Minutes:
+        https://arxiv.org/abs/1904.00962
+    """
+
+    def __init__(self, params, lr=1e-3, betas=(0.9, 0.999), eps=1e-8,
+                 weight_decay=0, adam=False):
+        if not 0.0 <= lr:
+            raise ValueError("Invalid learning rate: {}".format(lr))
+        if not 0.0 <= eps:
+            raise ValueError("Invalid epsilon value: {}".format(eps))
+        if not 0.0 <= betas[0] < 1.0:
+            raise ValueError("Invalid beta parameter at index 0: {}".format(betas[0]))
+        if not 0.0 <= betas[1] < 1.0:
+            raise ValueError("Invalid beta parameter at index 1: {}".format(betas[1]))
+        defaults = dict(lr=lr, betas=betas, eps=eps,
+                        weight_decay=weight_decay)
+        self.adam = adam
+        super(Lamb, self).__init__(params, defaults)
+
+    def step(self, closure=None):
+        """Performs a single optimization step.
+        Arguments:
+            closure (callable, optional): A closure that reevaluates the model
+                and returns the loss.
+        """
+        loss = None
+        if closure is not None:
+            loss = closure()
+
+        for group in self.param_groups:
+            for p in group['params']:
+                if p.grad is None:
+                    continue
+                grad = p.grad.data
+                if grad.is_sparse:
+                    raise RuntimeError('Lamb does not support sparse gradients, consider SparseAdam instad.')
+
+                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_like(p.data)
+
+                exp_avg, exp_avg_sq = state['exp_avg'], state['exp_avg_sq']
+                beta1, beta2 = group['betas']
+
+                state['step'] += 1
+
+                if group['weight_decay'] != 0:
+                    grad.add_(group['weight_decay'], p.data)
+
+                # Decay the first and second moment running average coefficient
+                exp_avg.mul_(beta1).add_(1 - beta1, grad)
+                exp_avg_sq.mul_(beta2).addcmul_(1 - beta2, grad, grad)
+                denom = exp_avg_sq.sqrt().add_(group['eps'])
+
+                bias_correction1 = 1 - beta1 ** state['step']
+                bias_correction2 = 1 - beta2 ** state['step']
+                # Apply bias to lr to avoid broadcast.
+                step_size = group['lr'] * math.sqrt(bias_correction2) / bias_correction1
+
+                adam_step = exp_avg / denom
+                # L2 norm uses sum, but here since we're dividing, use mean to avoid overflow.
+                r1 = p.data.pow(2).mean().sqrt()
+                r2 = adam_step.pow(2).mean().sqrt()
+                r = 1 if r1 == 0 or r2 == 0 else  min(r1/r2, 10)
+                state['r1'] = r1
+                state['r2'] = r2
+                state['r'] = r
+                if self.adam:
+                    r = 1
+
+                p.data.add_(-step_size * r, adam_step)
+
+        return loss