ColossalAI/colossalai/nn/optimizer/cpu_adam.py

170 lines
7.4 KiB
Python
Raw Normal View History

2022-03-17 07:05:41 +00:00
import math
from typing import Optional
import torch
from colossalai.kernel.op_builder import CPUAdamBuilder
from colossalai.registry import OPTIMIZERS
from .nvme_optimizer import NVMeOptimizer
@OPTIMIZERS.register_module
class CPUAdam(NVMeOptimizer):
"""Implements Adam algorithm.
Supports parameters updating on both GPU and CPU, depanding on the device of parameters.
But the parameters and gradients should on the same device:
* Parameters on CPU and gradients on CPU is allowed.
* Parameters on GPU and gradients on GPU is allowed.
* Parameters on GPU and gradients on CPU is **not** allowed.
`CPUAdam` requires CUDA extensions which can be built during installation or runtime.
This version of CPU Adam accelates parameters updating on CPU with SIMD.
Support of AVX2 or AVX512 is required.
The GPU part is implemented in an naive way.
CPU Adam also supports the hybrid precision calculation, eg. fp32 parameters and fp16 gradients.
:class:`colossalai.nn.optimizer.CPUAdam` may be used as a drop-in replacement for ``torch.optim.AdamW``,
or ``torch.optim.Adam`` with ``adamw_mode=False``
Adam was been proposed in `Adam: A Method for Stochastic Optimization`_.
Arguments:
model_params (iterable): iterable of parameters of 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)
amsgrad (boolean, optional): whether to use the AMSGrad variant of this
algorithm from the paper `On the Convergence of Adam and Beyond`_
(default: False) NOT SUPPORTED yet in CPUAdam!
adamw_mode (boolean, optional): Apply L2 regularization or weight decay
True for decoupled weight decay(also known as AdamW) (default: True)
simd_log (boolean, optional): whether to show if you are using SIMD to
accelerate. (default: False)
2022-08-12 10:01:02 +00:00
nvme_offload_fraction (float, optional): Fraction of optimizer states to be offloaded to NVMe. Defaults to 0.0.
nvme_offload_dir (Optional[str], optional): Directory to save NVMe offload files.
If it's ``None``, a random temporary directory will be used. Defaults to None.
2022-04-01 08:27:03 +00:00
.. _Adam\: A Method for Stochastic Optimization:
https://arxiv.org/abs/1412.6980
.. _On the Convergence of Adam and Beyond:
https://openreview.net/forum?id=ryQu7f-RZ
"""
# Number of fp32 shards for per parameter
# Param weight, grad, momentum and variance
num_fp32_shards_per_param = 4
def __init__(self,
model_params,
lr=1e-3,
bias_correction=True,
2022-03-17 07:05:41 +00:00
betas=(0.9, 0.999),
eps=1e-8,
weight_decay=0,
adamw_mode=True,
nvme_offload_fraction: float = 0.0,
nvme_offload_dir: Optional[str] = None):
2022-03-17 07:05:41 +00:00
default_args = dict(lr=lr, betas=betas, eps=eps, weight_decay=weight_decay, bias_correction=bias_correction)
super(CPUAdam, self).__init__(model_params, default_args, nvme_offload_fraction, nvme_offload_dir)
self.adamw_mode = adamw_mode
cpu_adam = CPUAdamBuilder().load()
self.cpu_adam_op = cpu_adam.CPUAdamOptimizer(lr, betas[0], betas[1], eps, weight_decay, adamw_mode)
2022-03-17 07:05:41 +00:00
def torch_adam_update(self,
data,
grad,
exp_avg,
exp_avg_sq,
lr,
beta1,
beta2,
eps,
weight_decay,
bias_correction1,
bias_correction2,
use_adamw=False):
# FIXME(ver217): remove the below line when replace torch adam with fused adam
grad = grad.float()
2022-03-17 07:05:41 +00:00
if weight_decay != 0:
if use_adamw:
data.mul_(1 - lr * weight_decay)
else:
grad = grad.add(data, alpha=weight_decay)
# 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)
# TODO(jiaruifang) dose not support amsgrad
denom = (exp_avg_sq.sqrt() / math.sqrt(bias_correction2)).add_(eps)
step_size = lr / bias_correction1
data.addcdiv_(exp_avg, denom, value=-step_size)
@torch.no_grad()
def step(self, closure=None, div_scale: float = -1):
loss = None
if closure is not None:
with torch.enable_grad():
loss = closure()
self._pre_step('exp_avg', 'exp_avg_sq')
2022-03-17 07:05:41 +00:00
for _, group in enumerate(self.param_groups):
for _, p in enumerate(group['params']):
if p.grad is None:
continue
state = self.state[p]
2022-03-17 07:05:41 +00:00
target_device = p.device
if len(state) == 0:
state['step'] = 0
# gradient momentums
state['exp_avg'] = torch.zeros_like(p, dtype=torch.float, device=target_device)
# gradient variances
state['exp_avg_sq'] = torch.zeros_like(p, dtype=torch.float, device=target_device)
self._post_state_init(p)
state['step'] += 1
beta1, beta2 = group['betas']
2022-03-17 07:05:41 +00:00
if target_device.type == 'cpu':
assert p.data.numel() == p.grad.data.numel(), "parameter and gradient should have the same size"
2022-03-17 07:05:41 +00:00
assert state['exp_avg'].device.type == 'cpu', "exp_avg should stay on cpu"
assert state['exp_avg_sq'].device.type == 'cpu', "exp_avg should stay on cpu"
self._pre_update(p, 'exp_avg', 'exp_avg_sq')
self.cpu_adam_op.step(state['step'], group['lr'], beta1, beta2, group['eps'], group['weight_decay'],
group['bias_correction'], p.data, p.grad.data, state['exp_avg'],
state['exp_avg_sq'], div_scale)
self._post_update(p, 'exp_avg', 'exp_avg_sq')
2022-03-17 07:05:41 +00:00
elif target_device.type == 'cuda':
assert div_scale == -1, "div_scale should remain default"
2022-03-17 07:05:41 +00:00
assert state['exp_avg'].device.type == 'cuda', "exp_avg should stay on cuda"
assert state['exp_avg_sq'].device.type == 'cuda', "exp_avg should stay on cuda"
bias_correction1 = 1 - beta1**state['step']
bias_correction2 = 1 - beta2**state['step']
2022-03-17 07:05:41 +00:00
# adam on cuda
self.torch_adam_update(p.data, p.grad.data, state['exp_avg'], state['exp_avg_sq'], group['lr'],
beta1, beta2, group['eps'], group['weight_decay'], bias_correction1,
bias_correction2, self.adamw_mode)
2022-03-17 07:05:41 +00:00
else:
raise RuntimeError
self._post_step()
return loss