[2ceedb]: / 2DNet / src / tuils / lrs_scheduler.py

Download this file

422 lines (346 with data), 17.5 kB

  1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
-------------------------------------------------
Description : lrs_scheduler
Reference:
1. https://towardsdatascience.com/transfer-learning-using-pytorch-4c3475f4495
2. https://discuss.pytorch.org/t/solved-learning-rate-decay/6825/5
3. https://discuss.pytorch.org/t/adaptive-learning-rate/320/34
4. https://github.com/pytorch/pytorch/blob/master/torch/optim/lr_scheduler.py
5. https://github.com/bckenstler/CLR
6. https://github.com/fastai/fastai/blob/master/fastai/sgdr.py
7. https://github.com/NVIDIA/nvvl/blob/master/examples/pytorch_superres/model/clr.py
Email : autuanliu@163.com
Date:2018/3/22
"""
from torch.optim import lr_scheduler
import math
from torch.optim.optimizer import Optimizer
import torch
class WarmRestart(lr_scheduler.CosineAnnealingLR):
"""This class implements Stochastic Gradient Descent with Warm Restarts(SGDR): https://arxiv.org/abs/1608.03983.
Set the learning rate of each parameter group using a cosine annealing schedule, When last_epoch=-1, sets initial lr as lr.
This can't support scheduler.step(epoch). please keep epoch=None.
"""
def __init__(self, optimizer, T_max=10, T_mult=2, eta_min=0, last_epoch=-1):
"""implements SGDR
Parameters:
----------
T_max : int
Maximum number of epochs.
T_mult : int
Multiplicative factor of T_max.
eta_min : int
Minimum learning rate. Default: 0.
last_epoch : int
The index of last epoch. Default: -1.
"""
self.T_mult = T_mult
super().__init__(optimizer, T_max, eta_min, last_epoch)
def get_lr(self):
if self.last_epoch == self.T_max:
self.last_epoch = 0
self.T_max *= self.T_mult
return [self.eta_min + (base_lr - self.eta_min) * (1 + math.cos(math.pi * self.last_epoch / self.T_max)) / 2 for base_lr in self.base_lrs]
def cyclical_lr(step_sz, min_lr=0.001, max_lr=1, mode='triangular', scale_func=None, scale_md='cycles', gamma=1.):
"""implements a cyclical learning rate policy (CLR).
The method cycles the learning rate between two boundaries with some constant frequency, as detailed in this
paper (https://arxiv.org/abs/1506.01186). The amplitude of the cycle can be scaled on a per-iteration or per-cycle basis.
This function has three built-in policies, as put forth in the paper.
Note:
-----
1. The difficulty in minimizing the loss arise from saddle rather than poor local minima(Dauphin, 2015).
2. Set stepsize equal to 2~10 times he number of iterations in an epoch.
3. It's best to stop training at the end of a cycle which is when the learning rate is at the minimum value and the accuracy peaks.(back to min learning rate at the training end)
4. LR range test: The triangular learning rate policy provides a simple mechanism to do this. Set base lr to the minimum value and set max lr to the
maximum value. Set both the stepsize and max iter to the same number of iterations. In this case, the learning rate will increase linearly from the minimum
value to the maximum value during this short run. Next, plot the accuracy versus learning rate.
Note the learning rate value when the accuracy starts to increase and when the accuracy slows, becomes ragged, or starts to fall. These two learning rates
are good choices for bounds; that is, set base lr to the first value and set max lr to the latter value. Alternatively, one can use the rule of
thumb that the optimum learning rate is usually within a factor of two of the largest one that converges and set base lr to 1/3 or 1/4 of max lr
5. The optimum learning rate will be between the bounds and near optimal learning rates will be used throughout training.
Notes: the learning rate of optimizer should be 1
Parameters:
----------
min_lr : float
lower boundary in the cycle. which is equal to the optimizer's initial learning rate.
max_lr : float
upper boundary in the cycle. Functionally, it defines the cycle amplitude (max_lr - base_lr).
step_sz : int
(2~10)*(len(datasets)/minibatch)
mode : str, optional
one of {triangular, triangular2, exp_range}. Default 'triangular'.
"triangular": A basic triangular cycle with no amplitude scaling.
"triangular2": A basic triangular cycle that scales initial amplitude by half each cycle.
"exp_range": A cycle that scales initial amplitude by gamma**(cycle iterations) at each cycle iteration.
scale_func : lambda function, optional
Custom scaling policy defined by a single argument lambda function, where 0 <= scale_fn(x) <= 1 for all x >= 0.
scale_md : str, optional
{'cycles', 'iterations'}. Defines whether scale_fn is evaluated on cycle number or cycle iterations (training
iterations since start of cycle). Default is 'cycles'.
gamma : float, optional
constant in 'exp_range' scaling function: gamma**(cycle iterations)
Returns:
--------
lambda function
Examples:
--------
>>> optimizer = optim.Adam(model.parameters(), lr=1.)
>>> step_size = 2*len(train_loader)
>>> clr = cyclical_lr(step_size, min_lr=0.001, max_lr=0.005)
>>> scheduler = lr_scheduler.LambdaLR(optimizer, [clr])
>>> # some other operations
>>> scheduler.step()
>>> optimizer.step()
"""
if scale_func == None:
if mode == 'triangular':
scale_fn = lambda x: 1.
scale_mode = 'cycles'
elif mode == 'triangular2':
scale_fn = lambda x: 1 / (2.**(x - 1))
scale_mode = 'cycles'
elif mode == 'exp_range':
scale_fn = lambda x: gamma**(x)
scale_mode = 'iterations'
else:
raise ValueError('The {} is not valid value!'.format(mode))
else:
scale_fn = scale_func
scale_mode = scale_md
lr_lambda = lambda iters: min_lr + (max_lr - min_lr) * rel_val(iters, step_sz, scale_mode)
def rel_val(iteration, stepsize, mode):
cycle = math.floor(1 + iteration / (2 * stepsize))
x = abs(iteration / stepsize - 2 * cycle + 1)
if mode == 'cycles':
return max(0, (1 - x)) * scale_fn(cycle)
elif mode == 'iterations':
return max(0, (1 - x)) * scale_fn(iteration)
else:
raise ValueError('The {} is not valid value!'.format(scale_mode))
return lr_lambda
def clr_reset(scheduler, thr):
"""learning rate scheduler reset if iteration = thr
Parameters:
----------
scheduler : instance of optim.lr_scheduler
instance of optim.lr_scheduler
thr : int
the reset point
Examples:
--------
>>> # some other operations(note the order of operations)
>>> scheduler.step()
>>> scheduler = clr_reset(scheduler, 1000)
>>> optimizer.step()
"""
if scheduler.last_epoch == thr:
scheduler.last_epoch = -1
return scheduler
def warm_restart(scheduler, T_mult=2):
"""warm restart policy
Parameters:
----------
T_mult: int
default is 2, Stochastic Gradient Descent with Warm Restarts(SGDR): https://arxiv.org/abs/1608.03983.
Examples:
--------
>>> # some other operations(note the order of operations)
>>> scheduler.step()
>>> scheduler = warm_restart(scheduler, T_mult=2)
>>> optimizer.step()
"""
if scheduler.last_epoch == scheduler.T_max:
scheduler.last_epoch = -1
scheduler.T_max *= T_mult
return scheduler
class AdamW(Optimizer):
"""Implements Adam algorithm.
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)
amsgrad (boolean, optional): whether to use the AMSGrad variant of this
algorithm from the paper `On the Convergence of Adam and Beyond`_
"""
def __init__(self, params, lr=1e-3, betas=(0.9, 0.999), eps=1e-8,
weight_decay=0, amsgrad=False):
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, amsgrad=amsgrad)
#super(AdamW, self).__init__(params, defaults)
super().__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('Adam does not support sparse gradients, please consider SparseAdam instead')
amsgrad = group['amsgrad']
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)
if amsgrad:
# Maintains max of all exp. moving avg. of sq. grad. values
state['max_exp_avg_sq'] = torch.zeros_like(p.data)
exp_avg, exp_avg_sq = state['exp_avg'], state['exp_avg_sq']
if amsgrad:
max_exp_avg_sq = state['max_exp_avg_sq']
beta1, beta2 = group['betas']
state['step'] += 1
# 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)
if amsgrad:
# 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
denom = max_exp_avg_sq.sqrt().add_(group['eps'])
else:
denom = exp_avg_sq.sqrt().add_(group['eps'])
bias_correction1 = 1 - beta1 ** state['step']
bias_correction2 = 1 - beta2 ** state['step']
step_size = group['lr'] * math.sqrt(bias_correction2) / bias_correction1
if group['weight_decay'] != 0:
decayed_weights = torch.mul(p.data, group['weight_decay'])
p.data.addcdiv_(-step_size, exp_avg, denom)
p.data.sub_(decayed_weights)
else:
p.data.addcdiv_(-step_size, exp_avg, denom)
return loss
import math
import torch
from torch.optim.optimizer import Optimizer, required
class RAdam(Optimizer):
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)
self.buffer = [[None, None, None] for ind in range(10)]
super(RAdam, self).__init__(params, defaults)
def __setstate__(self, state):
super(RAdam, self).__setstate__(state)
def step(self, closure=None):
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.float()
if grad.is_sparse:
raise RuntimeError('RAdam does not support sparse gradients')
p_data_fp32 = p.data.float()
state = self.state[p]
if len(state) == 0:
state['step'] = 0
state['exp_avg'] = torch.zeros_like(p_data_fp32)
state['exp_avg_sq'] = torch.zeros_like(p_data_fp32)
else:
state['exp_avg'] = state['exp_avg'].type_as(p_data_fp32)
state['exp_avg_sq'] = state['exp_avg_sq'].type_as(p_data_fp32)
exp_avg, exp_avg_sq = state['exp_avg'], state['exp_avg_sq']
beta1, beta2 = group['betas']
exp_avg_sq.mul_(beta2).addcmul_(1 - beta2, grad, grad)
exp_avg.mul_(beta1).add_(1 - beta1, grad)
state['step'] += 1
buffered = self.buffer[int(state['step'] % 10)]
if state['step'] == buffered[0]:
N_sma, step_size = buffered[1], buffered[2]
else:
buffered[0] = state['step']
beta2_t = beta2 ** state['step']
N_sma_max = 2 / (1 - beta2) - 1
N_sma = N_sma_max - 2 * state['step'] * beta2_t / (1 - beta2_t)
buffered[1] = N_sma
if N_sma > 5:
step_size = group['lr'] * math.sqrt((1 - beta2_t) * (N_sma - 4) / (N_sma_max - 4) * (N_sma - 2) / N_sma * N_sma_max / (N_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_data_fp32.add_(-group['weight_decay'] * group['lr'], p_data_fp32)
if N_sma > 5:
denom = exp_avg_sq.sqrt().add_(group['eps'])
p_data_fp32.addcdiv_(-step_size, exp_avg, denom)
else:
p_data_fp32.add_(-step_size, exp_avg)
p.data.copy_(p_data_fp32)
return loss
class AdamW(Optimizer):
def __init__(self, params, lr=1e-3, betas=(0.9, 0.999), eps=1e-8,
weight_decay=0, amsgrad=False, use_variance=True, warmup = 4000):
defaults = dict(lr=lr, betas=betas, eps=eps,
weight_decay=weight_decay, amsgrad=amsgrad, use_variance=True, warmup = warmup)
print('======== Warmup: {} ========='.format(warmup))
super(AdamW, self).__init__(params, defaults)
def __setstate__(self, state):
super(AdamW, self).__setstate__(state)
def step(self, closure=None):
global iter_idx
iter_idx += 1
grad_list = list()
mom_list = list()
mom_2rd_list = list()
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.float()
if grad.is_sparse:
raise RuntimeError('Adam does not support sparse gradients, please consider SparseAdam instead')
amsgrad = group['amsgrad']
p_data_fp32 = p.data.float()
state = self.state[p]
if len(state) == 0:
state['step'] = 0
state['exp_avg'] = torch.zeros_like(p_data_fp32)
state['exp_avg_sq'] = torch.zeros_like(p_data_fp32)
else:
state['exp_avg'] = state['exp_avg'].type_as(p_data_fp32)
state['exp_avg_sq'] = state['exp_avg_sq'].type_as(p_data_fp32)
exp_avg, exp_avg_sq = state['exp_avg'], state['exp_avg_sq']
beta1, beta2 = group['betas']
state['step'] += 1
exp_avg_sq.mul_(beta2).addcmul_(1 - beta2, grad, grad)
exp_avg.mul_(beta1).add_(1 - beta1, grad)
denom = exp_avg_sq.sqrt().add_(group['eps'])
bias_correction1 = 1 - beta1 ** state['step']
bias_correction2 = 1 - beta2 ** state['step']
if group['warmup'] > state['step']:
scheduled_lr = 1e-6 + state['step'] * (group['lr'] - 1e-6) / group['warmup']
# print(scheduled_lr)
else:
scheduled_lr = group['lr']
step_size = scheduled_lr * math.sqrt(bias_correction2) / bias_correction1
if group['weight_decay'] != 0:
p_data_fp32.add_(-group['weight_decay'] * scheduled_lr, p_data_fp32)
p_data_fp32.addcdiv_(-step_size, exp_avg, denom)
p.data.copy_(p_data_fp32)
return loss