-
Notifications
You must be signed in to change notification settings - Fork 0
/
optimizer.py
52 lines (38 loc) · 1.56 KB
/
optimizer.py
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
import numpy as np
from torch.optim.lr_scheduler import _LRScheduler
import torch
class NoamLR(_LRScheduler):
def __init__(self, optimizer, step_size, d_model, n_warmup_steps=2500):
self.step_size = step_size
self.d_model = d_model
self.n_warmup_steps = n_warmup_steps
self.step_count = 0
super(NoamLR, self).__init__(optimizer)
def step(self):
self.step_count += 1
new_lr = np.power(self.d_model, -0.5) * np.min([
np.power(self.step_count, -0.5),
np.power(self.n_warmup_steps, -1.5) * self.step_count])
for param_group in self.optimizer.param_groups:
param_group['lr'] = new_lr
class ScheduledOptim(object):
'''A simple wrapper class for learning rate scheduling'''
def __init__(self, optimizer, d_model, n_warmup_steps):
self.optimizer = optimizer
self.d_model = d_model
self.n_warmup_steps = n_warmup_steps
self.n_current_steps = 0
def step(self):
"Step by the inner optimizer"
self.optimizer.step()
def zero_grad(self):
"Zero out the gradients by the inner optimizer"
self.optimizer.zero_grad()
def update_learning_rate(self):
''' Learning rate scheduling per step '''
self.n_current_steps += 1
new_lr = np.power(self.d_model, -0.5) * np.min([
np.power(self.n_current_steps, -0.5),
np.power(self.n_warmup_steps, -1.5) * self.n_current_steps])
for param_group in self.optimizer.param_groups:
param_group['lr'] = new_lr