-
Notifications
You must be signed in to change notification settings - Fork 21
/
config.py
55 lines (44 loc) · 1.42 KB
/
config.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
51
52
53
54
55
import time
import torch
cuda_condition = torch.cuda.is_available()
device = torch.device('cuda:0' if cuda_condition else 'cpu')
VocabPath = 'checkpoint/pretrain/vocab.txt'
# ## mlm模型文件路径 ## #
SourcePath = 'data/src_data/src_data.txt'
CorpusPath = 'data/train_data/train.txt'
TestPath = 'data/test_data/test.txt'
PronunciationPath = 'data/char_meta.txt'
# Debug开关
Debug = False
# attention_mask开关
AttentionMask = True
# 使用预训练模型开关
# 目前未配置预外部训练模型
UsePretrain = True
# mask方式
# True表示mask时会mask每个字,而且对于出现频率较低的字会多次mask,频率的限制由WordGenTimes决定
# False表示按照bert的mask方式进行mask
AllMask = True
# ## MLM训练调试参数开始 ## #
MLMEpochs = 16
# 设置低频字的阈值,出现次数低于该频次的字将会被重复制作成WordGenTimes个样本
WordGenTimes = 3
if WordGenTimes > 1:
RanWrongDivisor = 1.0
else:
RanWrongDivisor = 0.15
MLMLearningRate = 1e-4
BatchSize = 16
SentenceLength = 512
FinetunePath = 'checkpoint/finetune/mlm_trained_%s.model' % SentenceLength
# ## MLM训练调试参数结束 ## #
# ## MLM通用参数 ## #
DropOut = 0.1
MaskRate = 0.15
VocabSize = len(open(VocabPath, 'r', encoding='utf-8').readlines())
HiddenSize = 768
HiddenLayerNum = 12
IntermediateSize = 3072
AttentionHeadNum = 12
def get_time():
return time.strftime("%Y-%m-%d %H:%M:%S", time.localtime())