-
Notifications
You must be signed in to change notification settings - Fork 9
Expand file tree
/
Copy pathconfigs.py
More file actions
132 lines (122 loc) · 4.31 KB
/
Copy pathconfigs.py
File metadata and controls
132 lines (122 loc) · 4.31 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
def get_java_config():
conf = {
'workdir': './java/',
# data_params
# training data
'train_name': 'train.methname.h5',
'train_api': 'train.apiseq.h5',
'train_tokens': 'train.tokens.h5',
'train_desc': 'train.desc.h5',
# test data
'valid_name': 'test.methname.h5',
'valid_api': 'test.apiseq.h5',
'valid_tokens': 'test.tokens.h5',
'valid_desc': 'test.desc.h5',
# use data (computing code vectors)
'use_codebase': 'use.rawcode.txt', # 'use.rawcode.h5'
'use_names': 'use.methname.h5',
'use_apis': 'use.apiseq.h5',
'use_tokens': 'use.tokens.h5',
# results data(code vectors)
'use_codevecs': 'use.codevecs.normalized.h5', # 'use.codevecs.h5',
# parameters
'name_len': 6,
'api_len': 30,
'tokens_len': 50,
'desc_len': 30,
'n_words': 10000, # len(vocabulary) + 1
# vocabulary info
'vocab_name': 'vocab.methname.pkl',
'vocab_api': 'vocab.apiseq.pkl',
'vocab_tokens': 'vocab.tokens.pkl',
'vocab_desc': 'vocab.desc.pkl',
# training_params
'batch_size': 12,
'chunk_size': 100000,
'nb_epoch': 2000,
'validation_split': 0.2,
# 'optimizer': 'adam',
'lr': 0.001,
'valid_every': 10,
'n_eval': 100,
'evaluate_all_threshold': {
'mode': 'all',
'top1': 0.4,
},
'log_every': 100,
'save_every': 10,
'reload': -1,
# 970,#epoch that the model is reloaded from . If reload=0, then train from scratch
# model_params
'emb_size': 100,
'n_hidden': 400, # number of hidden dimension of code/desc representation
# recurrent
'lstm_dims': 200, # * 2
'init_embed_weights_methname': None, # 'word2vec_100_methname.h5',
'init_embed_weights_tokens': None, # 'word2vec_100_tokens.h5',
'init_embed_weights_desc': None, # 'word2vec_100_desc.h5',
'margin': 0.05,
'sim_measure': 'cos', # similarity measure: gesd, cosine, aesd
}
return conf
def get_python_config():
conf = {
'workdir': './python/',
# data_params
# training data
'train_name': 'train.methname.npy',
'train_api': 'train.apiseq.npy',
'train_tokens': 'train.tokens.npy',
'train_desc': 'train.desc.npy',
# test data
'valid_name': 'small.test.methname.npy',
'valid_api': 'small.test.apiseq.npy',
'valid_tokens': 'small.test.tokens.npy',
'valid_desc': 'small.test.desc.npy',
# use data (computing code vectors)
'use_codebase': 'full.rawcode.txt', # 'use.rawcode.h5'
'use_names': 'train.methname.npy',
'use_apis': 'train.apiseq.npy',
'use_tokens': 'train.apiseq.npy',
# results data(code vectors)
'use_codevecs': 'use.codevecs.normalized.h5', # 'use.codevecs.h5',
# parameters
'name_len': 5,
'api_len': 45,
'tokens_len': 55,
'desc_len': 15,
'n_words': 10002, # len(vocabulary) + 1
# vocabulary info
'vocab_name': 'vocab.methname.pkl',
'vocab_api': 'vocab.apiseq.pkl',
'vocab_tokens': 'vocab.tokens.pkl',
'vocab_desc': 'vocab.desc.pkl',
# training_params
'batch_size': 64,
'chunk_size': 100000,
'nb_epoch': 50,
'validation_split': 0.2,
# 'optimizer': 'adam',
'lr': 0.001,
'valid_every': 10,
'n_eval': 100,
'evaluate_all_threshold': {
'mode': 'all',
'top1': 0.4,
},
'log_every': 100,
'save_every': 3,
'reload': -1,
# 970,#epoch that the model is reloaded from . If reload=0, then train from scratch
# model_params
'emb_size': 100,
'n_hidden': 400, # number of hidden dimension of code/desc representation
# recurrent
'lstm_dims': 200, # * 2
'init_embed_weights_methname': None, # 'word2vec_100_methname.h5',
'init_embed_weights_tokens': None, # 'word2vec_100_tokens.h5',
'init_embed_weights_desc': None, # 'word2vec_100_desc.h5',
'margin': 0.05,
'sim_measure': 'cos', # similarity measure: gesd, cosine, aesd
}
return conf