-
Notifications
You must be signed in to change notification settings - Fork 0
/
experiments_hm_script_count_correct_answers.py
91 lines (82 loc) · 3.15 KB
/
experiments_hm_script_count_correct_answers.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
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
from utils_data import create_calibrated_df
from utils_constants import CORRECTNESS
split = 'test'
# BERT
df = create_calibrated_df([
'output_bert_seed0_%s.csv' % split,
'output_bert_seed3_%s.csv' % split,
'output_bert_seed42_%s.csv' % split,
])
print('BERT: N. Correct = %5d; total N = %5d' % (len(df[df[CORRECTNESS]]), len(df)))
# XLNet
df = create_calibrated_df([
'output_xlnet_seed_2_%s.csv' % split,
'output_xlnet_seed_3_%s.csv' % split,
'output_xlnet_seed_4_%s.csv' % split,
])
print('XLNet: N. Correct = %5d; total N = %5d' % (len(df[df[CORRECTNESS]]), len(df)))
# DistilBERT
df = create_calibrated_df([
'output_distilbert_seed1_%s.csv' % split,
'output_distilbert_seed3_%s.csv' % split,
'output_distilbert_seed42_%s.csv' % split,
])
print('DistilBERT: N. Correct = %5d; total N = %5d' % (len(df[df[CORRECTNESS]]), len(df)))
# BERT DistilBERT
df = create_calibrated_df([
'output_bert_seed0_%s.csv' % split,
'output_bert_seed3_%s.csv' % split,
'output_bert_seed42_%s.csv' % split,
'output_distilbert_seed1_%s.csv' % split,
'output_distilbert_seed3_%s.csv' % split,
'output_distilbert_seed42_%s.csv' % split,
])
print('BERT-DistilBERT: N. Correct = %5d; total N = %5d' % (len(df[df[CORRECTNESS]]), len(df)))
# BERT XLNet
df = create_calibrated_df([
'output_bert_seed0_%s.csv' % split,
'output_bert_seed3_%s.csv' % split,
'output_bert_seed42_%s.csv' % split,
'output_xlnet_seed_2_%s.csv' % split,
'output_xlnet_seed_3_%s.csv' % split,
'output_xlnet_seed_4_%s.csv' % split,
])
print('BERT-XLNet: N. Correct = %5d; total N = %5d' % (len(df[df[CORRECTNESS]]), len(df)))
# DistilBERT XLNet
df = create_calibrated_df([
'output_distilbert_seed1_%s.csv' % split,
'output_distilbert_seed3_%s.csv' % split,
'output_distilbert_seed42_%s.csv' % split,
'output_xlnet_seed_2_%s.csv' % split,
'output_xlnet_seed_3_%s.csv' % split,
'output_xlnet_seed_4_%s.csv' % split,
])
print('DistilBERT-XLNet: N. Correct = %5d; total N = %5d' % (len(df[df[CORRECTNESS]]), len(df)))
# BERT - DistilBERT - XLNet
df = create_calibrated_df([
'output_bert_seed0_%s.csv' % split,
'output_bert_seed3_%s.csv' % split,
'output_bert_seed42_%s.csv' % split,
'output_distilbert_seed1_%s.csv' % split,
'output_distilbert_seed3_%s.csv' % split,
'output_distilbert_seed42_%s.csv' % split,
'output_xlnet_seed_2_%s.csv' % split,
'output_xlnet_seed_3_%s.csv' % split,
'output_xlnet_seed_4_%s.csv' % split,
])
print('BERT - DistilBERT - XLNet: N. Correct = %5d; total N = %5d' % (len(df[df[CORRECTNESS]]), len(df)))
# single models
model_names = [
'output_bert_seed0_%s.csv' % split,
'output_bert_seed3_%s.csv' % split,
'output_bert_seed42_%s.csv' % split,
'output_distilbert_seed1_%s.csv' % split,
'output_distilbert_seed3_%s.csv' % split,
'output_distilbert_seed42_%s.csv' % split,
'output_xlnet_seed_2_%s.csv' % split,
'output_xlnet_seed_3_%s.csv' % split,
'output_xlnet_seed_4_%s.csv' % split,
]
for model_name in model_names:
df = create_calibrated_df([model_name])
print('%s: N. Correct = %5d; total N = %5d' % (model_name, len(df[df[CORRECTNESS]]), len(df)))