-
Notifications
You must be signed in to change notification settings - Fork 58
/
results2xlsx.py
95 lines (67 loc) · 2.53 KB
/
results2xlsx.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
92
93
94
95
from __future__ import absolute_import, division, print_function
import os
import argparse
import yaml
import numpy as np
import openpyxl
from openpyxl import Workbook
from utils.core_utils import load_meta
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.add_argument('--folder', default='results', type=str)
parser.add_argument('--del_empty_dir', action='store_true')
args = parser.parse_args()
metas = {}
for subdir, dirs, files in os.walk(args.folder):
if len(dirs):
continue
if not len(files):
if args.del_empty_dir:
print('deleting folder %s' % subdir)
os.rmdir(os.path.abspath(subdir))
if 'model.h5' not in files:
print('model.h5 not found in %s' % subdir)
continue
try:
meta = load_meta(os.path.join(subdir, 'model.h5'))
metas[subdir.split(os.sep)[-1]] = meta
except KeyError:
print('meta not found in %s' % os.path.join(subdir, 'model.h5'))
training_args = list(set([arg for model in metas for arg in
metas[model]['training_args']]))
datasets = {}
for model in metas:
args = metas[model]['training_args']
meta = metas[model]
try:
key = args['dataset']
if type(key) in (list, set):
key = key[0]
key = key.split(os.sep)[-2]
except KeyError:
key = 'unknown'
if key not in datasets:
datasets[key] = {}
datasets[key][model] = meta
wb = Workbook()
columns = ['path'] + ['epoch', 'best_val_ler'] + training_args
for name in datasets:
ws = wb.create_sheet(name)
cell_range = ws['A1':'%s1'
% openpyxl.utils.get_column_letter(len(columns))][0]
for i, cell in zip(range(len(cell_range)), cell_range):
cell.value = columns[i]
for row, (model, meta) in enumerate(datasets[name].items(), start=2):
ws['A%d' % row] = model
for key in ('epoch', 'epochs'):
if key in meta:
ws['B%d' % row] = meta[key][np.argmin(meta['val_decoder_ler'])]
break
ws['C%d' % row] = np.min(meta['val_decoder_ler'])
for arg, val in meta['training_args'].items():
col = openpyxl.utils.get_column_letter(
training_args.index(arg) + 4)
if type(val) in (list, set):
val = ', '.join(val)
ws['%s%d' % (col, row)] = val
wb.save('results.xlsx')