forked from moneyDboat/data_grand
-
Notifications
You must be signed in to change notification settings - Fork 0
/
test_ensemble.py
70 lines (54 loc) · 1.71 KB
/
test_ensemble.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
# -*- coding: utf-8 -*-
"""
# @Author : captain
# @Time : 2018/8/31 20:21
# @Ide : PyCharm
"""
import torch
import models
import data
import numpy as np
import torch.nn.functional as F
import pandas as pd
from sklearn import metrics
import fire
from config import DefaultConfig
def val_ensemble():
files = []
ensmble = np.zeros([10000, 19])
for file in files:
prob = np.load(file)
ensmble += prob
result = np.argmax(ensmble, axis=1)
cal_f1(result)
def cal_f1(result):
val = pd.read_csv('/data/yujun/captain/datasets/word/val_set.csv')
val_label = val['class'].as_matrix()
val_label -= 1
f1score = np.mean(metrics.f1_score(result, val_label, average=None))
print('F1 Score: {}'.format(f1score))
def gen_test(model_path=None, device=0):
saved_model = torch.load(model_path)
config = saved_model['config']
config.device = int(device)
print('Load model from {}'.format(model_path))
_, val_iter, _, _, _ = util.load_data(config)
model = getattr(models, config.model)(config)
model.load_state_dict(saved_model['state_dict'])
torch.cuda.set_device(config.device)
model = model.cuda()
print(model)
model.eval()
probs_list = []
with torch.no_grad():
for batch in val_iter:
text, label = batch.text, batch.label
text, label = text.cuda(), label.cuda()
outputs = model(text)
probs = F.softmax(outputs, dim=1)
probs_list.append(probs.cpu().numpy())
prob_cat = np.concatenate(probs_list, axis=0)
np.save('val_{}.npy'.format(model_path), prob_cat)
print('Prob result val_{}.npy saved!'.format(model_path))
if __name__ == '__main__':
fire.Fire()