-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathsolution.py
201 lines (177 loc) · 7.49 KB
/
solution.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
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
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
import random
from collections import defaultdict
from flask import Flask, request, jsonify
import numpy as np
from utils import *
from solvers import *
import re
import traceback
import time
start_loop1 = time.time()
solver_param = defaultdict(dict)
solver_param[17]["train_size"] = 0.9
solver_param[18]["train_size"] = 0.85
solver_param[19]["train_size"] = 0.85
solver_param[20]["train_size"] = 0.85
class CuttingEdgeStrongGeneralAI(object):
def __init__(self, train_path='public_set/train'):
self.train_path = train_path
self.classifier = classifier.Solver()
solver_classes = [
solver1,
solver2,
solver3,
solver4,
solver5,
solver6,
solver7,
solver8,
solver9,
solver10,
solver10,
solver10,
solver13,
solver14,
solver15,
solver16,
solver17,
solver17,
solver17,
solver17,
solver21,
solver22,
solver23,
solver24,
solver25,
solver26
]
#solver_classes = [solver4]
self.solvers = self.solver_loading(solver_classes)
self.clf_fitting()
def solver_loading(self, solver_classes):
solvers = []
start_loop = time.time()
for i, solver_class in enumerate(solver_classes):
solver_index = i + 1
train_tasks = load_tasks(self.train_path, task_num=solver_index)
solver_path = os.path.join("data", "models", "solver{}.pkl".format(solver_index))
solver = solver_class.Solver(**solver_param[solver_index])
if os.path.exists(solver_path):
print("Loading Solver {}".format(solver_index))
solver.load(solver_path)
else:
print("Fitting Solver {}...".format(solver_index))
try:
print("Fitting Solver")
#solver = solver_class.Solver(**solver_param[solver_index])
solver.fit(train_tasks)
solver.save(solver_path)
except Exception as e:
print('Exception during fitting: {}'.format(e))
print("Solver {} is ready!\n".format(solver_index))
solvers.append(solver)
print(time.time() - start_loop)
return solvers
def clf_fitting(self):
tasks = []
for filename in os.listdir(self.train_path):
if filename.endswith(".json"):
data = read_config(os.path.join(self.train_path, filename))
tasks.append(data)
print("Fitting Classifier...")
self.classifier.fit(tasks)
print("Classifier is ready!")
return self
def not_so_strong_task_solver(self, task):
question = task['question']
if question['type'] == 'choice':
# pick a random answer
choice = random.choice(question['choices'])
answer = choice['id']
elif question['type'] == 'multiple_choice':
# pick a random number of random choices
min_choices = question.get('min_choices', 1)
max_choices = question.get('max_choices', len(question['choices']))
n_choices = random.randint(min_choices, max_choices)
random.shuffle(question['choices'])
answer = [
choice['id']
for choice in question['choices'][:n_choices]
]
elif question['type'] == 'matching':
# match choices at random
random.shuffle(question['choices'])
answer = {
left['id']: choice['id']
for left, choice in zip(question['left'], question['choices'])
}
elif question['type'] == 'text':
if question.get('restriction') == 'word':
# pick a random word from the text
words = [word for word in task['text'].split() if len(word) > 1]
answer = random.choice(words)
else:
# random text generated with https://fish-text.ru
print("AAAAAAAAAAAAAAAAAA")
answer = (
'Для современного мира реализация намеченных плановых заданий позволяет '
'выполнить важные задания по разработке новых принципов формирования '
'материально-технической и кадровой базы. Господа, реализация намеченных '
'плановых заданий играет определяющее значение для модели развития. '
'Сложно сказать, почему сделанные на базе интернет-аналитики выводы призывают '
'нас к новым свершениям, которые, в свою очередь, должны быть в равной степени '
'предоставлены сами себе. Ясность нашей позиции очевидна: базовый вектор '
'развития однозначно фиксирует необходимость существующих финансовых и '
'административных условий.'
)
else:
raise RuntimeError('Unknown question type: {}'.format(question['type']))
return answer
def take_exam(self, exam):
answers = {}
# pprint.pprint(exam)
if "tasks" in exam:
variant = exam["tasks"]
if isinstance(variant, dict):
if "tasks" in variant.keys():
variant = variant["tasks"]
else:
variant = exam
task_number = self.classifier.predict(variant)
for i,task_n in enumerate(task_number):
if "left" in variant[i]["question"] and len(variant[i]["question"]["left"]) == 4:
task_number[i] = 26
if "left" in variant[i]["question"] and len(variant[i]["question"]["left"]) == 5:
task_number[i] = 8
#print("Classifier results: ", task_number)
for i, task in enumerate(variant):
#if (int(task_number[i])):
task_id = task['id']
task_index, task_type = i + 1, task["question"]["type"]
try:
prediction = self.solvers[task_number[i] - 1].predict_from_model(task)
print("Prediction: ", prediction)
except Exception as e:
print(traceback.format_exc())
prediction = self.not_so_strong_task_solver(task)
if isinstance(prediction, np.ndarray):
prediction = list(prediction)
answers[task_id] = prediction
#else:
# answers[task['id']] = '0'
return answers
app = Flask(__name__)
ai = CuttingEdgeStrongGeneralAI()
@app.route('/ready')
def http_ready():
return 'OK'
print(time.time() - start_loop1)
@app.route('/take_exam', methods=['POST'])
def http_take_exam():
request_data = request.get_json()
answers = ai.take_exam(request_data)
return jsonify({
'answers': answers
})
if __name__ == '__main__':
app.run(host='0.0.0.0', port=8000)