-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathgenerate_questions.py
175 lines (151 loc) · 6.57 KB
/
generate_questions.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
from transformers import (WEIGHTS_NAME, GPT2Config, GPT2Tokenizer)
from generative_qg import GenerativeGPT2QG, GenerativeGPT2QGWrapper
from generative_qa import GenerativeGPT2QA2, GenerativeGPT2QD, GenerativeGPT2QA2Wrapper, GenerativeGPT2QDWrapper
import random
import sys
import csv
import numpy as np
import torch
from tqdm import tqdm, trange
from math import exp
import argparse
parser = argparse.ArgumentParser()
parser.add_argument(
"--name",
default="train_fake_1000000",
type=str,
help=
"The input data dir. Should contain the .tsv files (or other data files) for the task."
)
parser.add_argument(
"--epochs",
default=3125,
type=int,
help=
"The input data dir. Should contain the .tsv files (or other data files) for the task."
)
args = parser.parse_args()
dir = "/scratch/yyv959/commonsenseqa/"
qg_model_path = "/scratch/yyv959/commonsenseqa/outputs/gpt2-large/qg/"
qg_model = GenerativeGPT2QG.from_pretrained(qg_model_path)
qg_tokenizer = GPT2Tokenizer.from_pretrained(qg_model_path)
qg_model.add_tokenizer(qg_tokenizer)
qa_model_path = "/scratch/yyv959/commonsenseqa/outputs/gpt2-large/qa-v2/"
qa_model = GenerativeGPT2QA2.from_pretrained(qa_model_path)
qa_tokenizer = GPT2Tokenizer.from_pretrained(qa_model_path)
qa_model.add_tokenizer(qa_tokenizer)
qd_model_path = "/scratch/yyv959/commonsenseqa/outputs/gpt2-large/qd/"
qd_model = GenerativeGPT2QD.from_pretrained(qd_model_path)
qd_tokenizer = GPT2Tokenizer.from_pretrained(qd_model_path)
qd_model.add_tokenizer(qd_tokenizer)
distractor_size = 4
data = []
#qg_model.cuda()
qg_model = GenerativeGPT2QGWrapper(qg_model)
qa_model = GenerativeGPT2QA2Wrapper(qa_model)
qd_model = GenerativeGPT2QDWrapper(qd_model)
qg_model.eval()
qa_model.eval()
qd_model.eval()
#qa_model = torch.nn.DataParallel(qa_model)
#qd_model = torch.nn.DataParallel(qd_model)
#qg_model = torch.nn.DataParallel(qg_model)
qa_model.cuda()
qd_model.cuda()
qg_model.cuda()
with open(dir + args.name + ".csv",
'w',
encoding='utf8',
newline='') as tsv_file:
tsv_writer = csv.writer(tsv_file, delimiter=',', lineterminator='\n')
tsv_writer.writerow([
"id", "question", "concept", "true_answer", "wrong1", "wrong2",
"wrong3", "wrong4"
])
questions = []
for i in trange(args.epochs):
with torch.no_grad():
#qg_model.cuda()
questions += qg_model(80, 62, sample=True, tmp=1)
#qg_model.cpu()
qg_model.cpu()
qg_model = None
with torch.no_grad():
for question in tqdm(questions):
question = question.split("\t")[0]
question = "Q: " + question
input_ids_qa = qa_tokenizer.tokenize(question)
input_ids_qa += ["A", ":"]
input_ids_qa = torch.tensor(
qa_tokenizer.convert_tokens_to_ids(input_ids_qa), dtype=torch.long)
input_ids_qa = input_ids_qa.view(1, -1)
input_ids_qd = qd_tokenizer.tokenize(question)
input_ids_qd += ["A", ":"]
input_ids_qd = torch.tensor(
qd_tokenizer.convert_tokens_to_ids(input_ids_qd), dtype=torch.long)
input_ids_qd = input_ids_qd.view(1, -1)
#qa_model.cuda()
res = qa_model(input_ids_qa.cuda(),
12,
sample=False,
tmp=1.0,
label=None,
top_p=0.9)
#qa_model.cpu()
ans = res.split("\t")[1]
# distractors = set({})
distractors = qd_model(input_ids_qd.cuda(),
12,
num_distractors=distractor_size+1,
sample=True,
tmp=1,
label=None,
top_p=1.0)
distractors = set(distractors)
distractors = list(distractors)[:4]
distractors = set(distractors)
#if len(distractors) > distractor_size:
# distractors = list(distractors)
while len(distractors) < distractor_size:
new_distractors = qd_model(input_ids_qd.cuda(),
12,
num_distractors=distractor_size -
len(distractors),
sample=True,
tmp=1,
label=None,
top_p=1.0)
distractors = distractors.union(set(new_distractors))
if len(distractors) != distractor_size:
x = input(str(distractors))
#qd_model.cuda()
# for _ in range(distractor_size):
# question, distractor = qd_model.generate(input_ids_qd.cuda(),
# 12,
# sample=True,
# tmp=1.0,
# top_p=1.0,
# label=None)
# while distractor in distractors or distractor == ans:
# question, distractor = qd_model.generate(
# input_ids_qd.cuda(),
# 12,
# sample=True,
# tmp=1.0,
# top_p=1.0,
# label=None)
# distractors.add(distractor.strip())
#qd_model.cpu()
#output = [question] + [ans] + list(distractors)
tsv_writer.writerow(
["n/a",
question.replace("Q: ", "").strip(), "n/a",
ans.strip()] + list(distractors))
#data.append(tuple(output))
#random.shuffle(data)
#with open(dir + "train_fake_100000" + ".csv", 'w', encoding='utf8', newline='') as tsv_file:
# tsv_writer = csv.writer(tsv_file, delimiter=',', lineterminator='\n')
#
# tsv_writer.writerow(["id", "question", "concept", "true_answer", "wrong1" , "wrong2", "wrong3" , "wrong4"])
# for q, t, w1, w2, w3, w4 in data:
# tsv_writer.writerow(["n/a",q,"n/a",t,w1,w2,w3,w4])