File size: 6,580 Bytes
45f037d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
import pathlib
import textwrap

import google.generativeai as genai

import PIL.Image
import time
import os

import random
import numpy as np
def seed_everything(seed):
    random.seed(seed)
    np.random.seed(seed)
    os.environ['PYTHONHASHSEED'] = str(seed)
    # torch.manual_seed(seed)
    # torch.cuda.manual_seed(seed)
    # torch.backends.cudnn.deterministic = True
    # env.seed(seed)
seed_everything(1)

def random_generate_questions():
    if os.path.isdir("questions-hole-pos-50"):
        print("Questions are already generated")
        return
    os.makedirs("questions-hole-pos-50", exist_ok=False)
    os.makedirs("answers-hole-pos-50", exist_ok=False)
    for level in range(3,9):
        os.makedirs("questions-hole-pos-50/level%d"%(level))
        os.makedirs("answers-hole-pos-50/level%d"%(level))
        for question_id in range(100):
            # random ask a grid
            row_num = random.sample(range(level), 1)[0]
            col_num = random.sample(range(level), 1)[0]
            # find the GT
            with open("utils/maps-50-text/level%d/%d.txt"%(level, question_id), "r") as f:
                contents = f.read()
                rows = contents.split('\n')
                grid_content = rows[row_num][col_num]
            if grid_content == "H":
                answer = "Y"
            else:
                answer = "N"
            # write question and answer
            question = "Is there a hole in row %d, column %d?"%(row_num+1, col_num+1)
            with open("questions-hole-pos-50/level%d/%d.txt"%(level, question_id), "w") as f:
                f.write(question)
            with open("answers-hole-pos-50/level%d/%d.txt"%(level, question_id), "w") as f:
                f.write(answer)



levels = [3,4,5,6,7,8]
in_context_example_num = 0 # 0, 1, 2, 4, 8


GOOGLE_API_KEY='YOUR-API-KEY'
genai.configure(api_key=GOOGLE_API_KEY)
model = genai.GenerativeModel('gemini-pro-vision')
if in_context_example_num > 0:
    output_path = "output/output_img_%d/"%(in_context_example_num)
    input_backup_path = "input/input_backup_img_%d/"%(in_context_example_num)
else:
    output_path = "output/output_img/"
    input_backup_path = "input/input_backup_img/"

os.makedirs(output_path, exist_ok=True)
os.makedirs(input_backup_path, exist_ok=True)

EXAMPLE_DICT = {
    3: [],
    4: [],
    5: [],
    6: [],
    7: [],
    8: [],
}

# Prepare examples
for level in levels:
    for example_id in range(8):
        img_input = PIL.Image.open("example/level%d/img/%d.png"%(level, example_id))
        with open("example/level%d/question/%d.txt"%(level, example_id), "r") as f:
            question_input = f.read()
        with open("example/level%d/answer/%d.txt"%(level, example_id), "r") as f:
            answer_input = f.read()
        this_example = (img_input, question_input, answer_input)
        EXAMPLE_DICT[level].append(this_example)

# import ipdb; ipdb.set_trace()
for level in levels:
    os.makedirs(output_path + "level%d"%(level), exist_ok=True)
    os.makedirs(input_backup_path + "level%d"%(level), exist_ok=True)
    start_idx = 0
    end_idx = 100
    runned_term = 0
    map_path = "maps/level%d/img/"%(level)
    while True:
        try:
            curr_id = start_idx + runned_term
            if curr_id >= end_idx:
                break
            img_input = PIL.Image.open(map_path + "%d.png"%(curr_id))
            prompt_input_1 = '''
In this task, you will analyze a maze to determine if there is a hole in a specific position.
The following figure illustrates the appearances of the player, holes, lands, and the goal within the maze. You will need to focus on the appearance of the hole.
'''
            prompt_input_2 = '''
Here is an example to illustrate how to analyze and answer the question:
'''
            prompt_input_3 = '''
Example question: Is there a hole in row 3, column 3?

In this example:
- We check the position in row 3, column 3.
- According to the image, it is a land square. It does not contain a hole.
- Therefore, you will output "<Output> No".

Your output should be: "<Output> No" or "<Output> Yes", depending on whether there is a hole at the specified position.
'''
            question_examples = []
            answer_examples = []
            image_examples = []
            if in_context_example_num > 0:
                example_indices = random.sample(range(8), in_context_example_num)
                for example_index in example_indices:
                    this_example = EXAMPLE_DICT[level][example_index]
                    image_examples.append(this_example[0])
                    question_examples.append(this_example[1] + "\n")
                    answer_examples.append(this_example[2] + "\n\n")
            prompt_input_4 = "\n\nNow you will analyze the following maze and answer the question: "
            with open("maps/level%d/question/%d.txt"%(level, curr_id), "r") as f:
                prompt_input_5 = f.read()
            prompt_img_1 = PIL.Image.open('prompt-visual-images/system-figure-1.png')
            prompt_img_2 = PIL.Image.open('prompt-visual-images/system-figure-2.png')
            model_input_seq = [prompt_input_1, prompt_img_1, prompt_input_2, prompt_img_2, prompt_input_3]
            if in_context_example_num > 0:
                assert len(question_examples) == len(image_examples)
                assert len(question_examples) == in_context_example_num
                model_input_seq.append("## Example:\n")
                for example_index in range(in_context_example_num):
                    model_input_seq.append(image_examples[example_index])
                    model_input_seq.append(question_examples[example_index])
                    model_input_seq.append(answer_examples[example_index])
            model_input_seq += [prompt_input_4, img_input, prompt_input_5]

            response = model.generate_content(model_input_seq)
            with open(input_backup_path + "level%d/%d.txt"%(level, curr_id), "w") as f:
                contents = ""
                for input_prompt_index in range(len(model_input_seq)):
                    if type(model_input_seq[input_prompt_index]) == type("string"):
                        contents += model_input_seq[input_prompt_index]
                f.write(contents)
            with open(output_path + "level%d/%d.txt"%(level, curr_id), "w") as f:
                f.write(response.text)
            # import ipdb; ipdb.set_trace()
            # pass
            time.sleep(2)
            runned_term += 1
        except:
            time.sleep(2)
            pass