File size: 6,203 Bytes
c3f3b0b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
"""Utils for data load, save, and process (e.g., prompt construction)"""

import os
import json
import yaml
import re


DOMAIN_CAT2SUB_CAT = {
  'Art and Design': ['Art', 'Art_Theory', 'Design', 'Music'],
  'Business': ['Accounting', 'Economics', 'Finance', 'Manage','Marketing'],
  'Science': ['Biology', 'Chemistry', 'Geography', 'Math', 'Physics',],
  'Health and Medicine': ['Basic_Medical_Science', 'Clinical_Medicine', 'Diagnostics_and_Laboratory_Medicine', 'Pharmacy', 'Public_Health'],
  'Humanities and Social Science': ['History', 'Literature', 'Sociology', 'Psychology'],
  'Tech and Engineering': ['Agriculture', 'Architecture_and_Engineering', 'Computer_Science', 'Electronics', 'Energy_and_Power', 'Materials', 'Mechanical_Engineering'],
}


CAT_SHORT2LONG = {
    'acc': 'Accounting',
    'agri': 'Agriculture',
    'arch': 'Architecture_and_Engineering',
    'art': 'Art',
    'art_theory': 'Art_Theory',
    'bas_med': 'Basic_Medical_Science',
    'bio': 'Biology',
    'chem': 'Chemistry',
    'cli_med': 'Clinical_Medicine',
    'cs': 'Computer_Science',
    'design': 'Design',
    'diag_med': 'Diagnostics_and_Laboratory_Medicine',
    'econ': 'Economics',
    'elec': 'Electronics',
    'ep': 'Energy_and_Power',
    'fin': 'Finance',
    'geo': 'Geography',
    'his': 'History',
    'liter': 'Literature',
    'manage': 'Manage',
    'mark': 'Marketing',
    'mate': 'Materials',
    'math': 'Math',
    'mech': 'Mechanical_Engineering',
    'music': 'Music',
    'phar': 'Pharmacy',
    'phys': 'Physics',
    'psy': 'Psychology',
    'pub_health': 'Public_Health',
    'socio': 'Sociology'
}

# DATA SAVING
def save_json(filename, ds):
    with open(filename, 'w') as f:
        json.dump(ds, f, indent=4)


def get_multi_choice_info(options):
    """
    Given the list of options for multiple choice question
    Return the index2ans and all_choices
    """
    
    start_chr = 'A'
    all_choices = []
    index2ans = {}
    for i, option in enumerate(options):
        index2ans[chr(ord(start_chr) + i)] = option
        all_choices.append(chr(ord(start_chr) + i))

    return index2ans, all_choices

def load_yaml(file_path):
    with open(file_path, 'r') as stream:
        try:
            yaml_dict = yaml.safe_load(stream)
        except yaml.YAMLError as exc:
            print(exc)

    return yaml_dict


def parse_img_path(text):
    matches = re.findall("<img='(.*?)'>", text)
    return matches

def process_single_sample(data):
    question = data['question']
    o_imgs_paths = []
    for option in data['options']:
        current_o_imgs_paths = parse_img_path(option)
        for img_path in current_o_imgs_paths:
            o_imgs_paths.append(img_path)

    if len(o_imgs_paths) > 1:  # multiple images in options, used for random selection
        return {'id': data['id'], 'question': question, 'options': data['options'], 'answer': data['answer'],
             'image': None, 'question_type': data['question_type']}
    else:
        return {'id': data['id'], 'question': question, 'options': data['options'], 'answer': data['answer'],
             'image': data['image_1'], 'question_type': data['question_type']}


# DATA SAVING
def save_json(filename, ds):
    with open(filename, 'w') as f:
        json.dump(ds, f, indent=4)

def save_jsonl(filename, data):
    """
    Save a dictionary of data to a JSON Lines file with the filename as key and caption as value.

    Args:
        filename (str): The path to the file where the data should be saved.
        data (dict): The dictionary containing the data to save where key is the image path and value is the caption.
    """
    with open(filename, 'w', encoding='utf-8') as f:
        for img_path, caption in data.items():
            # Extract the base filename without the extension
            base_filename = os.path.basename(img_path)
            # Create a JSON object with the filename as the key and caption as the value
            json_record = json.dumps({base_filename: caption}, ensure_ascii=False)
            # Write the JSON object to the file, one per line
            f.write(json_record + '\n')

def save_args(args, path_dir):
    argsDict = args.__dict__
    with open(path_dir + 'setting.txt', 'w') as f:
        f.writelines('------------------ start ------------------' + '\n')
        for eachArg, value in argsDict.items():
            f.writelines(eachArg + ' : ' + str(value) + '\n')
        f.writelines('------------------- end -------------------')



# DATA PROCESSING
def construct_prompt(sample, config):
    question = sample['question']
    options = eval(sample['options'])
    example = ""
    if sample['question_type'] == 'multiple-choice':
        start_chr = 'A'
        prediction_range = []
        index2ans = {}
        for option in options:
            prediction_range.append(start_chr)
            example += f"({start_chr}) {option}\n"
            index2ans[start_chr] = option
            start_chr = chr(ord(start_chr) + 1)
        empty_prompt_sample_structure = config['multi_choice_example_format']
        empty_prompt = empty_prompt_sample_structure.format(question, example)
        res_dict = {}
        res_dict['index2ans'] = index2ans
        res_dict['correct_choice'] = sample['answer']
        res_dict['all_choices'] = prediction_range
        res_dict['empty_prompt'] = empty_prompt
        if config['task_instructions']:
            res_dict['final_input_prompt'] = config['task_instructions'].strip() + '\n\n' + empty_prompt
        else:
            res_dict['final_input_prompt'] = empty_prompt

        res_dict['gt_content'] = options[ord(sample['answer'].upper()) - ord('A')]
    else:
        empty_prompt_sample_structure = config['short_ans_example_format']
        empty_prompt = empty_prompt_sample_structure.format(question)
        res_dict = {}
        res_dict['empty_prompt'] = empty_prompt
        if config['task_instructions']:
            res_dict['final_input_prompt'] = config['task_instructions'].strip() + '\n\n' + empty_prompt
        else:
            res_dict['final_input_prompt'] = empty_prompt
        res_dict['gt_content'] = sample['answer']

    res_dict.update(sample)
    return res_dict