Hugging Face's logo

Spaces:
ASLP-lab
/
OSUM-EChat
Running on Zero

App Files Community
1
OSUM-EChat
File size: 7,629 Bytes 
841f290
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
 
# 负责wav & text instruct 的处理
import codecs
import json
import random

from gxl_ai_utils.utils import utils_file
import torch
import os

# asr+X
task_names = [
    "age",
    "gender",
    "style",
    "emotion",
    "caption"
]
map_dict = {
    "<TRANSCRIBE> <STYLE>": "style",
    "<TRANSCRIBE> <AGE>": "age",
    "<TRANSCRIBE> <GENDER>": "gender",
    "<TRANSCRIBE> <EMOTION>": "emotion",
    "<TRANSCRIBE> <CAPTION>": "caption",
    "<STYLE>": "style_only_X",
    "<AGE>": "age_only_X",
    "<GENDER>": "gender_only_X",
    "<EMOTION>": "emotion_only_X",
    "<CAPTION>": "caption_only_X",
}
question_info_dir = "/mnt/sfs/asr/code/osum_osum_echat/examples/wenetspeech/whisper/conf/language_follow_prompt/Q_with_asr"
# question_info_dir = "D:\osum_echat_workspace_2\osum_using_ofical\osum_osum_echat\examples\wenetspeech\whisper\conf\language_follow_prompt\Q_with_asr"
question_big_dict = {}
question_wav_big_dict = {}

# for task_name in task_names:
#     task_json_path = os.path.join(question_info_dir, f"{task_name}.json")
#     dict_i = utils_file.load_dict_from_json(task_json_path)
#     question_strs = list(dict_i.values())
#     question_big_dict[task_name] = question_strs
#
#
# for task_name in task_names:
#     task_json_path = os.path.join(question_info_dir, f"{task_name}.json")
#     dict_i = utils_file.load_dict_from_json(task_json_path)
#     question_strs = list(dict_i.keys())  # 获取字典的键作为问题字符串列表
#     question_wav_big_dict[task_name] =  []
#     for i in range(1, 21):
#         new_question_strs = [f"/mnt/sfs/asr/update_data/Q_with_asr_added_by_20250406/{task_name}_{x}_speaker_{i}.wav" for x in question_strs]
#         question_wav_big_dict[task_name].extend(new_question_strs)

question_info_dir = "/mnt/sfs/asr/code/osum_osum_echat/examples/wenetspeech/whisper/conf/language_follow_prompt/Q"
# question_info_dir = "D:\osum_echat_workspace_2\osum_using_ofical\osum_osum_echat\examples\wenetspeech\whisper\conf\language_follow_prompt\Q"
# for task_name in task_names:
#     task_json_path = os.path.join(question_info_dir, f"{task_name}.json")
#     dict_i = utils_file.load_dict_from_json(task_json_path)
#     question_strs = list(dict_i.values())
#     question_big_dict[task_name+"_only_X"] = question_strs
# for task_name in task_names:
#     task_json_path = os.path.join(question_info_dir, f"{task_name}.json")
#     dict_i = utils_file.load_dict_from_json(task_json_path)
#     question_strs = list(dict_i.keys())  # 获取字典的键作为问题字符串列表
#     question_wav_big_dict[task_name+"_only_X"] =  []
#     for i in range(1,21):
#         new_question_strs = [f"/mnt/sfs/asr/update_data/Q_only_X_added_by_20250406/{task_name}_{x}_speaker_{i}.wav" for x in question_strs]
#         question_wav_big_dict[task_name+"_only_X"].extend(new_question_strs)


# for answer
answer_info_dir = "/mnt/sfs/asr/code/osum_osum_echat/examples/wenetspeech/whisper/conf/language_follow_prompt/Q_with_asr/answer"
# answer_info_dir = "D:\osum_echat_workspace_2\osum_using_ofical\osum_osum_echat\examples\wenetspeech\whisper\conf\language_follow_prompt\Q_with_asr\\answer"
answer_big_dict = {}
# for task_name in task_names:
#     task_json_path = os.path.join(answer_info_dir, f"{task_name}.json")
#     dict_i = utils_file.load_dict_from_json(task_json_path)
#     answer_big_dict[task_name] = dict_i

answer_info_dir = "/mnt/sfs/asr/code/osum_osum_echat/examples/wenetspeech/whisper/conf/language_follow_prompt/Q/answer"
# answer_info_dir = "D:\osum_echat_workspace_2\osum_using_ofical\osum_osum_echat\examples\wenetspeech\whisper\conf\language_follow_prompt\Q\\answer"
# for task_name in task_names:
#     task_json_path = os.path.join(answer_info_dir, f"{task_name}.json")
#     dict_i = utils_file.load_dict_from_json(task_json_path)
#     answer_big_dict[task_name+"_only_X"] = dict_i



def get_question_prompt_by_task(task_tag):
    key_str = map_dict.get(task_tag, None)
    if key_str is None:
        return None
    question_strs = question_big_dict.get(key_str, [])
    if len(question_strs) == 0:
        return None
    # 随机算一个句子
    question_str = random.choice(question_strs)
    return question_str

def get_question_wav_path_by_task(task_tag):
    key_str = map_dict.get(task_tag, None)
    if key_str is None:
        return None
    question_strs = question_wav_big_dict.get(key_str, [])
    if len(question_strs) == 0:
        return None
    # 随机算一个句子
    question_str = random.choice(question_strs)
    return question_str

def get_answer_prompt_by_task(task_tag, answer_tag, asr_txt=None):
    key_str = map_dict.get(task_tag, None)
    if key_str is None:
        utils_file.logging_error(f"task_tag {task_tag} not in map_dict")
        return None
    answer_dict = answer_big_dict.get(key_str, {})
    answer_strs = answer_dict.get(answer_tag, [])
    if len(answer_strs) == 0:
        utils_file.logging_error(f"answer_tag {answer_tag} not in answer_dict of {task_tag}")
        return None
    # 随机算一个句子
    answer_str = random.choice(answer_strs)
    if asr_txt is not None:
        if "{}" not in answer_str:
            utils_file.logging_error(f"answer_str {answer_str} not contain","{}", 'but asr_txt is not None')
            return None
        answer_str = answer_str.format(asr_txt)
    else:
        if "{}" in answer_str:
            utils_file.logging_error(f"answer_str {answer_str} contain","{}", 'but asr_txt is None')
            return None
    return answer_str

if __name__ == '__main__':
    task_tag = "<CAPTION>"
    type_tag = "church_bells"
    question_str = get_question_wav_path_by_task(task_tag)
    print(question_str)
    question_str = get_question_wav_path_by_task(task_tag)
    print(question_str)
    question_str = get_question_wav_path_by_task(task_tag)
    print(question_str)
    question_str = get_question_wav_path_by_task(task_tag)
    print(question_str)

    question_str = get_question_prompt_by_task(task_tag)
    print(question_str)
    question_str = get_question_prompt_by_task(task_tag)
    print(question_str)
    question_str = get_question_prompt_by_task(task_tag)
    print(question_str)
    question_str = get_question_prompt_by_task(task_tag)
    print(question_str)

    question_str = get_answer_prompt_by_task(task_tag, type_tag)
    print(question_str)
    question_str = get_answer_prompt_by_task(task_tag, type_tag)
    print(question_str)
    question_str = get_answer_prompt_by_task(task_tag, type_tag)
    print(question_str)
    question_str = get_answer_prompt_by_task(task_tag, type_tag)
    print(question_str)

    question_str = get_answer_prompt_by_task(task_tag, type_tag, "I am a child")
    print(question_str)
    question_str = get_answer_prompt_by_task(task_tag, type_tag, "I am a child")
    print(question_str)
    question_str = get_answer_prompt_by_task(task_tag, type_tag, "I am a child")
    print(question_str)
    question_str = get_answer_prompt_by_task(task_tag, type_tag, "I am a child")
    print(question_str)

    # caption_json_path = "D:\osum_echat_workspace_2\osum_using_ofical\osum_osum_echat\examples\wenetspeech\whisper\conf\language_follow_prompt\Q_with_asr\\answer\caption.json"
    # caption_dict = utils_file.load_dict_from_json(caption_json_path)
    # new_caption_dict = {}
    # utils_file.print_list(list(caption_dict.keys()))
    # for k, v in caption_dict.items():
    #     new_k = k.lower()
    #     new_caption_dict[new_k] = v
    # with codecs.open(caption_json_path, 'w', encoding='utf-8') as f:
    #     json.dump(new_caption_dict, f, ensure_ascii=False, indent=4)