OliveSensorAPI/scripts/qa_generation/main.py

109 lines
3.7 KiB
Python
Raw Normal View History

2024-03-07 17:56:07 +08:00
import os
import json
import time
2024-03-07 17:56:07 +08:00
from tqdm import tqdm
import concurrent.futures
2024-03-07 17:56:07 +08:00
from datetime import datetime
import numpy as np
2024-03-07 17:56:07 +08:00
from config.config import result_dir, storage_interval, window_size, overlap_size, multi_process_num
2024-03-07 17:56:07 +08:00
from model.qwen import call_qwen_single_turn
from util.logger import get_logger
from util.data_loader import get_file_list, get_txt_content, capture_qa, merge_sub_qa_generation, save_to_file
2024-03-07 17:56:07 +08:00
logger = get_logger()
"""
每个线程产生 QA 对以及存储到子文件中
"""
def single_thread_generate(thread_num, interval, model_caller, storage_jsonl_path, contents):
storage_counter = 0
storage_list = []
for content in tqdm(contents):
try:
response = model_caller(content)
captured_qa = capture_qa(response)
if captured_qa is None:
continue
storage_list.extend(captured_qa)
storage_counter += 1
if storage_counter % interval == 0:
save_to_file(storage_jsonl_path, storage_list)
storage_counter = 0
storage_list = []
except Exception as exc:
logger.error("QA generation error : %s" % (exc))
if storage_list:
save_to_file(storage_jsonl_path, storage_list)
storage_list = []
2024-03-07 17:56:07 +08:00
"""
生成 QA
model_name: 可调用的模型名称暂时只实现了 qwen
interval: 存储间隔即每隔多少条存一次文件过密的间隔会增大 IO 开销
"""
def generate_qa(
model_name: str = 'qwen',
interval: int = 10,
2024-03-07 17:56:07 +08:00
):
current_time = datetime.now().strftime("%Y-%m-%d_%H-%M-%S")
if model_name == 'qwen':
model_caller = call_qwen_single_turn
else:
logger.warning('This model is currently not supported and will call the default model - qwen.')
model_caller = call_qwen_single_turn
model_name = 'qwen'
logger.info(f'The called model is: {model_name}.')
logger.info(f'The storage interval is: {interval}.')
file_list = get_file_list()
storage_counter = 0
storage_list = []
2024-03-08 14:25:34 +08:00
for file_path in file_list:
2024-03-08 19:00:45 +08:00
contents = get_txt_content(file_path, window_size=window_size, overlap_size=overlap_size)
2024-03-07 17:56:07 +08:00
storage_list = []
2024-03-08 14:25:34 +08:00
_, file_name = os.path.split(file_path)
storage_jsonl_path = os.path.join(
result_dir, f'{current_time}-{file_name}-{model_name}.jsonl')
2024-03-07 17:56:07 +08:00
logger.info(f'The generated QA will be stored in {storage_jsonl_path}.')
# 基于并发个数切分 contents 内容
contents_array = np.array(contents)
chunks = np.array_split(contents_array, multi_process_num)
# 构建并发参数 list
parameters_list = list()
for thread_num, chunk in enumerate(chunks):
parameters_list.append(
[thread_num, interval, model_caller, storage_jsonl_path + f'-{thread_num}', list(chunk)]
)
# 并发生成 QA 对
with concurrent.futures.ThreadPoolExecutor(max_workers=multi_process_num) as executor:
# 创建一个Future列表它们将对应每个worker_function的结果
futures = [executor.submit(single_thread_generate, *parameters) for parameters in parameters_list]
for future in concurrent.futures.as_completed(futures):
try:
future.result()
except Exception as exc:
logger.error("Thread generated an exception: %s" % (exc))
2024-03-07 17:56:07 +08:00
merge_sub_qa_generation(result_dir, storage_jsonl_path)
2024-03-07 17:56:07 +08:00
if __name__ == '__main__':
2024-03-08 14:25:34 +08:00
# 创建generated文件夹
os.makedirs('./data/generated', exist_ok=True)
2024-03-08 19:00:45 +08:00
generate_qa(interval=storage_interval)