报错，'Connection aborted.', ConnectionResetError(104, 'Connection reset by peer')，感谢作者的回复

[iMoriton]

import json
from pygtrans import Translate

# 加载原始数据集
with open('CoT_data.json', 'r', encoding='utf-8') as f:
    original_data = json.load(f)

# 初始化翻译客户端
client = Translate(
    domain='com.bo',
    proxies={'http': 'http://172.xx.xx.xxx:7899',
             'https': 'http://172.xx.xx.xxx:7899'}
)

# 创建纯藏语数据集容器
translated_data = [{
    "instruction": "",
    "input": "",
    "output": ""
} for _ in original_data]

# 分字段批量处理核心逻辑
for field in ['instruction', 'input', 'output']:
    # 提取字段内容并记录位置
    texts = []
    indexes = []
    for idx, item in enumerate(original_data):
        if field in item and item[field]:
            texts.append(item[field])
            indexes.append(idx)
    results = []
    # 批量翻译当前字段
    if texts:
        # print(texts)
        for i in range(0, len(texts), 100):
            results.extend(client.translate(texts[i:i + 100], target='bo'))
        # 回填翻译结果
        for result_idx, idx in enumerate(indexes):
            translated_data[idx][field] = results[result_idx].translatedText

# 保存结果
with open('tibetan_instructions.json', 'w', encoding='utf-8') as f:
    json.dump(translated_data, f, ensure_ascii=False, indent=2)

[lemisky]

这必然是网络问题吧

你试试直接请求谷歌, 用以验证网络是否畅通

import requests

# 没有全局 VPN 的话, 使用 proxies 设置下代理
r = requests.get('https://google.com/')

print(r)

"""
>>>  <Response [200]>
"""