Search TypeError: argument 'params': 'list' object cannot be converted to 'PyString'

[Neasontlj]

2024-06-11 10:29:37,169 [INFO] [Start Generate Column] 武汉历史沿革
2024-06-11 10:29:37,983 [ERROR] [Search]: Can not search '['武汉', '历史', '演变', '楚河汉街', '黄鹤楼', '辛亥革命']'. Error: https://duckduckgo.com/news.js TypeError: argument 'params': 'list' object cannot be converted to 'PyString'
2024-06-11 10:29:37,983 [INFO] [Search News Count] 0
2024-06-11 10:29:37,984 [INFO] [Picked News Count] 0
2024-06-11 10:29:37,984 [INFO] [Start Generate Column] 名人与武汉
2024-06-11 10:29:40,420 [ERROR] [Search]: Can not search '['武汉名人', '文化名人', '历史人物', '屈原', '李白', '张之洞']'. Error: https://duckduckgo.com/news.js TypeError: argument 'params': 'list' object cannot be converted to 'PyString'
2024-06-11 10:29:40,420 [INFO] [Search News Count] 0
2024-06-11 10:29:40,420 [INFO] [Picked News Count] 0
2024-06-11 10:29:40,421 [INFO] [Start Generate Column] 武汉地方文化
2024-06-11 10:29:41,159 [ERROR] [Search]: Can not search '['武汉文化', '民俗', '戏曲', '热干面', '豆皮', '楚剧']'. Error: https://duckduckgo.com/news.js TypeError: argument 'params': 'list' object cannot be converted to 'PyString'
2024-06-11 10:29:41,161 [INFO] [Search News Count] 0
2024-06-11 10:29:41,161 [INFO] [Picked News Count] 0
2024-06-11 10:29:41,161 [INFO] [Markdown Generation Failed] Due to have not any column data.
报错信息 以上

[Neasontlj]

2024-06-11 10:29:32,167 [INFO] [Outline Generated] {'report_title': '武汉历史文化特辑', 'column_list': [{'column_title': '武汉历史沿革', 'column_requirement': '内容需涵盖武汉的历史变迁、重要历史事件和文化遗迹', 'search_keyw', '辛亥革命']}, {'column_title': '名人与武汉', 'column_requirement': '聚焦对武汉有重大影响的历史人物，如屈原、李白、张之洞等', 'search_keywords': ['武汉名人', '文化名人', '历史人物', '屈原', '李白', '张之洞']}, {'column_title':容', 'search_keywords': ['武汉文化', '民俗', '戏曲', '热干面', '豆皮', '楚剧']}]}
2024-06-11 10:29:37,169 [INFO] [Start Generate Column] 武汉历史沿革
2024-06-11 10:29:37,983 [ERROR] [Search]: Can not search '['武汉', '历史', '演变', '楚河汉街', '黄鹤楼', '辛亥革命']'. Error: https://duckduckgo.com/news.js TypeError: argument 'params': 'list' object cannot be converted to 'PyString'
2024-06-11 10:29:37,983 [INFO] [Search News Count] 0
2024-06-11 10:29:37,984 [INFO] [Picked News Count] 0
2024-06-11 10:29:37,984 [INFO] [Start Generate Column] 名人与武汉
2024-06-11 10:29:40,420 [ERROR] [Search]: Can not search '['武汉名人', '文化名人', '历史人物', '屈原', '李白', '张之洞']'. Error: https://duckduckgo.com/news.js TypeError: argument 'params': 'list' object cannot be converted to 'PyString'
2024-06-11 10:29:40,420 [INFO] [Search News Count] 0
2024-06-11 10:29:40,420 [INFO] [Picked News Count] 0
2024-06-11 10:29:40,421 [INFO] [Start Generate Column] 武汉地方文化
2024-06-11 10:29:41,159 [ERROR] [Search]: Can not search '['武汉文化', '民俗', '戏曲', '热干面', '豆皮', '楚剧']'. Error: https://duckduckgo.com/news.js TypeError: argument 'params': 'list' object cannot be converted to 'PyString'
2024-06-11 10:29:41,161 [INFO] [Search News Count] 0
2024-06-11 10:29:41,161 [INFO] [Picked News Count] 0
2024-06-11 10:29:41,161 [INFO] [Markdown Generation Failed] Due to have not any column data.

[Maplemx]

this issue is caused by model output in wrong format.

we expect search_keywords output a string contains all keywords and use space to split

but the search_keywords output in your log is a list, that cause the error

but this is a understandable and usual model output mistake, we can try to handle it by check the output data type like:

if isinstance(column_info["search_keywords"], list):
    column_info["search_keywords"] = " ".combine(column_info["search_keywords"]

I will update code to add some compatible logic

[YueYongDev]

i also get this error, is there any update about this issue?

[FuzzyCatZ]

your [Search News Count] is 0. This might because the NEWS_TIME_LIMIT setting is too short to return any results. I suggest changing it from 'd' to 'w'."