# auto_paper_reading **Repository Path**: Stanwade/auto_paper_reading ## Basic Information - **Project Name**: auto_paper_reading - **Description**: paper reading assistant - **Primary Language**: Unknown - **License**: MIT - **Default Branch**: master - **Homepage**: None - **GVP Project**: No ## Statistics - **Stars**: 0 - **Forks**: 0 - **Created**: 2024-12-25 - **Last Updated**: 2025-02-13 ## Categories & Tags **Categories**: Uncategorized **Tags**: None ## README ### NOTE!!!! We can support more journals now!!! ### 环境配置 `python -m pip install -r requirements.txt` 百炼平台的API的配置可以参考[获取API key](https://help.aliyun.com/zh/model-studio/developer-reference/get-api-key)和[添加API Key](https://help.aliyun.com/zh/model-studio/developer-reference/configure-api-key-through-environment-variables) ### 直接对pdf文件夹进行分析 先跑`get_fileid_from_dir.py`,添加`-d`参数为pdf存放目录,该脚本可以搜索目录及其子目录中的所有pdf文件,给参数时需要使用'/'分隔符的路径,否则代码可能运行不通过; ``` python .\get_fileid_from_dir.py -d C:/Users/LZH/Nutstore/1/library/CNS ``` 跑`analysis_with_LLM.py`脚本,需要添加`-j`参数为上一步生成的json文件,; ``` python .\analysis_with_LLM.py -j .\CNS_microscopy_pdf_id.json ``` 运行结束后就可以在存放pdf的文件目录中看到生成的分析markdown。 这种方式进行分析时,需要把pdf上传到qwen平台,然后获取fileid(保存在json文件中),因为百炼平台有文件总量的限制,若需要删除远程的文件,可以运行`clear_fileid.py`,但是它会删除所有的远程文件。 ### 从网页数据进行分析 0. create & edit `./config.yml` (at this repo's root folder), you can get values from your account and use url corresponding: ``` (in config.yml) api-key: "sk-xxxxxxx" base-url: "https://dashscope.aliyuncs.com/compatible-mode/v1" ``` You can get api-key by following this link: https://help.aliyun.com/zh/model-studio/developer-reference/get-api-key edit json file "links_config.json" to add the journal and sub journal with their base-url and url-postfix 1. run `fetch_and_save_xlsx.py` 1. in this file's `if __name__ == '__main__':` section, you can change the year you want. you can choose year, journal, sub journal name in the function "fetch_articles" 2. run `read_conclude_through_web.py` 1. you can also change the year in `if __name__ == '__main__'` section. ----------------------- ### Qwen Doc: Json Output run `generate_conclusion.py` to get a survey report.(Under building)