---
# System prepended metadata

title: ask_self

---

## ask_self
如果你不想使用文件檔案，希望直接輸入文件內容，你可以使用ask_self，使用***info*** 參數將文件的內容傳給語言模型，***info***參數可為***str*** 或者 ***list of str***。


### example

```python!=
install_requires = [
    "pypdf",
    "langchain>=0.1.0",
    "chromadb==0.4.14",
    "openai==0.27",
    "tiktoken",
    "lark==1.1.7",
    "scikit-learn<1.3.0",
    "jieba==0.42.1",
    "sentence-transformers==2.2.2",
    "torch==2.0.1",
    "transformers>=4.33.4", 
    "llama-cpp-python==0.2.6",
    "auto-gptq==0.3.1",
    "tqdm==4.65.0",
    "docx2txt==0.8",
    "rouge==1.0.1",
    "rouge-chinese==1.0.3",
    "bert-score==0.3.13",
    "click",
    "tokenizers>=0.13.3",
    "streamlit==1.28.2",
    "streamlit_option_menu==0.3.6",
]

ak = akasha.Doc_QA(
    verbose=True,
    max_doc_len=15000,
    model="openai:gpt-4",
)
response = ak.ask_self(prompt="langchain的套件版本?", info=install_requires)
```


```shell
langchain的套件版本是0.1.0或更高版本。
```