检查其他资源

• 为这个问题添加了一个非常描述性的标题。
• 使用集成搜索在LangChain文档中进行了搜索。
• 使用GitHub搜索查找类似的问题，但没有找到。
• 我确信这是LangChain中的一个bug,而不是我的代码。
• 通过更新到LangChain的最新稳定版本(或特定集成包)无法解决此bug。

示例代码

以下是使用PydanticOutputParser时，由于Langchain无法解析LLM输出而出现的代码：

HUGGINGFACEHUB_API_TOKEN = os.getenv("HUGGINGFACEHUB_API_TOKEN")

repo_id = "mistralai/Mistral-7B-Instruct-v0.3"

model_kwargs = {
    "max_new_tokens": 60, 
    "max_length": 200, 
    "temperature": 0.1, 
    "timeout": 6000
}

# Using HuggingFaceHub
llm = HuggingFaceHub(
    repo_id=repo_id,
    huggingfacehub_api_token = HUGGINGFACEHUB_API_TOKEN,
    model_kwargs = model_kwargs,
)

# Define your desired data structure.
class Suggestions(BaseModel):
    words: List[str] = Field(description="list of substitute words based on context")

    # Throw error in case of receiving a numbered-list from API
    @field_validator('words')
    def not_start_with_number(cls, field):
        for item in field:
            if item[0].isnumeric():
                raise ValueError("The word can not start with numbers!")
        return field

parser = PydanticOutputParser(pydantic_object=Suggestions)

prompt_template = """
Offer a list of suggestions to substitute the specified target_word based on the context.
{format_instructions}
target_word={target_word}
context={context}
"""

prompt_input_variables = ["target_word", "context"]
partial_variables = {"format_instructions":parser.get_format_instructions()}
prompt = PromptTemplate(
    template=prompt_template,
    input_variables=prompt_input_variables,
    partial_variables=partial_variables
)

model_input = prompt.format_prompt(
			target_word="behaviour",
			context="The behaviour of the students in the classroom was disruptive and made it difficult for the teacher to conduct the lesson."
)

output = llm(model_input.to_string())

parser.parse(output)

尝试使用OutputFixingParser修复错误时，又遇到了另一个错误，以下是代码库：

outputfixing_parser = OutputFixingParser.from_llm(parser=parser,llm=llm)
print(outputfixing_parser)
outputfixing_parser.parse(output)

错误信息和堆栈跟踪(如果适用)

---------------------------------------------------------------------------
ValidationError                           Traceback (most recent call last)
File [~\Desktop\llmai\llm_deep\Lib\site-packages\langchain_core\output_parsers\pydantic.py:33](~/Desktop/llmai/llm_deep/Lib/site-packages/langchain_core/output_parsers/pydantic.py:33), in PydanticOutputParser._parse_obj(self, obj)
     [32](~/Desktop/llmai/llm_deep/Lib/site-packages/langchain_core/output_parsers/pydantic.py:32) if issubclass(self.pydantic_object, pydantic.BaseModel):
---> [33](~/Desktop/llmai/llm_deep/Lib/site-packages/langchain_core/output_parsers/pydantic.py:33)     return self.pydantic_object.model_validate(obj)
     [34](~/Desktop/llmai/llm_deep/Lib/site-packages/langchain_core/output_parsers/pydantic.py:34) elif issubclass(self.pydantic_object, pydantic.v1.BaseModel):

File [~\Desktop\llmai\llm_deep\Lib\site-packages\pydantic\main.py:551](~/Desktop/llmai/llm_deep/Lib/site-packages/pydantic/main.py:551), in BaseModel.model_validate(cls, obj, strict, from_attributes, context)
    [550](~/Desktop/llmai/llm_deep/Lib/site-packages/pydantic/main.py:550) __tracebackhide__ = True
--> [551](~/Desktop/llmai/llm_deep/Lib/site-packages/pydantic/main.py:551) return cls.__pydantic_validator__.validate_python(
    [552](~/Desktop/llmai/llm_deep/Lib/site-packages/pydantic/main.py:552)     obj, strict=strict, from_attributes=from_attributes, context=context
    [553](~/Desktop/llmai/llm_deep/Lib/site-packages/pydantic/main.py:553) )

ValidationError: 1 validation error for Suggestions
words
  Field required [type=missing, input_value={'properties': {'words': ..., 'required': ['words']}, input_type=dict]
    For further information visit https://errors.pydantic.dev/2.7/v/missing

During handling of the above exception, another exception occurred:

OutputParserException                     Traceback (most recent call last)
Cell In[284], [line 1](vscode-notebook-cell:?execution_count=284&line=1)
----> [1](vscode-notebook-cell:?execution_count=284&line=1) parser.parse(output)

File [~\Desktop\llmai\llm_deep\Lib\site-packages\langchain_core\output_parsers\pydantic.py:64](~/Desktop/llmai/llm_deep/Lib/site-packages/langchain_core/output_parsers/pydantic.py:64), in PydanticOutputParser.parse(self, text)
...

OutputParserException: Failed to parse Suggestions from completion {"properties": {"words": {"description": "list of substitute words based on context", "items": {"type": "string"}, "title": "Words", "type": "array"}}, "required": ["words"]}. Got: 1 validation error for Suggestions
words
  Field required [type=missing, input_value={'properties': {'words': ..., 'required': ['words']}, input_type=dict]
    For further information visit https://errors.pydantic.dev/2.7/v/missing

使用OutputFixingParser时的错误

---------------------------------------------------------------------------
ValidationError                           Traceback (most recent call last)
File [~\Desktop\llmai\llm_deep\Lib\site-packages\langchain_core\output_parsers\pydantic.py:33](~\Desktop/llmai/llm_deep/Lib/site-packages/langchain_core/output_parsers/pydantic.py:33), in PydanticOutputParser._parse_obj(self, obj)
     [32](~/Desktop/llmai/llm_deep/Lib/site-packages/langchain_core/output_parsers/pydantic.py:32) if issubclass(self.pydantic_object, pydantic.BaseModel):
---> [33](~/Desktop/llmai/llm_deep/Lib/site-packages/langchain_core/output_parsers/pydantic.py:33)     return self.pydantic_object.model_validate(obj)
     [34](~/Desktop/llmai/llm_deep/Lib/site-packages/langchain_core/output_parsers/pydantic.py:34) elif issubclass(self.pydantic_object, pydantic.v1.BaseModel):

File [~\Desktop\llmai\llm_deep\Lib\site-packages\pydantic\main.py:551](~/Desktop/llmai/llm_deep/Lib/site-packages/pydantic/main.py:551), in BaseModel.model_validate(cls, obj, strict, from_attributes, context)
    [550](~/Desktop/llmai/llm_deep/Lib/site-packages/pydantic/main.py:550) __tracebackhide__ = True
--> [551](~/Desktop/llmai/llm_deep/Lib/site-packages/pydantic/main.py:551) return cls.__pydantic_validator__.validate_python(
    [552](~/Desktop/llmai/llm_deep/Lib/site-packages/pydantic/main.py:552)     obj, strict=strict, from_attributes=from_attributes, context=context
    [553](~/Desktop/llmai/llm_deep/Lib/site-packages/pydantic/main.py:553) )

ValidationError: 1 validation error for Suggestions
  Input should be a valid dictionary or instance of Suggestions [type=model_type, input_value=None, input_type=NoneType]
    For further information visit https://errors.pydantic.dev/2.7/v/model_type

During handling of the above exception, another exception occurred:

OutputParserException                     Traceback (most recent call last)
Cell In[265], [line 1](vscode-notebook-cell:?execution_count=265&line=1)
----> [1](vscode-notebook-cell:?execution_count=265&line=1) outputfixing_parser.parse(output)

File [~\Desktop\llmai\llm_deep\Lib\site-packages\langchain\output_parsers\fix.py:62](~/Desktop/llmai/llm_deep/Lib/site-packages/langchain/output_parsers/fix.py:62), in OutputFixingParser.parse(self, completion)
     [60](~/Desktop/llmai/llm_deep/Lib/site-packages/langchain/output_parsers/fix.py:60) except OutputParserException as e:
...
     [44](~/Desktop/llmai/llm_deep/Lib/site-packages/langchain_core/output_parsers/pydantic.py:44)     try:

OutputParserException: Failed to parse Suggestions from completion null. Got: 1 validation error for Suggestions
  Input should be a valid dictionary or instance of Suggestions [type=model_type, input_value=None, input_type=NoneType]
    For further information visit https://errors.pydantic.dev/2.7/v/model_type

描述

输出必须能够解析LLM输出并提取如下所示生成的json;

Suggestions(words=["conduct", "misconduct", "actions", "antics", "performance", "demeanor", "attitude", "behavior", "manner", "pupil actions"])

系统信息

系统信息

操作系统：Windows
操作系统版本：10.0.22631
Python版本：3.11.9 | 由conda-forge打包 | (主版本，2024年4月19日，18:27:10) [MSC v.1938 64位(AMD64)]

软件包信息

langchain_core: 0.2.4
langchain: 0.2.2
langchain_community: 0.2.4
langsmith: 0.1.73
langchain_google_community: 1.0.5
langchain_huggingface: 0.0.3
langchain_text_splitters: 0.2.1

未安装的软件包(不一定是个问题)

以下软件包未找到：
langgraph
langserve