[BUG] 简洁阐述问题 / langchain部署中使用api来调用llm模型和embedding模型出现bug，初始化知识库失败

[Jacob021219]

问题描述 / Problem Description
langchain部署中使用api来调用llm模型和embedding模型出现bug，初始化知识库失败

复现问题的步骤 / Steps to Reproduce

1. 我修改了model_settings.yaml，我使用阿里dashscope的api服务调用qwen-plus和text-embedding-v3分别作为LLM和embedding
2. 执行chatchat kb -r初始化向量库
3. 出现问题
   3.1openai.BadRequestError: Error code: 400 - {'error': {'message': '<400> InternalError.Algo.InvalidParameter: Value error, contents is neither str nor list of str.: input.contents', 'type': 'InvalidParameter', 'param': None, 'code': 'InvalidParameter'}, 'id': '928d36f0-6097-9f7f-964d-f521ae99bfa0', 'request_id': '928d36f0-6097-9f7f-964d-f521ae99bfa0'}
   3.2ValueError: Tag "<400>" does not correspond to any known color directive, make sure you did not misspelled it (or prepend '' to escape it)
   3.3ValueError: Tag "<400>" does not correspond to any known color directive, make sure you did not misspelled it (or prepend '' to escape it)

完整报错 / whole problem
(chatchat) PS D:\llm_project\Langchain-Chatchat-master> chatchat kb -r
C:\Users\Jacob.conda\envs\chatchat\lib\site-packages\pydantic_settings\main.py:426: UserWarning: Config key json_file is set in model_config but will be ignored because no JsonConfigSettingsSource source is configured. To use this config key,
add a JsonConfigSettingsSource source to the settings sources via the settings_customise_sources hook.
self._settings_warn_unused_config_keys(sources, self.model_config)
C:\Users\Jacob.conda\envs\chatchat\lib\site-packages\langchain_api\module_import.py:87: LangChainDeprecationWarning: Importing GuardrailsOutputParser from langchain.output_parsers is deprecated. Please replace the import with the following:
from langchain_community.output_parsers.rail_parser import GuardrailsOutputParser
warnings.warn(
C:\Users\Jacob.conda\envs\chatchat\lib\site-packages\pydantic_settings\main.py:426: UserWarning: Config key json_file is set in model_config but will be ignored because no JsonConfigSettingsSource source is configured. To use this config key,
add a JsonConfigSettingsSource source to the settings sources via the settings_customise_sources hook.
self._settings_warn_unused_config_keys(sources, self.model_config)
C:\Users\Jacob.conda\envs\chatchat\lib\site-packages\langchain_api\module_import.py:87: LangChainDeprecationWarning: Importing GuardrailsOutputParser from langchain.output_parsers is deprecated. Please replace the import with the following:
from langchain_community.output_parsers.rail_parser import GuardrailsOutputParser
warnings.warn(
recreating all vector stores
2025-12-26 15:25:55.437 | INFO | chatchat.server.knowledge_base.kb_cache.faiss_cache:load_vector_store:109 - loading vector store in 'samples/vector_store/text-embedding-v4' from disk.
Process Process-1:
Traceback (most recent call last):
File "D:\llm_project\Langchain-Chatchat-master\libs\chatchat-server\chatchat\server\knowledge_base\kb_cache\faiss_cache.py", line 126, in load_vector_store
vector_store = self.new_vector_store(
File "D:\llm_project\Langchain-Chatchat-master\libs\chatchat-server\chatchat\server\knowledge_base\kb_cache\faiss_cache.py", line 63, in new_vector_store
vector_store = FAISS.from_documents([doc], embeddings, normalize_L2=True)
File "C:\Users\Jacob.conda\envs\chatchat\lib\site-packages\langchain_core\vectorstores.py", line 550, in from_documents
return cls.from_texts(texts, embedding, metadatas=metadatas, **kwargs)
File "C:\Users\Jacob.conda\envs\chatchat\lib\site-packages\langchain_community\vectorstores\faiss.py", line 930, in from_texts
embeddings = embedding.embed_documents(texts)
File "C:\Users\Jacob.conda\envs\chatchat\lib\site-packages\langchain_openai\embeddings\base.py", line 480, in embed_documents
return self._get_len_safe_embeddings(texts, engine=engine)
File "C:\Users\Jacob.conda\envs\chatchat\lib\site-packages\langchain_openai\embeddings\base.py", line 323, in _get_len_safe_embeddings
response = self.client.create(
File "C:\Users\Jacob.conda\envs\chatchat\lib\site-packages\openai\resources\embeddings.py", line 132, in create
return self._post(
File "C:\Users\Jacob.conda\envs\chatchat\lib\site-packages\openai_base_client.py", line 1259, in post
return cast(ResponseT, self.request(cast_to, opts, stream=stream, stream_cls=stream_cls))
File "C:\Users\Jacob.conda\envs\chatchat\lib\site-packages\openai_base_client.py", line 1047, in request
raise self._make_status_error_from_response(err.response) from None
openai.BadRequestError: Error code: 400 - {'error': {'message': '<400> InternalError.Algo.InvalidParameter: Value error, contents is neither str nor list of str.: input.contents', 'type': 'InvalidParameter', 'param': None, 'code': 'InvalidParameter'}, 'id': '928d36f0-6097-9f7f-964d-f521ae99bfa0', 'request_id': '928d36f0-6097-9f7f-964d-f521ae99bfa0'}

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "D:\llm_project\Langchain-Chatchat-master\libs\chatchat-server\chatchat\init_database.py", line 40, in worker
folder2db(
File "D:\llm_project\Langchain-Chatchat-master\libs\chatchat-server\chatchat\server\knowledge_base\migrate.py", line 156, in folder2db
kb.create_kb()
File "D:\llm_project\Langchain-Chatchat-master\libs\chatchat-server\chatchat\server\knowledge_base\kb_service\base.py", line 98, in create_kb
self.do_create_kb()
File "D:\llm_project\Langchain-Chatchat-master\libs\chatchat-server\chatchat\server\knowledge_base\kb_service\faiss_kb_service.py", line 57, in do_create_kb
self.load_vector_store()
File "D:\llm_project\Langchain-Chatchat-master\libs\chatchat-server\chatchat\server\knowledge_base\kb_service\faiss_kb_service.py", line 32, in load_vector_store
return kb_faiss_pool.load_vector_store(
File "D:\llm_project\Langchain-Chatchat-master\libs\chatchat-server\chatchat\server\knowledge_base\kb_cache\faiss_cache.py", line 140, in load_vector_store
logger.exception(e)
File "C:\Users\Jacob.conda\envs\chatchat\lib\site-packages\loguru_logger.py", line 2099, in exception
__self._log("ERROR", False, options, __message, args, kwargs)
File "C:\Users\Jacob.conda\envs\chatchat\lib\site-packages\loguru_logger.py", line 2051, in _log
colored_message = Colorizer.prepare_simple_message(str(message))
File "C:\Users\Jacob.conda\envs\chatchat\lib\site-packages\loguru_colorizer.py", line 378, in prepare_simple_message
parser.feed(string)
File "C:\Users\Jacob.conda\envs\chatchat\lib\site-packages\loguru_colorizer.py", line 260, in feed
raise ValueError(
ValueError: Tag "<400>" does not correspond to any known color directive, make sure you did not misspelled it (or prepend '' to escape it)

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "C:\Users\Jacob.conda\envs\chatchat\lib\multiprocessing\process.py", line 314, in _bootstrap
self.run()
File "C:\Users\Jacob.conda\envs\chatchat\lib\multiprocessing\process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "D:\llm_project\Langchain-Chatchat-master\libs\chatchat-server\chatchat\init_database.py", line 61, in worker
logger.exception(e)
File "C:\Users\Jacob.conda\envs\chatchat\lib\site-packages\loguru_logger.py", line 2099, in exception
__self._log("ERROR", False, options, __message, args, kwargs)
File "C:\Users\Jacob.conda\envs\chatchat\lib\site-packages\loguru_logger.py", line 2051, in _log
colored_message = Colorizer.prepare_simple_message(str(message))
File "C:\Users\Jacob.conda\envs\chatchat\lib\site-packages\loguru_colorizer.py", line 378, in prepare_simple_message
parser.feed(string)
File "C:\Users\Jacob.conda\envs\chatchat\lib\site-packages\loguru_colorizer.py", line 260, in feed
raise ValueError(
ValueError: Tag "<400>" does not correspond to any known color directive, make sure you did not misspelled it (or prepend '' to escape it)

[dosubot]

Hi @Jacob021219! I'm Dosu and I’m helping the Langchain-Chatchat team.

你的问题主要有两个核心点：

1. embedding模型API报400错误：你用的 DashScope embedding 模型名是 text-embedding-v3，但根据当前 Langchain-Chatchat 的配置，DashScope（oneapi）平台只支持 text-embedding-v1 和 Embedding-V1，没有 text-embedding-v3。模型名不匹配会导致 API 请求参数不被识别，从而报“contents is neither str nor list of str”错误。建议你把 embedding 模型名改成 text-embedding-v1 或 Embedding-V1，并确保 model_settings.yaml 里 DEFAULT_EMBEDDING_MODEL 和平台 embed_models 列表一致，且 DashScope 服务端已加载该模型。参考

2. Loguru 日志报 ValueError：API 返回的错误信息里有 <400> 这样的标签，Loguru 日志库会把它当成颜色指令，导致日志输出报错。这个问题不影响主流程，但会让日志输出异常。可以考虑把 logger.exception(e) 里的 e 先转成普通字符串再输出，或者临时关闭 Loguru 的彩色输出。参考

如果你确实需要用 text-embedding-v3，目前代码和配置未必支持，建议优先用官方支持的 embedding 模型。如果一定要用 v3，可以考虑自定义 embedding 类，确保传给 API 的 input 参数是字符串或字符串列表，避免格式错误。相关讨论和代码示例

总结：请先把 embedding 模型名改成 text-embedding-v1 或 Embedding-V1，并确保配置和服务端一致，绝大多数情况下能解决初始化失败问题。如果还有问题欢迎补充你的 model_settings.yaml 配置片段。

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}