求助：开发环境找不到 embbeding 模型

MaxKB
, ,
1

按照文档”https://maxkb.cn/docs/dev_manual/dev_environment/“搭建了开发环境，但索引一直失败，local_model 服务报错，我增加了一点日志，最终输出：

ERROR:root:--------------------------------
ERROR:root:model_name: /opt/maxkb/model/embedding/shibing624_text2vec-base-chinese
ERROR:root:model_credential: {'cache_folder': '/opt/maxkb/model/'}
ERROR:root:model_kwargs: {'model_id': '42f63a3d-427e-11ef-b3ec-a8a1595801ab', 'use_local': True, 'streaming': True}
ERROR:root:--------------------------------
2025-03-07 16:42:09 [handle_exception ERROR] stat: path should be string, bytes, os.PathLike or integer, not NoneType:Traceback (most recent call last):
  File "/var/gitrepo/MaxKB/venv/lib/python3.11/site-packages/rest_framework/views.py", line 506, in dispatch
    response = handler(request, *args, **kwargs)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/apps/setting/views/model_apply.py", line 29, in post
    ModelApplySerializers(data={'model_id': model_id}).embed_documents(request.data))
    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/apps/setting/serializers/model_apply_serializers.py", line 58, in embed_documents
    model = get_embedding_model(self.data.get('model_id'))
            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/apps/setting/serializers/model_apply_serializers.py", line 24, in get_embedding_model
    embedding_model = ModelManage.get_model(model_id,
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/apps/common/config/embedding_config.py", line 28, in get_model
    model_instance = get_model(_id)
                     ^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/apps/setting/serializers/model_apply_serializers.py", line 25, in <lambda>
    lambda _id: get_model(model, use_local=True))
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/apps/setting/models_provider/__init__.py", line 42, in get_model
    return get_model_(model.provider, model.model_type, model.model_name, model.credential, str(model.id), **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/apps/setting/models_provider/__init__.py", line 27, in get_model_
    model = get_provider(provider).get_model(model_type, model_name,
            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/apps/setting/models_provider/base_model_provider.py", line 83, in get_model
    return model_info.model_class.new_instance(model_type, model_name, model_credential, **model_kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/apps/setting/models_provider/impl/local_model_provider/model/embedding.py", line 61, in new_instance
    return LocalEmbedding(model_name=model_name, cache_folder='/opt/maxkb/model/embedding',
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/venv/lib/python3.11/site-packages/langchain_huggingface/embeddings/huggingface.py", line 61, in __init__
    self.client = sentence_transformers.SentenceTransformer(
                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/venv/lib/python3.11/site-packages/sentence_transformers/SentenceTransformer.py", line 197, in __init__
    modules = self._load_sbert_model(
              ^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/venv/lib/python3.11/site-packages/sentence_transformers/SentenceTransformer.py", line 1296, in _load_sbert_model
    module = Transformer(model_name_or_path, cache_dir=cache_folder, **kwargs)
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/venv/lib/python3.11/site-packages/sentence_transformers/models/Transformer.py", line 38, in __init__
    self.tokenizer = AutoTokenizer.from_pretrained(
                     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/venv/lib/python3.11/site-packages/transformers/models/auto/tokenization_auto.py", line 944, in from_pretrained
    return tokenizer_class.from_pretrained(pretrained_model_name_or_path, *inputs, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/venv/lib/python3.11/site-packages/transformers/tokenization_utils_base.py", line 2052, in from_pretrained
    return cls._from_pretrained(
           ^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/venv/lib/python3.11/site-packages/transformers/tokenization_utils_base.py", line 2090, in _from_pretrained
    slow_tokenizer = (cls.slow_tokenizer_class)._from_pretrained(
                     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/venv/lib/python3.11/site-packages/transformers/tokenization_utils_base.py", line 2292, in _from_pretrained
    tokenizer = cls(*init_inputs, **init_kwargs)
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/venv/lib/python3.11/site-packages/transformers/models/bert/tokenization_bert.py", line 114, in __init__
    if not os.path.isfile(vocab_file):
           ^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "<frozen genericpath>", line 30, in isfile
TypeError: stat: path should be string, bytes, os.PathLike or integer, not NoneType

2025-03-07 16:42:09 [basehttp INFO] "POST /api/model/42f63a3d-427e-11ef-b3ec-a8a1595801ab/embed_documents HTTP/1.1" 200 114
ERROR:root:--------------------------------
ERROR:root:model_name: /opt/maxkb/model/embedding/shibing624_text2vec-base-chinese
ERROR:root:model_credential: {'cache_folder': '/opt/maxkb/model/'}
ERROR:root:model_kwargs: {'model_id': '42f63a3d-427e-11ef-b3ec-a8a1595801ab', 'use_local': True, 'streaming': True}
ERROR:root:--------------------------------
2025-03-07 16:42:10 [handle_exception ERROR] stat: path should be string, bytes, os.PathLike or integer, not NoneType:Traceback (most recent call last):
  File "/var/gitrepo/MaxKB/venv/lib/python3.11/site-packages/rest_framework/views.py", line 506, in dispatch
    response = handler(request, *args, **kwargs)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/apps/setting/views/model_apply.py", line 29, in post
    ModelApplySerializers(data={'model_id': model_id}).embed_documents(request.data))
    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/apps/setting/serializers/model_apply_serializers.py", line 58, in embed_documents
    model = get_embedding_model(self.data.get('model_id'))
            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/apps/setting/serializers/model_apply_serializers.py", line 24, in get_embedding_model
    embedding_model = ModelManage.get_model(model_id,
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/apps/common/config/embedding_config.py", line 28, in get_model
    model_instance = get_model(_id)
                     ^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/apps/setting/serializers/model_apply_serializers.py", line 25, in <lambda>
    lambda _id: get_model(model, use_local=True))
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/apps/setting/models_provider/__init__.py", line 42, in get_model
    return get_model_(model.provider, model.model_type, model.model_name, model.credential, str(model.id), **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/apps/setting/models_provider/__init__.py", line 27, in get_model_
    model = get_provider(provider).get_model(model_type, model_name,
            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/apps/setting/models_provider/base_model_provider.py", line 83, in get_model
    return model_info.model_class.new_instance(model_type, model_name, model_credential, **model_kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/apps/setting/models_provider/impl/local_model_provider/model/embedding.py", line 61, in new_instance
    return LocalEmbedding(model_name=model_name, cache_folder='/opt/maxkb/model/embedding',
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/venv/lib/python3.11/site-packages/langchain_huggingface/embeddings/huggingface.py", line 61, in __init__
    self.client = sentence_transformers.SentenceTransformer(
                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/venv/lib/python3.11/site-packages/sentence_transformers/SentenceTransformer.py", line 197, in __init__
    modules = self._load_sbert_model(
              ^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/venv/lib/python3.11/site-packages/sentence_transformers/SentenceTransformer.py", line 1296, in _load_sbert_model
    module = Transformer(model_name_or_path, cache_dir=cache_folder, **kwargs)
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/venv/lib/python3.11/site-packages/sentence_transformers/models/Transformer.py", line 38, in __init__
    self.tokenizer = AutoTokenizer.from_pretrained(
                     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/venv/lib/python3.11/site-packages/transformers/models/auto/tokenization_auto.py", line 944, in from_pretrained
    return tokenizer_class.from_pretrained(pretrained_model_name_or_path, *inputs, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/venv/lib/python3.11/site-packages/transformers/tokenization_utils_base.py", line 2052, in from_pretrained
    return cls._from_pretrained(
           ^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/venv/lib/python3.11/site-packages/transformers/tokenization_utils_base.py", line 2090, in _from_pretrained
    slow_tokenizer = (cls.slow_tokenizer_class)._from_pretrained(
                     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/venv/lib/python3.11/site-packages/transformers/tokenization_utils_base.py", line 2292, in _from_pretrained
    tokenizer = cls(*init_inputs, **init_kwargs)
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/var/gitrepo/MaxKB/venv/lib/python3.11/site-packages/transformers/models/bert/tokenization_bert.py", line 114, in __init__
    if not os.path.isfile(vocab_file):
           ^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "<frozen genericpath>", line 30, in isfile
TypeError: stat: path should be string, bytes, os.PathLike or integer, not NoneType

很奇怪的是，明明传了数据，为什么提示说 path 是 None 呢？

另外，我在第一次启动服务之前，/opt/maxkb/conf 下面的配置文件是这样配的：

EMBEDDING_MODEL_PATH: /opt/maxkb/model/embedding
EMBEDDING_MODEL_NAME: shibing624_text2vec-base-chinese

因为我看文档里写的是：

# 模型名称:如果模型名称是路径,则会加载目录下的模型,如果是模型名称,则会在https://huggingface.co/下载模型 模型的下载位置为EMBEDDING_MODEL_PATH

以为不配路径，会自动下载模型，然后并没有。后面我调用 installer/install_model.py 下载的模型，目前 /opt/maxkb 下面的目录结构如下：

.
├── conf
│   └── config_example.yml
└── model
    ├── base
    │   └── hub
    └── embedding
        └── shibing624_text2vec-base-chinese

然后 maxkb 模型的配置是这样的：

WX20250307-231507@2x
WX20250307-231507@2x1210×1160 66.1 KB

感觉一切都对，但索引就是出错，保存模型配置也会报错：

Verification failed, please check whether the parameters are correct: stat: path should be string, bytes, os.PathLike or integer, not NoneType
2

这些服务都启动了么？
异步任务： python main.py dev celery
本地模型服务: python main.py dev local_model
web服务: python main.py dev

1 个赞
3

源码部署相关的错误需要自己多研究下，官方没有资源一一排查，实在搞不定可以使用官方标准安装包部署。