bisheng-langchain 0.3.3b0__py3-none-any.whl → 0.3.3.dev2__py3-none-any.whl

bisheng_langchain/chains/qa_generation/base_v2.py

@@ -10,6 +10,7 @@ from collections import defaultdict, namedtuple
 from dataclasses import dataclass
 from langchain_core.callbacks import CallbackManagerForChainRun
 from langchain_core.language_models import BaseLanguageModel
+from langchain_core.prompts import HumanMessagePromptTemplate, PromptTemplate
 
 try:
     from llama_index.node_parser import SimpleNodeParser
@@ -133,6 +134,9 @@ class TrainsetGenerator:
             chunk_size: int = 1024,
             seed: int = 42,
             prompt: Optional[ChatPromptTemplate] = SEED_QUESTION_CHAT_PROMPT,
+            filter_lowquality_context: bool = False,
+            filter_lowquality_question: bool = False,
+            answer_prompt: Optional[HumanMessagePromptTemplate] = ANSWER_FORMULATE,
     ) -> None:
         self.generator_llm = generator_llm
         self.critic_llm = critic_llm
@@ -150,6 +154,11 @@ class TrainsetGenerator:
         self.threshold = 5.0
         self.rng = default_rng(seed)
         self.prompt = prompt
+        self.filter_lowquality_context = filter_lowquality_context
+        self.filter_lowquality_question = filter_lowquality_question
+        if answer_prompt is None:
+            answer_prompt = ANSWER_FORMULATE
+        self.answer_prompt = answer_prompt
 
     @classmethod
     def from_default(
@@ -158,6 +167,9 @@ class TrainsetGenerator:
             chunk_size: int = 512,
             trainset_distribution: dict = DEFAULT_TRAIN_DISTRIBUTION,
             prompt: Optional[ChatPromptTemplate] = SEED_QUESTION_CHAT_PROMPT,
+            filter_lowquality_context: bool = False,
+            filter_lowquality_question: bool = False,
+            answer_prompt: Optional[PromptTemplate] = ANSWER_FORMULATE,
     ):
         generator_llm = llm
         critic_llm = llm
@@ -167,6 +179,9 @@ class TrainsetGenerator:
             chunk_size=chunk_size,
             trainset_distribution=trainset_distribution,
             prompt=prompt,
+            filter_lowquality_context=filter_lowquality_context,
+            filter_lowquality_question=filter_lowquality_question,
+            answer_prompt=answer_prompt,
         )
 
     def _get_evolve_type(self) -> str:
@@ -221,7 +236,7 @@ class TrainsetGenerator:
 
     def _generate_answer(self, question: str, context: t.List[str]) -> t.List[str]:
         return [
-            self._qc_template(ANSWER_FORMULATE, qstn, context[i])
+            self._qc_template(self.answer_prompt, qstn, context[i])
             for i, qstn in enumerate(question.split("\n"))
         ]
 
@@ -309,14 +324,17 @@ class TrainsetGenerator:
             )
 
             text_chunk = " ".join([node.get_content() for node in nodes])
-            score = self._filter_context(text_chunk)
-            if not score:
-                continue
+            if self.filter_lowquality_context:
+                score = self._filter_context(text_chunk)
+                if not score:
+                    continue
             seed_question = self._seed_question(text_chunk)
 
             question = seed_question
-            # is_valid_question = self._filter_question(question)
-            is_valid_question = True
+            if self.filter_lowquality_question:
+                is_valid_question = self._filter_question(question)
+            else:
+                is_valid_question = True
             if is_valid_question:
                 context = [text_chunk] * len(question.split("\n"))
                 is_conv = len(context) > 1
@@ -354,7 +372,10 @@ class QAGenerationChainV2(Chain):
             llm: BaseLanguageModel,
             k: Optional[int] = None,
             chunk_size: int = 512,
-            prompt: Optional[ChatPromptTemplate] = SEED_QUESTION_CHAT_PROMPT,
+            filter_lowquality_context: bool = False,
+            filter_lowquality_question: bool = False,
+            question_prompt: Optional[ChatPromptTemplate] = SEED_QUESTION_CHAT_PROMPT,
+            answer_prompt: Optional[HumanMessagePromptTemplate] = ANSWER_FORMULATE,
             **kwargs: Any,
     ) -> QAGenerationChainV2:
         """
@@ -362,13 +383,21 @@ class QAGenerationChainV2(Chain):
 
         Args:
             llm: a language model
-            prompt: a prompt template
+            question_prompt: a prompt template for generate question
+            answer_prompt: a prompt template for generate answer
             **kwargs: additional arguments
 
         Returns:
             a QAGenerationChain class
         """
-        generator = TrainsetGenerator.from_default(llm, chunk_size=chunk_size, prompt=prompt)
+        generator = TrainsetGenerator.from_default(
+            llm, 
+            chunk_size=chunk_size, 
+            prompt=question_prompt, 
+            answer_prompt=answer_prompt,
+            filter_lowquality_context=filter_lowquality_context, 
+            filter_lowquality_question=filter_lowquality_question
+        )
         return cls(documents=documents, generator=generator, k=k, **kwargs)
 
     @property
@@ -395,13 +424,14 @@ class QAGenerationChainV2(Chain):
         dataset = self.generator.generate(documents=self.documents, train_size=self.k)
         df = dataset.to_pandas()
         qa_pairs = df.to_dict("records")
-        qa = ''
+        qa = []
         for pair in qa_pairs:
-            qa += json.dumps(
-                {
-                    "question": pair["question"],
-                    "answer": pair["ground_truth"][0],
-                }, ensure_ascii=False)
+            qa.append({
+                "question": pair["question"],
+                "answer": pair["ground_truth"][0],
+                "context": pair["ground_truth_context"][0],
+            })
+        qa = f'```json\n{json.dumps(qa, ensure_ascii=False, indent=4)}\n```'
         return {self.output_key: qa}
 
     async def _acall(

bisheng_langchain/chat_models/proxy_llm.py

@@ -276,7 +276,7 @@ class ProxyChatLLM(BaseChatModel):
             function_call: Optional[dict] = None
             async for stream_resp in self.acompletion_with_retry(messages=message_dicts, **params):
 
-                role = stream_resp['choices'][0]['delta'].get('role', role)
+                role = stream_resp['choices'][0]['delta'].get('role') or role
                 token = stream_resp['choices'][0]['delta'].get('content', '')
                 inner_completion += token or ''
                 _function_call = stream_resp['choices'][0]['delta'].get('function_call')

bisheng_langchain/document_loaders/elem_unstrcutured_loader.py

@@ -87,7 +87,8 @@ class ElemUnstructuredLoader(BasePDFLoader):
 
         resp = requests.post(self.unstructured_api_url, headers=self.headers, json=payload)
         if resp.status_code != 200:
-            raise Exception(f'file partition {os.path.basename(self.file_name)} failed resp={resp.text}')
+            raise Exception(
+                f'file partition {os.path.basename(self.file_name)} failed resp={resp.text}')
 
         resp = resp.json()
         if 200 != resp.get('status_code'):
@@ -97,6 +98,10 @@ class ElemUnstructuredLoader(BasePDFLoader):
             logger.info(f'partition_error resp={resp}')
         logger.info(f'unstruct_return code={resp.get("status_code")}')
 
+        if resp.get('b64_pdf'):
+            with open(self.file_path, 'wb') as f:
+                f.write(base64.b64decode(resp['b64_pdf']))
+
         content, metadata = merge_partitions(partitions)
         metadata['source'] = self.file_name
 
@@ -144,27 +149,33 @@ class ElemUnstructuredLoaderV0(BasePDFLoader):
             return res['text'], {'source': self.file_name}
         # 说明文件解析失败，pdf文件直接返回报错
         if self.file_name.endswith('.pdf'):
-            raise Exception(f'file text {os.path.basename(self.file_name)} failed resp={resp.text}')
+            raise Exception(
+                f'file text {os.path.basename(self.file_name)} failed resp={resp.text}')
         # 非pdf文件，先将文件转为pdf格式，让后再执行partition模式解析文档
         # 把文件转为pdf
-        resp = requests.post(self.unstructured_api_url, headers=self.headers, json={
-            'filename': os.path.basename(self.file_name),
-            'b64_data': [b64_data],
-            'mode': 'topdf',
-        })
+        resp = requests.post(self.unstructured_api_url,
+                             headers=self.headers,
+                             json={
+                                 'filename': os.path.basename(self.file_name),
+                                 'b64_data': [b64_data],
+                                 'mode': 'topdf',
+                             })
         if resp.status_code != 200 or resp.json().get('status_code') != 200:
-            raise Exception(f'file topdf {os.path.basename(self.file_name)} failed resp={resp.text}')
+            raise Exception(
+                f'file topdf {os.path.basename(self.file_name)} failed resp={resp.text}')
         # 解析pdf文件
         payload['mode'] = 'partition'
         payload['b64_data'] = [resp.json()['b64_pdf']]
         payload['filename'] = os.path.basename(self.file_name) + '.pdf'
         resp = requests.post(self.unstructured_api_url, headers=self.headers, json=payload)
         if resp.status_code != 200 or resp.json().get('status_code') != 200:
-            raise Exception(f'file partition {os.path.basename(self.file_name)} failed resp={resp.text}')
+            raise Exception(
+                f'file partition {os.path.basename(self.file_name)} failed resp={resp.text}')
         res = resp.json()
         partitions = res['partitions']
         if not partitions:
-            raise Exception(f'file partition empty {os.path.basename(self.file_name)} resp={resp.text}')
+            raise Exception(
+                f'file partition empty {os.path.basename(self.file_name)} resp={resp.text}')
         # 拼接结果为文本
         content, _ = merge_partitions(partitions)
         return content, {'source': self.file_name}

bisheng_langchain/gpts/assistant.py

@@ -131,7 +131,7 @@ if __name__ == "__main__":
 
     # set_debug(True)
     # chat_history = []
-    # query = "600519、300750股价多少？"
+    # query = "分析当日市场行情"
     chat_history = ['你好', '你好，有什么可以帮助你吗？', '福蓉科技股价多少?', '福蓉科技（股票代码：300049）的当前股价为48.67元。']
     query = '今天是什么时候？去年这个时候的股价是多少？'
     bisheng_assistant = BishengAssistant("config/base_scene.yaml")

bisheng_langchain/gpts/prompts/assistant_prompt_base.py

@@ -1 +1,19 @@
-ASSISTANT_PROMPT_DEFAULT = "You are a helpful assistant."
+ASSISTANT_PROMPT_DEFAULT = "You are a helpful assistant."
+
+# ASSISTANT_PROMPT_DEFAULT = """
+# # role
+# 你是一名股票市场研究员。
+
+# # skill 1：股票市场行情分析
+# 当用户要求分析当日市场行情时，以表格形式展示当日股市情况：
+# - 行：沪深300、中证500、中证1000 以及 小计
+# - 列指标：当日涨跌幅、今年累计涨跌幅、成交额(单位：亿元)、成交额占比
+# - 小计 说明：对"成交额"列进行求和，其他列为空即可
+# - 指标计算说明：当日涨跌幅=实时行情数据涨跌幅; 今年累计涨跌幅=(实时行情数据现价-今年第一个交易日历史行情数据开盘价)/今年第一个交易日历史行情数据开盘价*100%; 成交额 = 实时行情数据成交额； 成交额占比=该指数占所有所有指数成交额的百分比。
+
+# 为了完成上述任务，你可以调用以下工具：
+# 1. 股票实时行情工具：获取当日现价、涨跌幅、成交额
+# 2. 股票历史行情工具：获取历史某日开盘价
+
+# 附：指数代码：沪深300(000300)、中证500(000905)、中证1000(000852)
+# """

bisheng_langchain/gpts/prompts/react_agent_prompt.py

@@ -23,10 +23,10 @@ Valid "action" values: "Final Answer" or {tool_names}
 Provide only ONE action per $JSON_BLOB, as shown:
 
 ```
-{{{{
+{{
   "action": $TOOL_NAME,
   "action_input": $INPUT
-}}}}
+}}
 ```
 
 Follow this format:
@@ -42,10 +42,10 @@ Observation: action result
 Thought: I know what to respond
 Action:
 ```
-{{{{
+{{
   "action": "Final Answer",
   "action_input": "Final response to human"
-}}}}
+}}
 
 Begin! Reminder to ALWAYS respond with a valid json blob of a single action. Use tools if necessary. Respond directly if appropriate. Format is Action:```$JSON_BLOB```then Observation
 """

bisheng_langchain/rag/init_retrievers/keyword_retriever.py

@@ -16,7 +16,7 @@ from langchain.text_splitter import TextSplitter
 
 
 class KeywordRetriever(BaseRetriever):
-    keyword_store: ElasticKeywordsSearch
+    keyword_store: VectorStore
     text_splitter: TextSplitter
     search_type: str = 'similarity'
     search_kwargs: dict = Field(default_factory=dict)

bisheng_langchain/utils/azure_dalle_image_generator.py

@@ -0,0 +1,140 @@
+import os
+from typing import Callable, Dict, Optional, Union
+
+import openai
+from langchain_community.utilities.dalle_image_generator import DallEAPIWrapper
+from langchain_core.pydantic_v1 import Field, SecretStr, root_validator
+from langchain_core.utils import convert_to_secret_str, get_from_dict_or_env
+
+
+class AzureDallEWrapper(DallEAPIWrapper):
+    """`Azure OpenAI` Embeddings API.
+
+    To use, you should have the
+    environment variable ``AZURE_OPENAI_API_KEY`` set with your API key or pass it
+    as a named parameter to the constructor.
+
+    Example:
+        .. code-block:: python
+
+            from langchain_openai import AzureOpenAIEmbeddings
+
+            openai = AzureOpenAIEmbeddings(model="text-embedding-3-large")
+    """
+
+    azure_endpoint: Union[str, None] = None
+    """Your Azure endpoint, including the resource.
+
+        Automatically inferred from env var `AZURE_OPENAI_ENDPOINT` if not provided.
+
+        Example: `https://example-resource.azure.openai.com/`
+    """
+    deployment: Optional[str] = Field(default=None, alias='azure_deployment')
+    """A model deployment.
+
+        If given sets the base client URL to include `/deployments/{azure_deployment}`.
+        Note: this means you won't be able to use non-deployment endpoints.
+    """
+    openai_api_key: Optional[SecretStr] = Field(default=None, alias='api_key')
+    """Automatically inferred from env var `AZURE_OPENAI_API_KEY` if not provided."""
+    azure_ad_token: Optional[SecretStr] = None
+    """Your Azure Active Directory token.
+
+        Automatically inferred from env var `AZURE_OPENAI_AD_TOKEN` if not provided.
+
+        For more:
+        https://www.microsoft.com/en-us/security/business/identity-access/microsoft-entra-id.
+    """
+    azure_ad_token_provider: Union[Callable[[], str], None] = None
+    """A function that returns an Azure Active Directory token.
+
+        Will be invoked on every request.
+    """
+    openai_api_version: Optional[str] = Field(default=None, alias='api_version')
+    """Automatically inferred from env var `OPENAI_API_VERSION` if not provided."""
+    validate_base_url: bool = True
+    chunk_size: int = 2048
+    """Maximum number of texts to embed in each batch"""
+
+    @root_validator()
+    def validate_environment(cls, values: Dict) -> Dict:
+        """Validate that api key and python package exists in environment."""
+        # Check OPENAI_KEY for backwards compatibility.
+        # TODO: Remove OPENAI_API_KEY support to avoid possible conflict when using
+        # other forms of azure credentials.
+        openai_api_key = (values['openai_api_key'] or os.getenv('AZURE_OPENAI_API_KEY')
+                          or os.getenv('OPENAI_API_KEY'))
+        values['openai_api_key'] = (convert_to_secret_str(openai_api_key)
+                                    if openai_api_key else None)
+        values['openai_api_base'] = (values['openai_api_base'] if 'openai_api_base' in values else
+                                     os.getenv('OPENAI_API_BASE'))
+        values['openai_api_version'] = values['openai_api_version'] or os.getenv(
+            'OPENAI_API_VERSION', default='2023-05-15')
+        values['openai_api_type'] = get_from_dict_or_env(values,
+                                                         'openai_api_type',
+                                                         'OPENAI_API_TYPE',
+                                                         default='azure')
+        values['openai_organization'] = (values['openai_organization']
+                                         or os.getenv('OPENAI_ORG_ID')
+                                         or os.getenv('OPENAI_ORGANIZATION'))
+        values['openai_proxy'] = get_from_dict_or_env(values,
+                                                      'openai_proxy',
+                                                      'OPENAI_PROXY',
+                                                      default='')
+        values['azure_endpoint'] = values['azure_endpoint'] or os.getenv('AZURE_OPENAI_ENDPOINT')
+        azure_ad_token = values['azure_ad_token'] or os.getenv('AZURE_OPENAI_AD_TOKEN')
+        values['azure_ad_token'] = (convert_to_secret_str(azure_ad_token)
+                                    if azure_ad_token else None)
+        # For backwards compatibility. Before openai v1, no distinction was made
+        # between azure_endpoint and base_url (openai_api_base).
+        openai_api_base = values['openai_api_base']
+        if openai_api_base and values['validate_base_url']:
+            if '/openai' not in openai_api_base:
+                values['openai_api_base'] += '/openai'
+                raise ValueError('As of openai>=1.0.0, Azure endpoints should be specified via '
+                                 'the `azure_endpoint` param not `openai_api_base` '
+                                 '(or alias `base_url`). ')
+            if values['deployment']:
+                raise ValueError('As of openai>=1.0.0, if `deployment` (or alias '
+                                 '`azure_deployment`) is specified then '
+                                 '`openai_api_base` (or alias `base_url`) should not be. '
+                                 'Instead use `deployment` (or alias `azure_deployment`) '
+                                 'and `azure_endpoint`.')
+        client_params = {
+            'api_version':
+            values['openai_api_version'],
+            'azure_endpoint':
+            values['azure_endpoint'],
+            'azure_deployment':
+            values['deployment'],
+            'api_key':
+            (values['openai_api_key'].get_secret_value() if values['openai_api_key'] else None),
+            'azure_ad_token':
+            (values['azure_ad_token'].get_secret_value() if values['azure_ad_token'] else None),
+            'azure_ad_token_provider':
+            values['azure_ad_token_provider'],
+            'organization':
+            values['openai_organization'],
+            'base_url':
+            values['openai_api_base'],
+            'timeout':
+            values['request_timeout'],
+            'max_retries':
+            values['max_retries'],
+            'default_headers':
+            values['default_headers'],
+            'default_query':
+            values['default_query'],
+        }
+        if not values.get('client'):
+            sync_specific = {'http_client': values['http_client']}
+            values['client'] = openai.AzureOpenAI(**client_params, **sync_specific).images
+        if not values.get('async_client'):
+            async_specific = {'http_client': values['http_async_client']}
+            values['async_client'] = openai.AsyncAzureOpenAI(**client_params,
+                                                             **async_specific).images
+        return values
+
+    @property
+    def _llm_type(self) -> str:
+        return 'azure-openai-chat'

bisheng_langchain/vectorstores/milvus.py

@@ -105,6 +105,7 @@ class Milvus(MilvusLangchain):
                  index_params: Optional[dict] = None,
                  search_params: Optional[dict] = None,
                  drop_old: Optional[bool] = False,
+                 partition_key: Optional[str] = None,
                  *,
                  primary_field: str = 'pk',
                  text_field: str = 'text',
@@ -195,6 +196,7 @@ class Milvus(MilvusLangchain):
         self._vector_field = vector_field
         #  partion key for multi-tenancy
         self._partition_field = partition_field
+        self.partition_key = partition_key
 
         self.fields: list[str] = []
         # Create the connection to the server
@@ -438,13 +440,13 @@ class Milvus(MilvusLangchain):
             self.col.load()
 
     def add_texts(
-        self,
-        texts: Iterable[str],
-        metadatas: Optional[List[dict]] = None,
-        timeout: Optional[int] = None,
-        batch_size: int = 1000,
-        no_embedding: bool = False,
-        **kwargs: Any,
+            self,
+            texts: Iterable[str],
+            metadatas: Optional[List[dict]] = None,
+            timeout: Optional[int] = None,
+            batch_size: int = 1000,
+            no_embedding: bool = False,
+            **kwargs: Any,
     ) -> List[str]:
         """Insert text data into Milvus.
 
@@ -529,13 +531,13 @@ class Milvus(MilvusLangchain):
         return pks
 
     def similarity_search(
-        self,
-        query: str,
-        k: int = 4,
-        param: Optional[dict] = None,
-        expr: Optional[str] = None,
-        timeout: Optional[int] = None,
-        **kwargs: Any,
+            self,
+            query: str,
+            k: int = 4,
+            param: Optional[dict] = None,
+            expr: Optional[str] = None,
+            timeout: Optional[int] = None,
+            **kwargs: Any,
     ) -> List[Document]:
         """Perform a similarity search against the query string.
 
@@ -567,13 +569,13 @@ class Milvus(MilvusLangchain):
         return [doc for doc, _ in res]
 
     def similarity_search_by_vector(
-        self,
-        embedding: List[float],
-        k: int = 4,
-        param: Optional[dict] = None,
-        expr: Optional[str] = None,
-        timeout: Optional[int] = None,
-        **kwargs: Any,
+            self,
+            embedding: List[float],
+            k: int = 4,
+            param: Optional[dict] = None,
+            expr: Optional[str] = None,
+            timeout: Optional[int] = None,
+            **kwargs: Any,
     ) -> List[Document]:
         """Perform a similarity search against the query string.
 
@@ -605,13 +607,13 @@ class Milvus(MilvusLangchain):
         return [doc for doc, _ in res]
 
     def similarity_search_with_score(
-        self,
-        query: str,
-        k: int = 4,
-        param: Optional[dict] = None,
-        expr: Optional[str] = None,
-        timeout: Optional[int] = None,
-        **kwargs: Any,
+            self,
+            query: str,
+            k: int = 4,
+            param: Optional[dict] = None,
+            expr: Optional[str] = None,
+            timeout: Optional[int] = None,
+            **kwargs: Any,
     ) -> List[Tuple[Document, float]]:
         """Perform a search on a query string and return results with score.
 
@@ -651,13 +653,13 @@ class Milvus(MilvusLangchain):
         return res
 
     def similarity_search_with_score_by_vector(
-        self,
-        embedding: List[float],
-        k: int = 4,
-        param: Optional[dict] = None,
-        expr: Optional[str] = None,
-        timeout: Optional[int] = None,
-        **kwargs: Any,
+            self,
+            embedding: List[float],
+            k: int = 4,
+            param: Optional[dict] = None,
+            expr: Optional[str] = None,
+            timeout: Optional[int] = None,
+            **kwargs: Any,
     ) -> List[Tuple[Document, float]]:
         """Perform a search on a query string and return results with score.
 
@@ -692,12 +694,12 @@ class Milvus(MilvusLangchain):
         output_fields = self.fields[:]
         output_fields.remove(self._vector_field)
         # partition for multi-tenancy
-        if 'partition_key' in kwargs:
+        if self.partition_key:
             # add parttion
             if expr:
-                expr = f"{expr} and {self._partition_field}==\"{kwargs['partition_key']}\""
+                expr = f"{expr} and {self._partition_field}==\"{self.partition_key}\""
             else:
-                expr = f"{self._partition_field}==\"{kwargs['partition_key']}\""
+                expr = f"{self._partition_field}==\"{self.partition_key}\""
 
         # Perform the search.
         res = self.col.search(
@@ -721,15 +723,15 @@ class Milvus(MilvusLangchain):
         return ret
 
     def max_marginal_relevance_search(
-        self,
-        query: str,
-        k: int = 4,
-        fetch_k: int = 20,
-        lambda_mult: float = 0.5,
-        param: Optional[dict] = None,
-        expr: Optional[str] = None,
-        timeout: Optional[int] = None,
-        **kwargs: Any,
+            self,
+            query: str,
+            k: int = 4,
+            fetch_k: int = 20,
+            lambda_mult: float = 0.5,
+            param: Optional[dict] = None,
+            expr: Optional[str] = None,
+            timeout: Optional[int] = None,
+            **kwargs: Any,
     ) -> List[Document]:
         """Perform a search and return results that are reordered by MMR.
 
@@ -774,15 +776,15 @@ class Milvus(MilvusLangchain):
         )
 
     def max_marginal_relevance_search_by_vector(
-        self,
-        embedding: list[float],
-        k: int = 4,
-        fetch_k: int = 20,
-        lambda_mult: float = 0.5,
-        param: Optional[dict] = None,
-        expr: Optional[str] = None,
-        timeout: Optional[int] = None,
-        **kwargs: Any,
+            self,
+            embedding: list[float],
+            k: int = 4,
+            fetch_k: int = 20,
+            lambda_mult: float = 0.5,
+            param: Optional[dict] = None,
+            expr: Optional[str] = None,
+            timeout: Optional[int] = None,
+            **kwargs: Any,
     ) -> List[Document]:
         """Perform a search and return results that are reordered by MMR.
 
@@ -869,18 +871,18 @@ class Milvus(MilvusLangchain):
 
     @classmethod
     def from_texts(
-        cls,
-        texts: List[str],
-        embedding: Embeddings,
-        metadatas: Optional[List[dict]] = None,
-        collection_name: str = 'LangChainCollection',
-        connection_args: dict[str, Any] = DEFAULT_MILVUS_CONNECTION,
-        consistency_level: str = 'Session',
-        index_params: Optional[dict] = None,
-        search_params: Optional[dict] = None,
-        drop_old: bool = False,
-        no_embedding: bool = False,
-        **kwargs: Any,
+            cls,
+            texts: List[str],
+            embedding: Embeddings,
+            metadatas: Optional[List[dict]] = None,
+            collection_name: str = 'LangChainCollection',
+            connection_args: dict[str, Any] = DEFAULT_MILVUS_CONNECTION,
+            consistency_level: str = 'Session',
+            index_params: Optional[dict] = None,
+            search_params: Optional[dict] = None,
+            drop_old: bool = False,
+            no_embedding: bool = False,
+            **kwargs: Any,
     ) -> Milvus:
         """Create a Milvus collection, indexes it with HNSW, and insert data.
 

{bisheng_langchain-0.3.3b0.dist-info → bisheng_langchain-0.3.3.dev2.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: bisheng-langchain
-Version: 0.3.3b0
+Version: 0.3.3.dev2
 Summary: bisheng langchain modules
 Home-page: https://github.com/dataelement/bisheng
 Author: DataElem

{bisheng_langchain-0.3.3b0.dist-info → bisheng_langchain-0.3.3.dev2.dist-info}/RECORD

@@ -23,7 +23,7 @@ bisheng_langchain/chains/conversational_retrieval/__init__.py,sha256=47DEQpj8HBS
 bisheng_langchain/chains/conversational_retrieval/base.py,sha256=XiqBqov6No-wTVCou6qyMT5p2JQgoQI7OLQOYH8XUos,5313
 bisheng_langchain/chains/qa_generation/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 bisheng_langchain/chains/qa_generation/base.py,sha256=VYGmLDB0bnlDQ6T8ivLP55wwFbMo9HOzlPEDUuRx5fU,4148
-bisheng_langchain/chains/qa_generation/base_v2.py,sha256=v16zyj1-f17SpibREyO1gz0b7FdJDyTOYmqV1FVFFbY,13839
+bisheng_langchain/chains/qa_generation/base_v2.py,sha256=2F2kGe3ermJraQu4oC-m8vm_ENBy_Zi4uHrJDcSOeJw,15460
 bisheng_langchain/chains/qa_generation/prompt.py,sha256=4eJk9aDUYDN1qaaYRPy9EobCIncnwS8BbQaDFzzePtM,1944
 bisheng_langchain/chains/qa_generation/prompt_v2.py,sha256=sQLanA_iOnLqrUIwzfTOTANt-1vJ44CM54HFDU8Jo1Q,8938
 bisheng_langchain/chains/question_answering/__init__.py,sha256=_gOZMc-SWprK6xc-Jj64jcr9nc-G4YkZbEYwfJNq_bY,8795
@@ -35,7 +35,7 @@ bisheng_langchain/chains/router/rule_router.py,sha256=R2YRUnwn7s_7DbsSn27uPn4cIV
 bisheng_langchain/chat_models/__init__.py,sha256=4-HTLE_SXO4hmNJu6yQxiQKBt2IFca_ezllVBLmvbEE,635
 bisheng_langchain/chat_models/host_llm.py,sha256=35_jTdUm85mk-t2MARZYGC8dIPVtf5XXlGfFE6hQ1Gc,23153
 bisheng_langchain/chat_models/minimax.py,sha256=JLs_f6vWD9beZYUtjD4FG28G8tZHrGUAWOwdLIuJomw,13901
-bisheng_langchain/chat_models/proxy_llm.py,sha256=wzVBZik9WC3-f7kyQ1eu3Ooibqpcocln08knf5lV1Nw,17082
+bisheng_langchain/chat_models/proxy_llm.py,sha256=3ZCWU4n4eEMQcWiZt1BgD6DuVlbW8IyZSWGPsRunQco,17084
 bisheng_langchain/chat_models/qwen.py,sha256=W73KxDRQBUZEzttEM4K7ZzPqbN-82O6YQmpX-HB_wZU,19971
 bisheng_langchain/chat_models/sensetime.py,sha256=fuQ5yYGO5F7o7iQ7us17MlL4TAWRRFCCpNN9bAF-ydc,17056
 bisheng_langchain/chat_models/wenxin.py,sha256=OBXmFWkUWZMu1lUz6hPAEawsbAcdgMWcm9WkJJLZyng,13671
@@ -54,7 +54,7 @@ bisheng_langchain/document_loaders/custom_kv.py,sha256=xWUPhcr1hjbdya4zgEHG4Fl0s
 bisheng_langchain/document_loaders/elem_html.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 bisheng_langchain/document_loaders/elem_image.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 bisheng_langchain/document_loaders/elem_pdf.py,sha256=K-TXILGNFLFjavhun_MFbUF4t2_WGA3Z-kbnr75lmW8,22243
-bisheng_langchain/document_loaders/elem_unstrcutured_loader.py,sha256=ODAveZ1Rbq9SOO_qyIgH_NL8L2QqUZXOqf3wCKGYaRE,6984
+bisheng_langchain/document_loaders/elem_unstrcutured_loader.py,sha256=FtoyfmE85CwZuMvr52_bqcHQCgypKCWMGwLZrzgQYbY,7353
 bisheng_langchain/document_loaders/universal_kv.py,sha256=ZdIgFIc2fH2kkvJNb7j2wi6FLS_PaaatVy6z_YNV2hw,4114
 bisheng_langchain/document_loaders/parsers/__init__.py,sha256=OOM_FJkwaU-zNS58fASw0TH8FNT6VXKb0VrvisgdrII,171
 bisheng_langchain/document_loaders/parsers/ellm_client.py,sha256=Y_CRYwBr-gFArOirF1b76KyI5N8eVpsLeDiIsKtYkpU,1641
@@ -70,7 +70,7 @@ bisheng_langchain/embeddings/interface/__init__.py,sha256=GNY3tibpRxpAdAfSvQmXBK
 bisheng_langchain/embeddings/interface/types.py,sha256=VdurbtsnjCPdlOjPFcK2Mg6r9bJYYHb3tepvkk-y3nM,461
 bisheng_langchain/embeddings/interface/wenxin.py,sha256=5d9gI4enmfkD80s0FHKiDt33O0mwM8Xc5WTubnMUy8c,3104
 bisheng_langchain/gpts/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-bisheng_langchain/gpts/assistant.py,sha256=lObPDRSSzRcz6jijj76zr9dtfWOIztCaII429BPTHsU,5444
+bisheng_langchain/gpts/assistant.py,sha256=jPGVjPhqx-z4nvEHVsprWLqAVbO99Uus_GADAPASXQE,5438
 bisheng_langchain/gpts/auto_optimization.py,sha256=WNsC19rgvuDYQlSIaYThq5RqCbuobDbzCwAJW4Ksw0c,3626
 bisheng_langchain/gpts/auto_tool_selected.py,sha256=21WETf9o0YS-QEBwv3mmZRObKWszefQkXEqAA6KzoaM,1582
 bisheng_langchain/gpts/load_tools.py,sha256=LiiK1OqFu7Ki-F_Rhfi1rgp0wBQCSrTDdqsgwciTOIU,8099
@@ -80,12 +80,12 @@ bisheng_langchain/gpts/agent_types/__init__.py,sha256=88tFt1GfrfIqa4hCg0cMJk7rTe
 bisheng_langchain/gpts/agent_types/llm_functions_agent.py,sha256=IXg5u8dSk-FcLvjrvvLcN5revGccXylXkD73ZWhaDWs,8715
 bisheng_langchain/gpts/agent_types/llm_react_agent.py,sha256=W6IJMwoFUj_a2cXZ_nXOpzaHBjP5IBrFxRTSejNVi9A,6678
 bisheng_langchain/gpts/prompts/__init__.py,sha256=pOnXvk6_PjqAoLrh68sI9o3o6znKGxoLMVFP-0XTCJo,704
-bisheng_langchain/gpts/prompts/assistant_prompt_base.py,sha256=Q69qWxtwaeGPPwMWXQ44OsQPg7wHB7nYT17wbqQe3kM,57
+bisheng_langchain/gpts/prompts/assistant_prompt_base.py,sha256=Yp9M1XbZb5jHeBG_txcwWA84Euvl89t0g-GbJMa5Ur0,1133
 bisheng_langchain/gpts/prompts/assistant_prompt_cohere.py,sha256=GLQ77oXqSlE7Xes2ObsFsNon5nOJOCRhQOKE5bUpgaI,2421
 bisheng_langchain/gpts/prompts/assistant_prompt_opt.py,sha256=TZsRK4XPMrUhGg0PoMyiE3wE-aG34UmlVflkCl_c0QI,4151
 bisheng_langchain/gpts/prompts/breif_description_prompt.py,sha256=w4A5et0jB-GkxEMQBp4i6GKX3RkVeu7NzWEjOZZAicM,5336
 bisheng_langchain/gpts/prompts/opening_dialog_prompt.py,sha256=VVF0JLHtetupVB0kabiFHWDHlQaa4nFLcbYXgIBA3nw,5965
-bisheng_langchain/gpts/prompts/react_agent_prompt.py,sha256=YGaO4S0r9AZhzt9Idyiw2mcd5QH1GLIIGc2oAN2YKdE,1976
+bisheng_langchain/gpts/prompts/react_agent_prompt.py,sha256=MA5FReipAYfe6ypOvg_SJSlHxsjNdCh1HFG22axjFLo,1968
 bisheng_langchain/gpts/prompts/select_tools_prompt.py,sha256=AyvVnrLEsQy7RHuGTPkcrMUxgA98Q0TzF-xweoc7GyY,1400
 bisheng_langchain/gpts/tools/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 bisheng_langchain/gpts/tools/api_tools/__init__.py,sha256=CkEjgIFM4GIv86V1B7SsFLaB6M86c54QuO8wIRizUZ8,1608
@@ -126,7 +126,7 @@ bisheng_langchain/rag/config/baseline_s2b_mix.yaml,sha256=rkPfzU2-mvjRrZ0zMHaQsn
 bisheng_langchain/rag/config/baseline_v2.yaml,sha256=RP-DwIRIS_ZK8ixbXi2Z28rKqHD56pWmr2o2WWIwq3Y,2382
 bisheng_langchain/rag/init_retrievers/__init__.py,sha256=qpLLAuqZPtumTlJj17Ie5AbDDmiUiDxYefg_pumqu-c,218
 bisheng_langchain/rag/init_retrievers/baseline_vector_retriever.py,sha256=oRKZZpxlLQAtsubIcAXeXpf1a9h6Pt6uOtNTLeD2jps,2362
-bisheng_langchain/rag/init_retrievers/keyword_retriever.py,sha256=NRT0fBx6HFR7j9IbRl_NBuqF7hnL-9v5GCqHpgnrfPQ,2523
+bisheng_langchain/rag/init_retrievers/keyword_retriever.py,sha256=Da4Q5BrfN0GckJaeAgPYMlzQAp9ll7ZGGyvs7OdCQ5c,2513
 bisheng_langchain/rag/init_retrievers/mix_retriever.py,sha256=Whxq4kjNPLsxnHcVo60usdFFwLTCD-1jO38q08LXkVQ,4653
 bisheng_langchain/rag/init_retrievers/smaller_chunks_retriever.py,sha256=RQ7QLEOOhBrkw-EimXVJqIGa96D-KkNDik2h9hzg9fU,3805
 bisheng_langchain/rag/prompts/__init__.py,sha256=IUCq9gzqGQN_6IDk0D_F5t3mOUI_KbmSzYnnXoX4VKE,223
@@ -147,12 +147,13 @@ bisheng_langchain/retrievers/mix_es_vector.py,sha256=dSrrsuMPSgGiu181EOzACyIKiDX
 bisheng_langchain/sql/__init__.py,sha256=2arRtNQ-kUvIsy_8v_PrLxf5r9W-S7mbqptG_l4_1RE,88
 bisheng_langchain/sql/base.py,sha256=WNHCy16UoxvDbroHnJq8CsZ9ot4NGflCm8Bgiv45kks,6152
 bisheng_langchain/utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+bisheng_langchain/utils/azure_dalle_image_generator.py,sha256=wbOIrVX8lN8zZni0GjUpwKBESy0TmlqrGLZCY5PbUsM,6819
 bisheng_langchain/utils/requests.py,sha256=vWGKyNTxApVeaVdKxqACfIT1Q8wMy-jC3kUv2Ce9Mzc,8688
 bisheng_langchain/vectorstores/__init__.py,sha256=zCZgDe7LyQ0iDkfcm5UJ5NxwKQSRHnqrsjx700Fy11M,213
 bisheng_langchain/vectorstores/elastic_keywords_search.py,sha256=Pm1rS50GJ0HWbjBsFDgs28SVuVbjGSRPOor6yJlnE7w,13347
-bisheng_langchain/vectorstores/milvus.py,sha256=lrnezKnYXhyH5M1g3a-Mcwpj9mwzAj44TKmzyUXlQYY,36297
+bisheng_langchain/vectorstores/milvus.py,sha256=8HHbIxoSbLYDFlFJSfmjLOfqGpOSZd24iVYWSYz3TX0,36637
 bisheng_langchain/vectorstores/retriever.py,sha256=hj4nAAl352EV_ANnU2OHJn7omCH3nBK82ydo14KqMH4,4353
-bisheng_langchain-0.3.3b0.dist-info/METADATA,sha256=3ZhnM1PA4jWzmVmse0P8h2HUeremsdpWHom39DphHqA,2486
-bisheng_langchain-0.3.3b0.dist-info/WHEEL,sha256=GJ7t_kWBFywbagK5eo9IoUwLW6oyOeTKmQ-9iHFVNxQ,92
-bisheng_langchain-0.3.3b0.dist-info/top_level.txt,sha256=Z6pPNyCo4ihyr9iqGQbH8sJiC4dAUwA_mAyGRQB5_Fs,18
-bisheng_langchain-0.3.3b0.dist-info/RECORD,,
+bisheng_langchain-0.3.3.dev2.dist-info/METADATA,sha256=9JMcguR6d_cAwkNX3ynYJi55rnk6PjZSqfUIXH9gnG4,2489
+bisheng_langchain-0.3.3.dev2.dist-info/WHEEL,sha256=GJ7t_kWBFywbagK5eo9IoUwLW6oyOeTKmQ-9iHFVNxQ,92
+bisheng_langchain-0.3.3.dev2.dist-info/top_level.txt,sha256=Z6pPNyCo4ihyr9iqGQbH8sJiC4dAUwA_mAyGRQB5_Fs,18
+bisheng_langchain-0.3.3.dev2.dist-info/RECORD,,