merge from master

README.md

@@ -59,7 +59,7 @@ docker run -d --gpus all -p 80:8501 registry.cn-beijing.aliyuncs.com/chatchat/ch
 该项目是一个可以实现 __完全本地化__推理的知识库增强方案, 重点解决数据安全保护，私域化部署的企业痛点。
 本开源方案采用```Apache License```，可以免费商用，无需付费。
 
-我们支持市面上主流的本地大预言模型和Embedding模型，支持开源的本地向量数据库。
+我们支持市面上主流的本地大语言模型和Embedding模型，支持开源的本地向量数据库。
 支持列表详见[Wiki](https://github.com/chatchat-space/Langchain-Chatchat/wiki/)
 
 
@@ -92,14 +92,14 @@ $ pip install -r requirements_webui.txt
 
 如需在本地或离线环境下运行本项目，需要首先将项目所需的模型下载至本地，通常开源 LLM 与 Embedding 模型可以从 [HuggingFace](https://huggingface.co/models) 下载。
 
-以本项目中默认使用的 LLM 模型 [THUDM/ChatGLM2-6B](https://huggingface.co/THUDM/chatglm2-6b) 与 Embedding 模型 [moka-ai/m3e-base](https://huggingface.co/moka-ai/m3e-base) 为例：
+以本项目中默认使用的 LLM 模型 [THUDM/ChatGLM3-6B](https://huggingface.co/THUDM/chatglm3-6b) 与 Embedding 模型 [BAAI/bge-large-zh](https://huggingface.co/BAAI/bge-large-zh) 为例：
 
 下载模型需要先[安装 Git LFS](https://docs.github.com/zh/repositories/working-with-files/managing-large-files/installing-git-large-file-storage)，然后运行
 
 ```Shell
 $ git lfs install
-$ git clone https://huggingface.co/THUDM/chatglm2-6b
+$ git clone https://huggingface.co/THUDM/chatglm3-6b
-$ git clone https://huggingface.co/moka-ai/m3e-base
+$ git clone https://huggingface.co/BAAI/bge-large-zh
 ```
 ### 3. 初始化知识库和配置文件
 
@@ -148,7 +148,7 @@ $ python startup.py -a
 [![Telegram](https://img.shields.io/badge/Telegram-2CA5E0?style=for-the-badge&logo=telegram&logoColor=white "langchain-chatglm")](https://t.me/+RjliQ3jnJ1YyN2E9)
 
 ### 项目交流群
-<img src="img/qr_code_76.jpg" alt="二维码" width="300" />
+<img src="img/qr_code_82.jpg" alt="二维码" width="300" />
 
 🎉 Langchain-Chatchat 项目微信交流群，如果你也对本项目感兴趣，欢迎加入群聊参与讨论交流。
 

configs/model_config.py.example

@@ -105,6 +105,7 @@ ONLINE_LLM_MODEL = {
         "version": "qwen-turbo",  # 可选包括 "qwen-turbo", "qwen-plus"
         "api_key": "",  # 请在阿里云控制台模型服务灵积API-KEY管理页面创建
         "provider": "QwenWorker",
+        "embed_model": "text-embedding-v1" # embedding 模型名称
     },
 
     # 百川 API，申请方式请参考 https://www.baichuan-ai.com/home#api-enter
@@ -241,7 +242,7 @@ MODEL_PATH = {
         "agentlm-13b": "THUDM/agentlm-13b",
         "agentlm-70b": "THUDM/agentlm-70b",
 
-        "Yi-34B-Chat": "https://huggingface.co/01-ai/Yi-34B-Chat",
+        "Yi-34B-Chat": "01-ai/Yi-34B-Chat",
     },
     "reranker":{
         "bge-reranker-large":"BAAI/bge-reranker-large",

knowledge_base/samples/content/wiki

@@ -1 +1 @@
-Subproject commit 4c6f1a9748563b5d3ae81a628ff8a71f3e5686d1
+Subproject commit 2f24adb218f23eab00d7fcd7ccf5072f2f35cb3c

requirements.txt

@@ -1,18 +1,18 @@
 # API requirements
 
 # On Windows system, install the cuda version manually from https://pytorch.org/
-torch>=2.1.2
+torch~=2.1.2
-torchvision>=0.16.2
+torchvision~=0.16.2
-torchaudio>=2.1.2
+torchaudio~=2.1.2
 
-langchain>=0.0.352
+langchain==0.0.352
-langchain-experimental>=0.0.47
+langchain-experimental==0.0.47
 pydantic==1.10.13
-fschat>=0.2.33
+fschat==0.2.34
-xformers>=0.0.23.post1
+xformers==0.0.23.post1
-openai>=1.6.0
+openai~=1.6.0
-sentence_transformers>=2.2.2
+sentence_transformers==2.2.2
-transformers>=4.36.2
+transformers==4.36.2
 fastapi>=0.105
 nltk>=3.8.1
 uvicorn>=0.24.0.post1
@@ -20,11 +20,11 @@ starlette~=0.27.0
 unstructured[all-docs]==0.11.0
 python-magic-bin; sys_platform == 'win32'
 SQLAlchemy==2.0.19
-faiss-cpu>=1.7.4
+faiss-cpu~=1.7.4
-accelerate>=0.24.1
+accelerate==0.24.1
-spacy>=3.7.2
+spacy~=3.7.2
-PyMuPDF>=1.23.8
+PyMuPDF~=1.23.8
-rapidocr_onnxruntime>=1.3.8
+rapidocr_onnxruntime~=1.3.8
 requests>=2.31.0
 pathlib>=1.0.1
 pytest>=7.4.3
@@ -37,20 +37,20 @@ websockets>=12.0
 numpy~=1.26.2
 pandas~=2.1.4
 einops>=0.7.0
-transformers_stream_generator>=0.0.4
+transformers_stream_generator==0.0.4
-vllm>=0.2.6; sys_platform == "linux"
+vllm==0.2.6; sys_platform == "linux"
 
 # optional document loaders
 
 rapidocr_paddle[gpu]>=1.3.0.post5
 jq>=1.6.0
 html2text
-beautifulsoup4>=4.12.2
+beautifulsoup4~=4.12.2
-pysrt>=1.1.2
+pysrt~=1.1.2
 
 # Online api libs dependencies
 
-zhipuai>=1.0.7
+zhipuai>=1.0.7，<=2.0.0
 dashscope>=1.13.6
 volcengine>=1.0.119
 pymilvus>=2.3.4
@@ -61,17 +61,15 @@ pymilvus>=2.3.4
 
 arxiv>=2.0.0
 youtube-search>=2.1.2
-duckduckgo-search>=4.1.0
+duckduckgo-search>=3.9.9
 metaphor-python>=0.1.23
 
 # WebUI requirements
 
-streamlit>=1.29.0
+streamlit~=1.29.0
 streamlit-option-menu>=0.3.6
-streamlit-antd-components>=0.2.4
+streamlit-chatbox==1.1.11
-streamlit-chatbox>=1.1.11
 streamlit-modal>=0.1.0
 streamlit-aggrid>=0.3.4.post3
 httpx[brotli,http2,socks]>=0.25.2
 watchdog>=3.0.0
-

requirements_api.txt

@@ -1,17 +1,18 @@
 # API requirements
 
-torch>=2.1.2
+# On Windows system, install the cuda version manually from https://pytorch.org/
-torchvision>=0.16.2
+torch~=2.1.2
-torchaudio>=2.1.2
+torchvision~=0.16.2
+torchaudio~=2.1.2
 
-langchain>=0.0.352
+langchain==0.0.352
-langchain-experimental>=0.0.47
+langchain-experimental==0.0.47
 pydantic==1.10.13
-fschat>=0.2.34
+fschat==0.2.34
-xformers>=0.0.23.post1
+xformers==0.0.23.post1
-openai>=1.6.0
+openai~=1.6.0
-sentence_transformers>=2.2.2
+sentence_transformers==2.2.2
-transformers>=4.36.2
+transformers==4.36.2
 fastapi>=0.105
 nltk>=3.8.1
 uvicorn>=0.24.0.post1
@@ -19,11 +20,11 @@ starlette~=0.27.0
 unstructured[all-docs]==0.11.0
 python-magic-bin; sys_platform == 'win32'
 SQLAlchemy==2.0.19
-faiss-cpu>=1.7.4
+faiss-cpu~=1.7.4
-accelerate>=0.24.1
+accelerate==0.24.1
-spacy>=3.7.2
+spacy~=3.7.2
-PyMuPDF>=1.23.8
+PyMuPDF~=1.23.8
-rapidocr_onnxruntime>=1.3.8
+rapidocr_onnxruntime~=1.3.8
 requests>=2.31.0
 pathlib>=1.0.1
 pytest>=7.4.3
@@ -36,16 +37,16 @@ websockets>=12.0
 numpy~=1.26.2
 pandas~=2.1.4
 einops>=0.7.0
-transformers_stream_generator>=0.0.4
+transformers_stream_generator==0.0.4
-vllm>=0.2.6; sys_platform == "linux"
+vllm==0.2.6; sys_platform == "linux"
 
 # optional document loaders
 
 rapidocr_paddle[gpu]>=1.3.0.post5
 jq>=1.6.0
 html2text
-beautifulsoup4>=4.12.2
+beautifulsoup4~=4.12.2
-pysrt>=1.1.2
+pysrt~=1.1.2
 
 # Online api libs dependencies
 
@@ -60,5 +61,5 @@ pymilvus>=2.3.4
 
 arxiv>=2.0.0
 youtube-search>=2.1.2
-duckduckgo-search>=4.1.0
+duckduckgo-search>=3.9.9
 metaphor-python>=0.1.23

requirements_webui.txt

@@ -1,9 +1,9 @@
 # WebUI requirements
 
-streamlit>=1.29.0 # do remember to add streamlit to environment variables if you use windows
+streamlit~=1.29.0
 streamlit-option-menu>=0.3.6
-streamlit-chatbox>=1.1.11
+streamlit-chatbox==1.1.11
 streamlit-modal>=0.1.0
 streamlit-aggrid>=0.3.4.post3
+httpx[brotli,http2,socks]>=0.25.2
 watchdog>=3.0.0
-

server/chat/file_chat.py

@@ -132,7 +132,7 @@ async def file_chat(query: str = Body(..., description="用户输入", examples=
 
         context = "\n".join([doc.page_content for doc in docs])
         if len(docs) == 0: ## 如果没有找到相关文档，使用Empty模板
-            prompt_template = get_prompt_template("knowledge_base_chat", "Empty")
+            prompt_template = get_prompt_template("knowledge_base_chat", "empty")
         else:
             prompt_template = get_prompt_template("knowledge_base_chat", prompt_name)
         input_msg = History(role="user", content=prompt_template).to_msg_template(False)

server/embeddings_api.py

@@ -6,32 +6,32 @@ from fastapi import Body
 from fastapi.concurrency import run_in_threadpool
 from typing import Dict, List
 
-
 online_embed_models = list_online_embed_models()
 
 
 def embed_texts(
-    texts: List[str],
+        texts: List[str],
-    embed_model: str = EMBEDDING_MODEL,
+        embed_model: str = EMBEDDING_MODEL,
-    to_query: bool = False,
+        to_query: bool = False,
 ) -> BaseResponse:
     '''
     对文本进行向量化。返回数据格式：BaseResponse(data=List[List[float]])
     TODO: 也许需要加入缓存机制，减少 token 消耗
     '''
     try:
-        if embed_model in list_embed_models(): # 使用本地Embeddings模型
+        if embed_model in list_embed_models():  # 使用本地Embeddings模型
             from server.utils import load_local_embeddings
 
             embeddings = load_local_embeddings(model=embed_model)
             return BaseResponse(data=embeddings.embed_documents(texts))
 
-        if embed_model in list_online_embed_models(): # 使用在线API
+        if embed_model in list_online_embed_models():  # 使用在线API
             config = get_model_worker_config(embed_model)
             worker_class = config.get("worker_class")
+            embed_model = config.get("embed_model")
             worker = worker_class()
             if worker_class.can_embedding():
-                params = ApiEmbeddingsParams(texts=texts, to_query=to_query)
+                params = ApiEmbeddingsParams(texts=texts, to_query=to_query, embed_model=embed_model)
                 resp = worker.do_embeddings(params)
                 return BaseResponse(**resp)
 
@@ -67,9 +67,10 @@ async def aembed_texts(
 
 
 def embed_texts_endpoint(
-    texts: List[str] = Body(..., description="要嵌入的文本列表", examples=[["hello", "world"]]),
+        texts: List[str] = Body(..., description="要嵌入的文本列表", examples=[["hello", "world"]]),
-    embed_model: str = Body(EMBEDDING_MODEL, description=f"使用的嵌入模型，除了本地部署的Embedding模型，也支持在线API({online_embed_models})提供的嵌入服务。"),
+        embed_model: str = Body(EMBEDDING_MODEL,
-    to_query: bool = Body(False, description="向量是否用于查询。有些模型如Minimax对存储/查询的向量进行了区分优化。"),
+                                description=f"使用的嵌入模型，除了本地部署的Embedding模型，也支持在线API({online_embed_models})提供的嵌入服务。"),
+        to_query: bool = Body(False, description="向量是否用于查询。有些模型如Minimax对存储/查询的向量进行了区分优化。"),
 ) -> BaseResponse:
     '''
     对文本进行向量化，返回 BaseResponse(data=List[List[float]])
@@ -78,9 +79,9 @@ def embed_texts_endpoint(
 
 
 def embed_documents(
-    docs: List[Document],
+        docs: List[Document],
-    embed_model: str = EMBEDDING_MODEL,
+        embed_model: str = EMBEDDING_MODEL,
-    to_query: bool = False,
+        to_query: bool = False,
 ) -> Dict:
     """
     将 List[Document] 向量化，转化为 VectorStore.add_embeddings 可以接受的参数

server/model_workers/base.py

@@ -113,6 +113,9 @@ class ApiModelWorker(BaseModelWorker):
         sys.stdout = sys.__stdout__
         sys.stderr = sys.__stderr__
 
+        new_loop = asyncio.new_event_loop()
+        asyncio.set_event_loop(new_loop)
+
         self.context_len = context_len
         self.semaphore = asyncio.Semaphore(self.limit_worker_concurrency)
         self.version = None