From bf6e4e3b42c39616e746024744bb90373260cc0b Mon Sep 17 00:00:00 2001
From: tanfeng <823018000@qq.com>
Date: Fri, 22 Nov 2024 16:57:08 +0800
Subject: [PATCH 1/6] =?UTF-8?q?=E9=80=82=E9=85=8D=E6=9C=80=E6=96=B0POC?=
 =?UTF-8?q?=E7=89=88=E6=9C=AC=E5=B1=95=E7=A4=BA=E9=A1=B5=E9=9D=A2=EF=BC=8C?=
 =?UTF-8?q?=E5=8A=9F=E8=83=BD=E5=A2=9E=E5=8A=A0?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 mxRAG/PocValidation/chat_with_ascend/app.py | 558 ++++++++++++++++----
 1 file changed, 441 insertions(+), 117 deletions(-)

diff --git a/mxRAG/PocValidation/chat_with_ascend/app.py b/mxRAG/PocValidation/chat_with_ascend/app.py
index 83d1f33d2..d5805bfb8 100644
--- a/mxRAG/PocValidation/chat_with_ascend/app.py
+++ b/mxRAG/PocValidation/chat_with_ascend/app.py
@@ -9,6 +9,9 @@ from loguru import logger
 from paddle.base import libpaddle
 from mx_rag.chain import SingleText2TextChain
 from mx_rag.embedding.local import TextEmbedding
+from mx_rag.embedding.service import TEIEmbedding
+from mx_rag.reranker.local import LocalReranker
+from mx_rag.reranker.service import TEIReranker
 from mx_rag.llm import Text2TextLLM, LLMParameterConfig
 from mx_rag.retrievers import Retriever
 from mx_rag.storage.document_store import SQLiteDocstore
@@ -17,39 +20,225 @@ from mx_rag.storage.vectorstore.vectorstore import SimilarityStrategy
 from mx_rag.utils import ClientParam
 from langchain.text_splitter import RecursiveCharacterTextSplitter
 from langchain_community.document_loaders import TextLoader
-from mx_rag.document.loader import DocxLoader, PdfLoader
-from mx_rag.knowledge import KnowledgeDB
-from mx_rag.knowledge.knowledge import KnowledgeStore
+from mx_rag.document.loader import DocxLoader, PdfLoader, ExcelLoader
+from mx_rag.knowledge import KnowledgeStore, KnowledgeDB, KnowledgeMgrStore, KnowledgeMgr
 from mx_rag.knowledge.handler import upload_files, LoaderMng
 
-global text2text_chain
-global text_knowledge_db
-global loader_mng
-global text_emb
 
+class CustomFormatter(argparse.ArgumentDefaultsHelpFormatter):
+    def _get_default_metavar_for_optional(self, action):
+        return action.type.__name__
+
+    def _get_default_metavar_for_positional(self, action):
+        return action.type.__name__
+
+
+parse = argparse.ArgumentParser(formatter_class=CustomFormatter)
+parse.add_argument("--embedding_path", type=str, default="/home/data/acge_text_embedding", help="embedding模型本地路径")
+parse.add_argument("--tei_emb", type=bool, default=False, help="是否使用TEI服务化的embedding模型")
+parse.add_argument("--embedding_url", type=str, default="http://127.0.0.1:8080/embed",help="使用TEI服务化的embedding模型url地址")
+parse.add_argument("--embedding_dim", type=int, default=1024, help="embedding模型向量维度")
+parse.add_argument("--llm_url", type=str, default="http://127.0.0.1:1025/v1/chat/completions", help="大模型url地址")
+parse.add_argument("--model_name", type=str, default="Llama3-8B-Chinese-Chat", help="大模型名称")
+parse.add_argument("--tei_reranker", type=bool, default=False, help="是否使用TEI服务化的reranker模型")
+parse.add_argument("--reranker_path", type=str, default=None, help="reranker模型本地路径")
+parse.add_argument("--reranker_url", type=str, default=None, help="使用TEI服务化的embedding模型url地址")
+parse.add_argument("--dev", type=int, default=0, help="使用的npu卡，可通过执行npu-smi info获取")
+parse.add_argument("--port", type=int, default=8080, help="web后端端口")
+
+args = parse.parse_args().__dict__
+tei_emb: bool = args.pop('tei_emb')
+embedding_path: str = args.pop('embedding_path')
+embedding_url: str = args.pop('embedding_url')
+embedding_dim: int = args.pop('embedding_dim')
+llm_url: str = args.pop('llm_url')
+model_name: str = args.pop('model_name')
+tei_reranker: bool = args.pop('tei_reranker')
+reranker_path: str = args.pop('reranker_path')
+reranker_url: str = args.pop('reranker_url')
+dev: int = args.pop('dev')
+port: int = args.pop('port')
 
+# 初始化向量数据库
+vector_store = MindFAISS(x_dim=embedding_dim,
+                         similarity_strategy=SimilarityStrategy.FLAT_L2,
+                         devs=[dev],
+                         load_local_index="./test_faiss.index",
+                         auto_save=True
+                         )
+# 初始化文档chunk关系数据库
+chunk_store = SQLiteDocstore(db_path="./test_sql.db")
+# 初始化知识管理关系数据库
+knowledge_store = KnowledgeStore(db_path="./test_sql.db")
+# 初始化知识库管理
+text_knowledge_db = KnowledgeDB(knowledge_store=knowledge_store,
+                                chunk_store=chunk_store,
+                                vector_store=vector_store,
+                                knowledge_name="test",
+                                white_paths=["/tmp"]
+                                )
+knowledge_mgr = KnowledgeMgr(KnowledgeMgrStore("./test_sql.db"))
+
+# 配置text生成text大模型chain，具体ip端口请根据实际情况适配修改
+llm = Text2TextLLM(base_url=llm_url, model_name=model_name, client_param=ClientParam(use_http=True))
+
+# 配置embedding模型，请根据模型具体路径适配
+if tei_emb:
+    text_emb = TEIEmbedding(url=embedding_url, client_param=ClientParam(use_http=True))
+else:
+    text_emb = TextEmbedding(model_path=embedding_path, dev_id=dev)
+
+# 配置rerank,请根据模型具体路径适配
+if tei_reranker:
+    reranker = TEIReranker(url=reranker_url, client_param=ClientParam(use_http=True))
+elif reranker_path is not None:
+    reranker = LocalReranker(model_path=reranker_path, dev_id=dev)
+else:
+    reranker = None
+
+
+# 历史问题改写
+def generate_question(history, llm, history_n: int = 5):
+    prompt = """现在你有一个上下文依赖问题补全任务，任务要求:请根据对话历史和用户当前的问句，重写问句。\n
+    历史问题依次是:\n
+    {}\n
+    用户当前的问句:\n
+    {}\n
+    注意如果当前问题不依赖历史问题直接返回none即可\n
+    请根据上述对话历史重写用户当前的问句,仅输出重写后的问句,不需要附加任何分析。\n
+    重写问句: \n
+    """
+    if len(history) <= 2:
+        return history
+    cur_query = history[-1][0]
+    history_qa = history[0:-1]
+    history_list = [f"第{idx + 1}轮：{q_a[0]}" for idx, q_a in enumerate(history_qa) if q_a[0] is not None]
+    history_list = history_list[:history_n]
+    history_str = "\n\n".join(history_list)
+    re_query = prompt.format(history_str, cur_query)
+    new_query = llm.chat(query=re_query, llm_config=LLMParameterConfig(max_tokens=512,
+                                                                       temperature=0.5,
+                                                                       top_p=0.95))
+    if new_query != "none":
+        history[-1][0] = "原始问题: " + cur_query
+        history += [[new_query, None]]
+    return history
+
+
+# 聊天对话框
 def bot_response(history,
+                 history_r,
                  max_tokens: int = 512,
                  temperature: float = 0.5,
-                 top_p: float = 0.95
+                 top_p: float = 0.95,
+                 history_n: int = 5,
+                 score_threshold: float = 0.5,
+                 top_k: int = 1,
+                 chat_type: str = "RAG检索增强对话",
+                 show_type: str = "不显示",
+                 is_rewrite:str = "否",
+                 knowledge_name: str = 'test',
                  ):
-    # 将最新的问题传给RAG
+    chat_type_mapping = {"RAG检索增强对话":1,
+                         "仅大模型对话":0}
+    show_type_mapping = {"对话结束后显示":1,
+                         "检索框单独显示":2,
+                         "不显示":0}
+    is_rewrite_mapping = {"是":1,
+                          "否":0}
+    # 初始化检索器
+    retrieve_cls = creat_retriever(knowledge_name, top_k, score_threshold)
+    # 初始化chain
+    text2text_chain = SingleText2TextChain(llm=llm, retriever=retrieve_cls, reranker=reranker)
+    # 历史问题改写
+    if is_rewrite_mapping.get(is_rewrite) == 1:
+        history = generate_question(history, llm, history_n)
+    history[-1][1] = '推理错误'
     try:
-        response = text2text_chain.query(history[-1][0],
-                                        LLMParameterConfig(max_tokens=max_tokens, temperature=temperature, top_p=top_p,
-                                                           stream=True))
-        # 返回迭代器
-        history[-1][1] = '推理错误'
-        for res in response:
-            history[-1][1] = '推理错误' if res['result'] is None else res['result']
-            yield history
-        yield history
+        # 仅使用大模型
+        if chat_type_mapping.get(chat_type) == 0:
+            response = llm.chat_streamly(query=history[-1][0],
+                                         llm_config=LLMParameterConfig(max_tokens=max_tokens,
+                                                                       temperature=temperature,
+                                                                       top_p=top_p))
+            # 返回迭代器
+            for res in response:
+                history[-1][1] = res
+                yield history, history_r
+        # 使用RAG增强回答
+        elif chat_type_mapping.get(chat_type) == 1:
+            response = text2text_chain.query(history[-1][0],
+                                             LLMParameterConfig(max_tokens=max_tokens,
+                                                                temperature=temperature,
+                                                                top_p=top_p,
+                                                                stream=True))
+            # 不展示检索内容
+            if show_type_mapping.get(show_type) == 0:
+                for res in response:
+                    history[-1][1] = '推理错误' if res['result'] is None else res['result']
+                    yield history, history_r
+            # 问答结尾展示
+            elif show_type_mapping.get(show_type) == 1:
+                for res in response:
+                    history[-1][1] = '推理错误' if res['result'] is None else res['result']
+                    q_docs = res['source_documents']
+                    yield history, history_r
+                # 有检索到信息
+                if len(q_docs) > 0:
+                    history_last = ''
+                    for i ,source in enumerate(q_docs):
+                        sources = "\n====检索信息来源" + str(i + 1) + ":" + source['metadata']['source'] + "====" + "\n" + \
+                                  "--->参考内容：" +source['page_content'] + "\n"
+                        history_last += sources
+                    history += [[None, history_last]]
+                yield history, history_r
+            # 检索窗口展示
+            else:
+                for res in response:
+                    history[-1][1] = '推理错误' if res['result'] is None else res['result']
+                    q_docs = res['source_documents']
+                    yield history, history_r
+                # 有检索到信息
+                if len(q_docs) > 0:
+                    history_r_last = ''
+                    for i ,source in enumerate(q_docs):
+                        sources = "\n====检索信息来源" + str(i + 1) + ":" + source['metadata']['source'] + "====" + "\n" + \
+                                  "--->参考内容：" +source['page_content'] + "\n"
+                        history_r_last += sources
+                    history_r += [[history[-1][0], history_r_last]]
+                yield history, history_r
     except Exception as err:
         logger.info(f"query failed, find exception: {err}")
-        history[-1][1] = "推理错误"
-        yield history
+        yield history, history_r
+
+
+# 检索对话框
+def re_response(history_r,
+                score_threshold: float = 0.5,
+                top_k: int = 1,
+                knowledge_name: str = 'test'
+                ):
+    # 初始化检索器
+    retrieve_cls = creat_retriever(knowledge_name, top_k, score_threshold)
+    q_docs = retrieve_cls.invoke(history_r[-1][0])
+    if len(q_docs) > 0:
+        history_r_last = ''
+        for i, source in enumerate(q_docs):
+            sources = "\n====检索信息来源" + str(i + 1) + ":" + source.metadata['source'] + "====" + "\n" + \
+                      "--->参考内容：" + source.page_content + "\n"
+            history_r_last += sources
+        history_r[-1][1] = history_r_last
+    else:
+        history_r[-1][1] = "未检索到相关信息"
+    return history_r
 
 
+# 检索信息
+def user_retriever(user_message, history_r):
+    return "",history_r + [[user_message, None]]
+
+
+# 聊天信息
 def user_query(user_message, history):
     return "", history + [[user_message, None]]
 
@@ -58,23 +247,137 @@ def clear_history(history):
     return []
 
 
-SAVE_FILE_PATH = "/tmp/document_files"
+# 创建新的知识库
+def creat_knowledge_db(knowledge_name: str, is_regist: bool=True):
+    knowledge_name = get_knowledge_rename(knowledge_name)
+    index_name, db_name = get_db_names(knowledge_name)
+    vector_store_cls = MindFAISS(x_dim=embedding_dim,
+                                 similarity_strategy=SimilarityStrategy.FLAT_L2,
+                                 devs=[dev],
+                                 load_local_index=index_name,
+                                 auto_save=True)
+    # 初始化文档chunk关系数据库
+    chunk_store_cls = SQLiteDocstore(db_path=db_name)
+    # 初始化知识管理关系数据库
+    knowledge_store_cls = KnowledgeStore(db_path=db_name)
+    # 初始化知识库管理
+    knowledge_db_cls = KnowledgeDB(knowledge_store=knowledge_store_cls,
+                                   chunk_store=chunk_store_cls,
+                                   vector_store=vector_store_cls,
+                                   knowledge_name=knowledge_name,
+                                   white_paths=["/tmp"])
+    if is_regist:
+        knowledge_mgr.register(knowledge_db_cls)
+    return knowledge_db_cls
+
+
+def creat_retriever(knowledge_name: str, top_k, score_threshold):
+    knowledge_name = get_knowledge_rename(knowledge_name)
+    index_name, db_name = get_db_names(knowledge_name)
+    vector_store_ret = MindFAISS(x_dim=embedding_dim,
+                                 similarity_strategy=SimilarityStrategy.FLAT_L2,
+                                 devs=[dev],
+                                 load_local_index=index_name,
+                                 auto_save=True)
+    # 初始化文档chunk关系数据库
+    chunk_store_ret = SQLiteDocstore(db_path=db_name)
+    retriever_cls = Retriever(vector_store=vector_store_ret,
+                              document_store=chunk_store_ret,
+                              embed_func=text_emb.embed_documents,
+                              k=top_k,
+                              score_threshold=score_threshold)
+    return retriever_cls
+
+
+# 删除知识库
+def delete_knowledge_db(knowledge_name: str):
+    knowledge_name = get_knowledge_rename(knowledge_name)
+    knowledge_db_lst = knowledge_mgr.get_all()
+    if knowledge_name in knowledge_db_lst:
+        knowledge_db_cls = creat_knowledge_db(knowledge_name, is_regist=False)
+        clear_document(knowledge_name)
+        knowledge_mgr.delete(knowledge_db_cls)
+    return get_knowledge_db()
+
+
+# 获取知识库文档列表
+def get_document(knowledge_name: str):
+    knowledge_name = get_knowledge_rename(knowledge_name)
+    knowledge_db_cls = creat_knowledge_db(knowledge_name, is_regist=False)
+    doc_names = knowledge_db_cls.get_all_documents()
+    return knowledge_name, doc_names, len(doc_names)
+
+
+# 清空知识库文档列表
+def clear_document(knowledge_name: str):
+    knowledge_name, doc_names, doc_cnt = get_document(knowledge_name)
+    if doc_cnt > 0:
+        knowledge_db_cls = creat_knowledge_db(knowledge_name, is_regist=False)
+        for doc_name in doc_names:
+            knowledge_db_cls.delete_file(doc_name)
+        return get_document(knowledge_name)
+    else:
+        return knowledge_name, doc_names, doc_cnt
+
+
+# 获取知识库列表
+def get_knowledge_db():
+    knowledge_db_lst = knowledge_mgr.get_all()
+    return knowledge_db_lst, len(knowledge_db_lst)
+
+
+def get_db_names(knowledge_name: str):
+    index_name = "./" + knowledge_name + "_faiss.index"
+    dn_name = "./" + knowledge_name + "_sql.db"
+    return index_name, dn_name
 
 
-def file_upload(files):
+def get_knowledge_rename(knowledge_name: str):
+    if knowledge_name is None or len(knowledge_name) == 0 or ' ' in knowledge_name:
+        return 'test'
+    else:
+        return knowledge_name
+
+
+# 上传知识库
+def file_upload(files,
+                knowledge_db_name: str = 'test',
+                chunk_size: int = 750,
+                chunk_overlap: int = 150
+                ):
+    save_file_path = "/tmp/document_files"
+    knowledge_db_name = get_knowledge_rename(knowledge_db_name)
     # 指定保存文件的文件夹
-    if not os.path.exists(SAVE_FILE_PATH):
-        os.makedirs(SAVE_FILE_PATH)
+    if not os.path.exists(save_file_path):
+        os.makedirs(save_file_path)
     if files is None or len(files) == 0:
         print('no file need save')
+    if knowledge_db_name in knowledge_mgr.get_all():
+        knowledge_db_cls = creat_knowledge_db(knowledge_db_name, is_regist=False)
+    else:
+        knowledge_db_cls = creat_knowledge_db(knowledge_db_name)
+    # 注册文档加载器
+    loader_mng = LoaderMng()
+    # 加载文档加载器，可以使用mxrag自有的，也可以使用langchain的
+    loader_mng.register_loader(loader_class=TextLoader, file_types=[".txt", ".md"])
+    loader_mng.register_loader(loader_class=DocxLoader, file_types=[".docx"])
+    loader_mng.register_loader(loader_class=PdfLoader, file_types=[".pdf"])
+    loader_mng.register_loader(loader_class=ExcelLoader, file_types=[".xlsx", ".xls"])
+
+    # 加载文档切分器，使用langchain的
+    loader_mng.register_splitter(splitter_class=RecursiveCharacterTextSplitter,
+                                 file_types=[".docx", ".txt", ".md", ".pdf", ".xlsx", ".xls"],
+                                 splitter_params={"chunk_size": chunk_size,
+                                                  "chunk_overlap": chunk_overlap,
+                                                  "keep_separator": False
+                                                  })
     for file in files:
         try:
             # 上传领域知识文档
-            shutil.copy(file.name, SAVE_FILE_PATH)
+            shutil.copy(file.name, save_file_path)
             # 知识库:chunk\embedding\add
-            upload_files(text_knowledge_db, [file.name], loader_mng=loader_mng, embed_func=text_emb.embed_documents,
-                         force=True)
-            print(f"file {file.name} save to {SAVE_FILE_PATH}.")
+            upload_files(knowledge_db_cls, [file.name], loader_mng=loader_mng, embed_func=text_emb.embed_documents, force=True)
+            print(f"file {file.name} save to {save_file_path}.")
         except Exception as err:
             logger.error(f"save failed, find exception: {err}")
 
@@ -83,6 +386,7 @@ def file_change(files, upload_btn):
     print("file changes")
 
 
+# 构建gradio框架
 def build_demo():
     with gr.Blocks() as demo:
         gr.HTML("<center><h1>检索增强生成(RAG)对话</h1><p>powered by MindX RAG</p><center>")
@@ -94,14 +398,51 @@ def build_demo():
                     files = gr.Files(
                         height=300,
                         file_count="multiple",
-                        file_types=[".docx", ".txt", ".md", ".pdf"],
+                        file_types=[".docx", ".txt", ".md", ".pdf", ".xlsx", ".xls"],
                         interactive=True,
                         label="上传知识库文档"
                     )
                 with gr.Row():
                     upload_btn = gr.Button("上传文件")
                 with gr.Row():
-                    with gr.Accordion(label='知识库情况'):
+                    with gr.TabItem(label='知识库情况'):
+                        set_knowledge_name = gr.Textbox(label='设置当前知识库',placeholder='在此输入知识库名称，默认使用test知识库')
+                        with gr.Row():
+                            creat_knowledge_btn = gr.Button('创建知识库')
+                            delete_knowledge_btn = gr.Button('删除知识库')
+                        knowledge_name_output = gr.Textbox(label='知识库列表')
+                        knowledge_number_output = gr.Textbox(label='知识库数量')
+                        with gr.Row():
+                            show_knowledge_btn = gr.Button('显示知识库')
+                    with gr.TabItem(label='文件情况'):
+                        knowledge_name = gr.Textbox('知识库名称')
+                        knowledge_file_output = gr.Textbox('知识库文件列表')
+                        knowledge_file_num_output = gr.Textbox('知识库文件数量')
+                        with gr.Row():
+                            knowledge_file_out_btn = gr.Button('显示文件情况')
+                            knowledge_clear_btn = gr.Button('清空知识库')
+                with gr.Row():
+                    with gr.According(label='文档切分参数设置',open=False):
+                        chunk_size = gr.Slider(
+                            minimum=50,
+                            maximum=5000,
+                            value=750,
+                            step=50,
+                            interactive=True,
+                            label="chunk_size",
+                            info="文本切分长度"
+                        )
+                        chunk_overlap = gr.Slider(
+                            minimum=10,
+                            maximum=500,
+                            value=150,
+                            step=10,
+                            interactive=True,
+                            label="chunk_overlap",
+                            info="文本切分填充长度"
+                        )
+                with gr.Row():
+                    with gr.According(label='大模型参数设置', open=False):
                         temperature = gr.Slider(
                             minimum=0.01,
                             maximum=2,
@@ -129,101 +470,84 @@ def build_demo():
                             label="最大tokens",
                             info="输入+输出最多的tokens数"
                         )
-            with gr.Column(scale=200):
-                chatbot = gr.Chatbot(height=500)
-                with gr.Row():
-                    msg = gr.Textbox(placeholder="在此输入问题...", container=False)
+                        is_rewrite = gr.Radio(['是','否'], value='否', label="是否根据历史提问重写问题？")
+                        history_n = gr.Slider(
+                            minimum=1,
+                            maximum=10,
+                            value=5,
+                            step=1,
+                            interactive=True,
+                            label="历史提问重新轮数",
+                            info="问题重写时所参考的历史提问轮数"
+                        )
                 with gr.Row():
-                    send_btn = gr.Button(value="发送", variant="primary")
-                    clean_btn = gr.Button(value="清空历史")
-            send_btn.click(user_query, [msg, chatbot], [msg, chatbot], queue=False).then(bot_response,
-                                                                                         [chatbot, max_tokens,
-                                                                                          temperature, top_p], chatbot)
+                    with gr.According(label='检索参数设置', open=False):
+                        score_threshold = gr.Slider(
+                            minimum=0,
+                            maximum=1,
+                            value=0.5,
+                            step=0.01,
+                            interactive=True,
+                            label="score_threshold",
+                            info="相似性检索阈值，值越大表示越相关,低于阈值不会被返回。"
+                        )
+                        top_k = gr.Slider(
+                            minimum=1,
+                            maximum=5,
+                            value=1,
+                            step=1,
+                            interactive=True,
+                            label="top_k",
+                            info="相似性检索返回条数"
+                        )
+                        show_type = gr.Radio(['对话结束后显示','检索框单独显示','不显示'], value="对话结束后显示", label="知识库文档匹配结果展示方式选择")
+            with gr.Column(scale=200):
+                with gr.Tabs():
+                    with gr.TabItem("对话窗口"):
+                        chat_type = gr.Radio(['RAG检索增强对话','仅大模型对话'], value="RAG检索增强对话", label="请选择对话模式？")
+                        chatbot = gr.Chatbot(height=550)
+                        with gr.Row():
+                            msg = gr.Textbox(placeholder="在此输入问题...", container=False)
+                        with gr.Row():
+                            send_btn = gr.Button(value="发送", variant="primary")
+                            clean_btn = gr.Button(value="清空历史")
+                    with gr.TabItem("检索窗口"):
+                        chatbot_r = gr.Chatbot(height=550)
+                        with gr.Row():
+                            msg_r = gr.Textbox(placeholder="在此输入问题...", container=False)
+                        with gr.Row():
+                            send_btn_r = gr.Button(value="文档检索", variant="primary")
+                            clean_btn_r = gr.Button(value="清空历史")
+            # RAG发送
+            send_btn.click(user_query, [msg, chatbot], [msg, chatbot], queue=False
+                           ).then(bot_response,
+                                  [chatbot, chatbot_r, max_tokens, temperature, top_p, history_n, score_threshold, top_k, chat_type, show_type, is_rewrite, set_knowledge_name],
+                                  [chatbot, chatbot_r])
+            # RAG清除历史
             clean_btn.click(clear_history, chatbot, chatbot)
+            # 上传文件
             files.change(file_change, [], [])
-            upload_btn.click(file_upload, files, []).then(clear_history, files, files)
+            upload_btn.click(file_upload, [files, set_knowledge_name, chunk_size, chunk_overlap], files)
+            # 管理所有知识库
+            creat_knowledge_btn.click(creat_knowledge_db, [set_knowledge_name],[]).then(get_knowledge_db, [],[knowledge_name_output, knowledge_number_output])
+            show_knowledge_btn.click(get_knowledge_db, [], [knowledge_name_output, knowledge_number_output])
+            delete_knowledge_btn.click(delete_knowledge_db, [set_knowledge_name], [knowledge_name_output, knowledge_number_output])
+            # 管理知识库里文件
+            knowledge_file_out_btn.click(get_document, [set_knowledge_name], [knowledge_name, knowledge_file_output, knowledge_file_num_output])
+            knowledge_clear_btn.click(clear_document, [set_knowledge_name], [knowledge_name, knowledge_file_output, knowledge_file_num_output])
+            # 检索发送
+            send_btn_r.click(user_retriever, [msg_r, chatbot_r], [msg_r, chatbot_r], queue=False
+                             ).then(re_response, [chatbot_r, score_threshold, top_k, set_knowledge_name], chatbot_r)
+            # 检索清除历史
+            clean_btn_r.click(clear_history, chatbot_r, chatbot_r)
     return demo
 
 
-def create_gradio(port):
+def create_gradio(ports):
     demo = build_demo()
     demo.queue()
-    demo.launch(share=True, server_name="0.0.0.0", server_port=port)
-
-
-class CustomFormatter(argparse.ArgumentDefaultsHelpFormatter):
-    def _get_default_metavar_for_optional(self, action):
-        return action.type.__name__
-
-    def _get_default_metavar_for_positional(self, action):
-        return action.type.__name__
+    demo.launch(share=True, server_name="0.0.0.0", server_port=ports)
 
 
 if __name__ == '__main__':
-    parse = argparse.ArgumentParser(formatter_class=CustomFormatter)
-    parse.add_argument("--embedding_path", type=str, default="/home/data/acge_text_embedding",
-                       help="embedding模型本地路径")
-    parse.add_argument("--embedding_dim", type=int, default=1024, help="embedding模型向量维度")
-    parse.add_argument("--llm_url", type=str, default="http://127.0.0.1:1025/v1/chat/completions", help="大模型url地址")
-    parse.add_argument("--model_name", type=str, default="Llama3-8B-Chinese-Chat", help="大模型名称")
-    parse.add_argument("--port", type=str, default=8080, help="web后端端口")
-
-    args = parse.parse_args()
-
-    try:
-        # 离线构建知识库,首先注册文档处理器
-        loader_mng = LoaderMng()
-        # 加载文档加载器，可以使用mxrag自有的，也可以使用langchain的
-        loader_mng.register_loader(loader_class=TextLoader, file_types=[".txt", ".md"])
-        loader_mng.register_loader(loader_class=DocxLoader, file_types=[".docx"])
-        loader_mng.register_loader(loader_class=PdfLoader, file_types=[".pdf"])
-
-        # 加载文档切分器，使用langchain的
-        loader_mng.register_splitter(splitter_class=RecursiveCharacterTextSplitter,
-                                     file_types=[".docx", ".txt", ".md", ".pdf"],
-                                     splitter_params={"chunk_size": 750,
-                                                      "chunk_overlap": 150,
-                                                      "keep_separator": False
-                                                      }
-                                     )
-
-        # 设置向量检索使用的npu卡，具体可以用的卡可执行npu-smi info查询获取
-        dev = 0
-
-        # 初始化向量数据库
-        vector_store = MindFAISS(x_dim=args.embedding_dim,
-                                 similarity_strategy=SimilarityStrategy.FLAT_L2,
-                                 devs=[dev],
-                                 load_local_index="./faiss.index",
-                                 auto_save=True
-                                 )
-        # 初始化文档chunk关系数据库
-        chunk_store = SQLiteDocstore(db_path="./sql.db")
-        # 初始化知识管理关系数据库
-        knowledge_store = KnowledgeStore(db_path="./sql.db")
-        # 初始化知识库管理
-        text_knowledge_db = KnowledgeDB(knowledge_store=knowledge_store,
-                                        chunk_store=chunk_store,
-                                        vector_store=vector_store,
-                                        knowledge_name="test",
-                                        white_paths=["/tmp"]
-                                        )
-
-        # 加载embedding模型，请根据模型具体路径适配
-        text_emb = TextEmbedding(model_path=args.embedding_path, dev_id=dev)
-
-        # Step2在线问题答复,初始化检索器
-        text_retriever = Retriever(vector_store=vector_store,
-                                   document_store=chunk_store,
-                                   embed_func=text_emb.embed_documents,
-                                   k=1,
-                                   score_threshold=0.4
-                                   )
-        # 配置text生成text大模型chain，具体ip端口请根据实际情况适配修改
-        llm = Text2TextLLM(base_url=args.llm_url, model_name=args.model_name, client_param=ClientParam(use_http=True))
-        text2text_chain = SingleText2TextChain(llm=llm,
-                                               retriever=text_retriever,
-                                               )
-        create_gradio(int(args.port))
-    except Exception as e:
-        logger.info(f"exception happed :{e}")
+    create_gradio(port)
-- 
Gitee


From 01ec67c2a0a8da930d0867ac111d47f4b2ed08f4 Mon Sep 17 00:00:00 2001
From: tanfeng <823018000@qq.com>
Date: Mon, 25 Nov 2024 11:07:33 +0800
Subject: [PATCH 2/6] =?UTF-8?q?=E9=80=82=E9=85=8D=E6=9C=80=E6=96=B0POC?=
 =?UTF-8?q?=E7=89=88=E6=9C=AC=E5=B1=95=E7=A4=BA=E9=A1=B5=E9=9D=A2=EF=BC=8C?=
 =?UTF-8?q?=E5=8A=9F=E8=83=BD=E5=A2=9E=E5=8A=A0?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 mxRAG/PocValidation/chat_with_ascend/app.py | 731 ++++++++++----------
 1 file changed, 359 insertions(+), 372 deletions(-)

diff --git a/mxRAG/PocValidation/chat_with_ascend/app.py b/mxRAG/PocValidation/chat_with_ascend/app.py
index d5805bfb8..5d6dea138 100644
--- a/mxRAG/PocValidation/chat_with_ascend/app.py
+++ b/mxRAG/PocValidation/chat_with_ascend/app.py
@@ -24,77 +24,148 @@ from mx_rag.document.loader import DocxLoader, PdfLoader, ExcelLoader
 from mx_rag.knowledge import KnowledgeStore, KnowledgeDB, KnowledgeMgrStore, KnowledgeMgr
 from mx_rag.knowledge.handler import upload_files, LoaderMng
 
-
-class CustomFormatter(argparse.ArgumentDefaultsHelpFormatter):
-    def _get_default_metavar_for_optional(self, action):
-        return action.type.__name__
-
-    def _get_default_metavar_for_positional(self, action):
-        return action.type.__name__
-
-
-parse = argparse.ArgumentParser(formatter_class=CustomFormatter)
-parse.add_argument("--embedding_path", type=str, default="/home/data/acge_text_embedding", help="embedding模型本地路径")
-parse.add_argument("--tei_emb", type=bool, default=False, help="是否使用TEI服务化的embedding模型")
-parse.add_argument("--embedding_url", type=str, default="http://127.0.0.1:8080/embed",help="使用TEI服务化的embedding模型url地址")
-parse.add_argument("--embedding_dim", type=int, default=1024, help="embedding模型向量维度")
-parse.add_argument("--llm_url", type=str, default="http://127.0.0.1:1025/v1/chat/completions", help="大模型url地址")
-parse.add_argument("--model_name", type=str, default="Llama3-8B-Chinese-Chat", help="大模型名称")
-parse.add_argument("--tei_reranker", type=bool, default=False, help="是否使用TEI服务化的reranker模型")
-parse.add_argument("--reranker_path", type=str, default=None, help="reranker模型本地路径")
-parse.add_argument("--reranker_url", type=str, default=None, help="使用TEI服务化的embedding模型url地址")
-parse.add_argument("--dev", type=int, default=0, help="使用的npu卡，可通过执行npu-smi info获取")
-parse.add_argument("--port", type=int, default=8080, help="web后端端口")
-
-args = parse.parse_args().__dict__
-tei_emb: bool = args.pop('tei_emb')
-embedding_path: str = args.pop('embedding_path')
-embedding_url: str = args.pop('embedding_url')
-embedding_dim: int = args.pop('embedding_dim')
-llm_url: str = args.pop('llm_url')
-model_name: str = args.pop('model_name')
-tei_reranker: bool = args.pop('tei_reranker')
-reranker_path: str = args.pop('reranker_path')
-reranker_url: str = args.pop('reranker_url')
-dev: int = args.pop('dev')
-port: int = args.pop('port')
-
-# 初始化向量数据库
-vector_store = MindFAISS(x_dim=embedding_dim,
-                         similarity_strategy=SimilarityStrategy.FLAT_L2,
-                         devs=[dev],
-                         load_local_index="./test_faiss.index",
-                         auto_save=True
-                         )
-# 初始化文档chunk关系数据库
-chunk_store = SQLiteDocstore(db_path="./test_sql.db")
-# 初始化知识管理关系数据库
-knowledge_store = KnowledgeStore(db_path="./test_sql.db")
-# 初始化知识库管理
-text_knowledge_db = KnowledgeDB(knowledge_store=knowledge_store,
-                                chunk_store=chunk_store,
-                                vector_store=vector_store,
-                                knowledge_name="test",
-                                white_paths=["/tmp"]
-                                )
+#初始化数据库管理
 knowledge_mgr = KnowledgeMgr(KnowledgeMgrStore("./test_sql.db"))
 
-# 配置text生成text大模型chain，具体ip端口请根据实际情况适配修改
-llm = Text2TextLLM(base_url=llm_url, model_name=model_name, client_param=ClientParam(use_http=True))
 
-# 配置embedding模型，请根据模型具体路径适配
-if tei_emb:
-    text_emb = TEIEmbedding(url=embedding_url, client_param=ClientParam(use_http=True))
-else:
-    text_emb = TextEmbedding(model_path=embedding_path, dev_id=dev)
+# 创建新的知识库
+def creat_knowledge_db(knowledge_name: str, is_regist: bool = True):
+    knowledge_name = get_knowledge_rename(knowledge_name)
+    index_name, db_name = get_db_names(knowledge_name)
+    vector_store_cls = MindFAISS(x_dim=embedding_dim,
+                                 similarity_strategy=SimilarityStrategy.FLAT_L2,
+                                 devs=[dev],
+                                 load_local_index=index_name,
+                                 auto_save=True)
+    # 初始化文档chunk关系数据库
+    chunk_store_cls = SQLiteDocstore(db_path=db_name)
+    # 初始化知识管理关系数据库
+    knowledge_store_cls = KnowledgeStore(db_path=db_name)
+    # 初始化知识库管理
+    knowledge_db_cls = KnowledgeDB(knowledge_store=knowledge_store_cls,
+                                   chunk_store=chunk_store_cls,
+                                   vector_store=vector_store_cls,
+                                   knowledge_name=knowledge_name,
+                                   white_paths=["/tmp"])
+    if is_regist:
+        knowledge_mgr.register(knowledge_db_cls)
+    return knowledge_db_cls
 
-# 配置rerank,请根据模型具体路径适配
-if tei_reranker:
-    reranker = TEIReranker(url=reranker_url, client_param=ClientParam(use_http=True))
-elif reranker_path is not None:
-    reranker = LocalReranker(model_path=reranker_path, dev_id=dev)
-else:
-    reranker = None
+
+# 创建检索器
+def creat_retriever(knowledge_name: str, top_k, score_threshold):
+    knowledge_name = get_knowledge_rename(knowledge_name)
+    index_name, db_name = get_db_names(knowledge_name)
+    vector_store_ret = MindFAISS(x_dim=embedding_dim,
+                                 similarity_strategy=SimilarityStrategy.FLAT_L2,
+                                 devs=[dev],
+                                 load_local_index=index_name,
+                                 auto_save=True)
+    # 初始化文档chunk关系数据库
+    chunk_store_ret = SQLiteDocstore(db_path=db_name)
+    retriever_cls = Retriever(vector_store=vector_store_ret,
+                              document_store=chunk_store_ret,
+                              embed_func=text_emb.embed_documents,
+                              k=top_k,
+                              score_threshold=score_threshold)
+    return retriever_cls
+
+
+# 删除知识库
+def delete_knowledge_db(knowledge_name: str):
+    knowledge_name = get_knowledge_rename(knowledge_name)
+    knowledge_db_lst = knowledge_mgr.get_all()
+    if knowledge_name in knowledge_db_lst:
+        knowledge_db_cls = creat_knowledge_db(knowledge_name, is_regist=False)
+        clear_document(knowledge_name)
+        knowledge_mgr.delete(knowledge_db_cls)
+    return get_knowledge_db()
+
+
+# 获取知识库中文档列表
+def get_document(knowledge_name: str):
+    knowledge_name = get_knowledge_rename(knowledge_name)
+    knowledge_db_cls = creat_knowledge_db(knowledge_name, is_regist=False)
+    doc_names = knowledge_db_cls.get_all_documents()
+    return knowledge_name, doc_names, len(doc_names)
+
+
+# 清空知识库中文档列表
+def clear_document(knowledge_name: str):
+    knowledge_name, doc_names, doc_cnt = get_document(knowledge_name)
+    if doc_cnt > 0:
+        knowledge_db_cls = creat_knowledge_db(knowledge_name, is_regist=False)
+        for doc_name in doc_names:
+            knowledge_db_cls.delete_file(doc_name)
+        return get_document(knowledge_name)
+    else:
+        return knowledge_name, doc_names, doc_cnt
+
+
+# 获取知识库列表
+def get_knowledge_db():
+    knowledge_db_lst = knowledge_mgr.get_all()
+    return knowledge_db_lst, len(knowledge_db_lst)
+
+
+# 上传知识库
+def file_upload(files,
+                knowledge_db_name: str = 'test',
+                chunk_size: int = 750,
+                chunk_overlap: int = 150
+                ):
+    save_file_path = "/tmp/document_files"
+    knowledge_db_name = get_knowledge_rename(knowledge_db_name)
+    # 指定保存文件的文件夹
+    if not os.path.exists(save_file_path):
+        os.makedirs(save_file_path)
+    if files is None or len(files) == 0:
+        print('no file need save')
+    if knowledge_db_name in knowledge_mgr.get_all():
+        knowledge_db_cls = creat_knowledge_db(knowledge_db_name, is_regist=False)
+    else:
+        knowledge_db_cls = creat_knowledge_db(knowledge_db_name)
+    # 注册文档加载器
+    loader_mng = LoaderMng()
+    # 加载文档加载器，可以使用mxrag自有的，也可以使用langchain的
+    loader_mng.register_loader(loader_class=TextLoader, file_types=[".txt", ".md"])
+    loader_mng.register_loader(loader_class=DocxLoader, file_types=[".docx"])
+    loader_mng.register_loader(loader_class=PdfLoader, file_types=[".pdf"])
+    loader_mng.register_loader(loader_class=ExcelLoader, file_types=[".xlsx", ".xls"])
+
+    # 加载文档切分器，使用langchain的
+    loader_mng.register_splitter(splitter_class=RecursiveCharacterTextSplitter,
+                                 file_types=[".docx", ".txt", ".md", ".pdf", ".xlsx", ".xls"],
+                                 splitter_params={"chunk_size": chunk_size,
+                                                  "chunk_overlap": chunk_overlap,
+                                                  "keep_separator": False
+                                                  })
+    for file in files:
+        try:
+            # 上传领域知识文档
+            shutil.copy(file.name, save_file_path)
+            # 知识库:chunk\embedding\add
+            upload_files(knowledge_db_cls, [file.name], loader_mng=loader_mng, embed_func=text_emb.embed_documents, force=True)
+            print(f"file {file.name} save to {save_file_path}.")
+        except Exception as err:
+            logger.error(f"save failed, find exception: {err}")
+
+
+def file_change(files, upload_btn):
+    print("file changes")
+
+
+def get_db_names(knowledge_name: str):
+    index_name = "./" + knowledge_name + "_faiss.index"
+    db_name = "./" + knowledge_name + "_sql.db"
+    return index_name, db_name
+
+
+def get_knowledge_rename(knowledge_name: str):
+    if knowledge_name is None or len(knowledge_name) == 0 or ' ' in knowledge_name:
+        return 'test'
+    else:
+        return knowledge_name
 
 
 # 历史问题改写
@@ -187,8 +258,9 @@ def bot_response(history,
                 if len(q_docs) > 0:
                     history_last = ''
                     for i ,source in enumerate(q_docs):
-                        sources = "\n====检索信息来源" + str(i + 1) + ":" + source['metadata']['source'] + "====" + "\n" + \
-                                  "--->参考内容：" +source['page_content'] + "\n"
+                        sources = "\n====检索信息来源:" + str(i + 1) + ":" + "[数据库名]:" + str(knowledge_name) + \
+                                  "[文件名]:" + source['metadata']['source'] + "====" + "\n" + \
+                                  "参考内容：" +source['page_content'] + "\n"
                         history_last += sources
                     history += [[None, history_last]]
                 yield history, history_r
@@ -202,8 +274,9 @@ def bot_response(history,
                 if len(q_docs) > 0:
                     history_r_last = ''
                     for i ,source in enumerate(q_docs):
-                        sources = "\n====检索信息来源" + str(i + 1) + ":" + source['metadata']['source'] + "====" + "\n" + \
-                                  "--->参考内容：" +source['page_content'] + "\n"
+                        sources = "\n====检索信息来源:" + str(i + 1) + ":" + "[数据库名]:" + str(knowledge_name) + \
+                                  "[文件名]:" + source['metadata']['source'] + "====" + "\n" + \
+                                  "参考内容：" +source['page_content'] + "\n"
                         history_r_last += sources
                     history_r += [[history[-1][0], history_r_last]]
                 yield history, history_r
@@ -224,8 +297,9 @@ def re_response(history_r,
     if len(q_docs) > 0:
         history_r_last = ''
         for i, source in enumerate(q_docs):
-            sources = "\n====检索信息来源" + str(i + 1) + ":" + source.metadata['source'] + "====" + "\n" + \
-                      "--->参考内容：" + source.page_content + "\n"
+            sources = "\n====检索信息来源:" + str(i + 1) + "[数据库名]:" + str(knowledge_name)  +\
+                      "[文件名]:" + source.metadata['source'] + "====" + "\n" + \
+                      "参考内容：" + source.page_content + "\n"
             history_r_last += sources
         history_r[-1][1] = history_r_last
     else:
@@ -247,307 +321,220 @@ def clear_history(history):
     return []
 
 
-# 创建新的知识库
-def creat_knowledge_db(knowledge_name: str, is_regist: bool=True):
-    knowledge_name = get_knowledge_rename(knowledge_name)
-    index_name, db_name = get_db_names(knowledge_name)
-    vector_store_cls = MindFAISS(x_dim=embedding_dim,
-                                 similarity_strategy=SimilarityStrategy.FLAT_L2,
-                                 devs=[dev],
-                                 load_local_index=index_name,
-                                 auto_save=True)
-    # 初始化文档chunk关系数据库
-    chunk_store_cls = SQLiteDocstore(db_path=db_name)
-    # 初始化知识管理关系数据库
-    knowledge_store_cls = KnowledgeStore(db_path=db_name)
-    # 初始化知识库管理
-    knowledge_db_cls = KnowledgeDB(knowledge_store=knowledge_store_cls,
-                                   chunk_store=chunk_store_cls,
-                                   vector_store=vector_store_cls,
-                                   knowledge_name=knowledge_name,
-                                   white_paths=["/tmp"])
-    if is_regist:
-        knowledge_mgr.register(knowledge_db_cls)
-    return knowledge_db_cls
-
-
-def creat_retriever(knowledge_name: str, top_k, score_threshold):
-    knowledge_name = get_knowledge_rename(knowledge_name)
-    index_name, db_name = get_db_names(knowledge_name)
-    vector_store_ret = MindFAISS(x_dim=embedding_dim,
-                                 similarity_strategy=SimilarityStrategy.FLAT_L2,
-                                 devs=[dev],
-                                 load_local_index=index_name,
-                                 auto_save=True)
-    # 初始化文档chunk关系数据库
-    chunk_store_ret = SQLiteDocstore(db_path=db_name)
-    retriever_cls = Retriever(vector_store=vector_store_ret,
-                              document_store=chunk_store_ret,
-                              embed_func=text_emb.embed_documents,
-                              k=top_k,
-                              score_threshold=score_threshold)
-    return retriever_cls
-
-
-# 删除知识库
-def delete_knowledge_db(knowledge_name: str):
-    knowledge_name = get_knowledge_rename(knowledge_name)
-    knowledge_db_lst = knowledge_mgr.get_all()
-    if knowledge_name in knowledge_db_lst:
-        knowledge_db_cls = creat_knowledge_db(knowledge_name, is_regist=False)
-        clear_document(knowledge_name)
-        knowledge_mgr.delete(knowledge_db_cls)
-    return get_knowledge_db()
-
-
-# 获取知识库文档列表
-def get_document(knowledge_name: str):
-    knowledge_name = get_knowledge_rename(knowledge_name)
-    knowledge_db_cls = creat_knowledge_db(knowledge_name, is_regist=False)
-    doc_names = knowledge_db_cls.get_all_documents()
-    return knowledge_name, doc_names, len(doc_names)
-
-
-# 清空知识库文档列表
-def clear_document(knowledge_name: str):
-    knowledge_name, doc_names, doc_cnt = get_document(knowledge_name)
-    if doc_cnt > 0:
-        knowledge_db_cls = creat_knowledge_db(knowledge_name, is_regist=False)
-        for doc_name in doc_names:
-            knowledge_db_cls.delete_file(doc_name)
-        return get_document(knowledge_name)
-    else:
-        return knowledge_name, doc_names, doc_cnt
-
-
-# 获取知识库列表
-def get_knowledge_db():
-    knowledge_db_lst = knowledge_mgr.get_all()
-    return knowledge_db_lst, len(knowledge_db_lst)
-
-
-def get_db_names(knowledge_name: str):
-    index_name = "./" + knowledge_name + "_faiss.index"
-    dn_name = "./" + knowledge_name + "_sql.db"
-    return index_name, dn_name
-
-
-def get_knowledge_rename(knowledge_name: str):
-    if knowledge_name is None or len(knowledge_name) == 0 or ' ' in knowledge_name:
-        return 'test'
+if __name__ == '__main__':
+    class CustomFormatter(argparse.ArgumentDefaultsHelpFormatter):
+        def _get_default_metavar_for_optional(self, action):
+            return action.type.__name__
+
+        def _get_default_metavar_for_positional(self, action):
+            return action.type.__name__
+
+    parse = argparse.ArgumentParser(formatter_class=CustomFormatter)
+    parse.add_argument("--embedding_path", type=str, default="/home/data/acge_text_embedding",
+                       help="embedding模型本地路径")
+    parse.add_argument("--tei_emb", type=bool, default=False, help="是否使用TEI服务化的embedding模型")
+    parse.add_argument("--embedding_url", type=str, default="http://127.0.0.1:8080/embed",
+                       help="使用TEI服务化的embedding模型url地址")
+    parse.add_argument("--embedding_dim", type=int, default=1024, help="embedding模型向量维度")
+    parse.add_argument("--llm_url", type=str, default="http://127.0.0.1:1025/v1/chat/completions", help="大模型url地址")
+    parse.add_argument("--model_name", type=str, default="Llama3-8B-Chinese-Chat", help="大模型名称")
+    parse.add_argument("--tei_reranker", type=bool, default=False, help="是否使用TEI服务化的reranker模型")
+    parse.add_argument("--reranker_path", type=str, default=None, help="reranker模型本地路径")
+    parse.add_argument("--reranker_url", type=str, default=None, help="使用TEI服务化的embedding模型url地址")
+    parse.add_argument("--dev", type=int, default=0, help="使用的npu卡，可通过执行npu-smi info获取")
+    parse.add_argument("--port", type=int, default=8080, help="web后端端口")
+
+    args = parse.parse_args().__dict__
+    embedding_path: str = args.pop('embedding_path')
+    tei_emb: bool = args.pop('tei_emb')
+    embedding_url: str = args.pop('embedding_url')
+    embedding_dim: int = args.pop('embedding_dim')
+    llm_url: str = args.pop('llm_url')
+    model_name: str = args.pop('model_name')
+    tei_reranker: bool = args.pop('tei_reranker')
+    reranker_path: str = args.pop('reranker_path')
+    reranker_url: str = args.pop('reranker_url')
+    dev: int = args.pop('dev')
+    port: int = args.pop('port')
+
+    # 初始化test数据库
+    knowledge_db = creat_knowledge_db('test')
+    # 配置text生成text大模型chain，具体ip端口请根据实际情况适配修改
+    llm = Text2TextLLM(base_url=llm_url, model_name=model_name, client_param=ClientParam(use_http=True))
+    # 配置embedding模型，请根据模型具体路径适配
+    if tei_emb:
+        text_emb = TEIEmbedding(url=embedding_url, client_param=ClientParam(use_http=True))
     else:
-        return knowledge_name
-
-
-# 上传知识库
-def file_upload(files,
-                knowledge_db_name: str = 'test',
-                chunk_size: int = 750,
-                chunk_overlap: int = 150
-                ):
-    save_file_path = "/tmp/document_files"
-    knowledge_db_name = get_knowledge_rename(knowledge_db_name)
-    # 指定保存文件的文件夹
-    if not os.path.exists(save_file_path):
-        os.makedirs(save_file_path)
-    if files is None or len(files) == 0:
-        print('no file need save')
-    if knowledge_db_name in knowledge_mgr.get_all():
-        knowledge_db_cls = creat_knowledge_db(knowledge_db_name, is_regist=False)
+        text_emb = TextEmbedding(model_path=embedding_path, dev_id=dev)
+    # 配置rerank,请根据模型具体路径适配
+    if tei_reranker:
+        reranker = TEIReranker(url=reranker_url, client_param=ClientParam(use_http=True))
+    elif reranker_path is not None:
+        reranker = LocalReranker(model_path=reranker_path, dev_id=dev)
     else:
-        knowledge_db_cls = creat_knowledge_db(knowledge_db_name)
-    # 注册文档加载器
-    loader_mng = LoaderMng()
-    # 加载文档加载器，可以使用mxrag自有的，也可以使用langchain的
-    loader_mng.register_loader(loader_class=TextLoader, file_types=[".txt", ".md"])
-    loader_mng.register_loader(loader_class=DocxLoader, file_types=[".docx"])
-    loader_mng.register_loader(loader_class=PdfLoader, file_types=[".pdf"])
-    loader_mng.register_loader(loader_class=ExcelLoader, file_types=[".xlsx", ".xls"])
-
-    # 加载文档切分器，使用langchain的
-    loader_mng.register_splitter(splitter_class=RecursiveCharacterTextSplitter,
-                                 file_types=[".docx", ".txt", ".md", ".pdf", ".xlsx", ".xls"],
-                                 splitter_params={"chunk_size": chunk_size,
-                                                  "chunk_overlap": chunk_overlap,
-                                                  "keep_separator": False
-                                                  })
-    for file in files:
-        try:
-            # 上传领域知识文档
-            shutil.copy(file.name, save_file_path)
-            # 知识库:chunk\embedding\add
-            upload_files(knowledge_db_cls, [file.name], loader_mng=loader_mng, embed_func=text_emb.embed_documents, force=True)
-            print(f"file {file.name} save to {save_file_path}.")
-        except Exception as err:
-            logger.error(f"save failed, find exception: {err}")
-
-
-def file_change(files, upload_btn):
-    print("file changes")
-
-
-# 构建gradio框架
-def build_demo():
-    with gr.Blocks() as demo:
-        gr.HTML("<center><h1>检索增强生成(RAG)对话</h1><p>powered by MindX RAG</p><center>")
-        with gr.Row():
-            with gr.Column(scale=100):
-                with gr.Row():
-                    model_select = gr.Dropdown(choices=["Default"], value="Default", container=False, interactive=True)
-                with gr.Row():
-                    files = gr.Files(
-                        height=300,
-                        file_count="multiple",
-                        file_types=[".docx", ".txt", ".md", ".pdf", ".xlsx", ".xls"],
-                        interactive=True,
-                        label="上传知识库文档"
-                    )
-                with gr.Row():
-                    upload_btn = gr.Button("上传文件")
-                with gr.Row():
-                    with gr.TabItem(label='知识库情况'):
-                        set_knowledge_name = gr.Textbox(label='设置当前知识库',placeholder='在此输入知识库名称，默认使用test知识库')
-                        with gr.Row():
-                            creat_knowledge_btn = gr.Button('创建知识库')
-                            delete_knowledge_btn = gr.Button('删除知识库')
-                        knowledge_name_output = gr.Textbox(label='知识库列表')
-                        knowledge_number_output = gr.Textbox(label='知识库数量')
-                        with gr.Row():
-                            show_knowledge_btn = gr.Button('显示知识库')
-                    with gr.TabItem(label='文件情况'):
-                        knowledge_name = gr.Textbox('知识库名称')
-                        knowledge_file_output = gr.Textbox('知识库文件列表')
-                        knowledge_file_num_output = gr.Textbox('知识库文件数量')
-                        with gr.Row():
-                            knowledge_file_out_btn = gr.Button('显示文件情况')
-                            knowledge_clear_btn = gr.Button('清空知识库')
-                with gr.Row():
-                    with gr.According(label='文档切分参数设置',open=False):
-                        chunk_size = gr.Slider(
-                            minimum=50,
-                            maximum=5000,
-                            value=750,
-                            step=50,
+        reranker = None
+
+    # 构建gradio框架
+    def build_demo():
+        with (gr.Blocks() as demo):
+            gr.HTML("<center><h1>检索增强生成(RAG)对话</h1><p>powered by MindX RAG</p><center>")
+            with gr.Row():
+                with gr.Column(scale=100):
+                    with gr.Row():
+                        model_select = gr.Dropdown(choices=["Default"], value="Default", container=False, interactive=True)
+                    with gr.Row():
+                        files = gr.Files(
+                            height=300,
+                            file_count="multiple",
+                            file_types=[".docx", ".txt", ".md", ".pdf", ".xlsx", ".xls"],
                             interactive=True,
-                            label="chunk_size",
-                            info="文本切分长度"
+                            label="上传知识库文档"
                         )
-                        chunk_overlap = gr.Slider(
-                            minimum=10,
-                            maximum=500,
-                            value=150,
-                            step=10,
-                            interactive=True,
-                            label="chunk_overlap",
-                            info="文本切分填充长度"
-                        )
-                with gr.Row():
-                    with gr.According(label='大模型参数设置', open=False):
-                        temperature = gr.Slider(
-                            minimum=0.01,
-                            maximum=2,
-                            value=0.5,
-                            step=0.01,
-                            interactive=True,
-                            label="温度",
-                            info="Token生成的随机性"
-                        )
-                        top_p = gr.Slider(
-                            minimum=0.01,
-                            maximum=1,
-                            value=0.95,
-                            step=0.05,
-                            interactive=True,
-                            label="Top P",
-                            info="累计概率总和阈值"
-                        )
-                        max_tokens = gr.Slider(
-                            minimum=100,
-                            maximum=1024,
-                            value=512,
-                            step=1,
-                            interactive=True,
-                            label="最大tokens",
-                            info="输入+输出最多的tokens数"
-                        )
-                        is_rewrite = gr.Radio(['是','否'], value='否', label="是否根据历史提问重写问题？")
-                        history_n = gr.Slider(
-                            minimum=1,
-                            maximum=10,
-                            value=5,
-                            step=1,
-                            interactive=True,
-                            label="历史提问重新轮数",
-                            info="问题重写时所参考的历史提问轮数"
-                        )
-                with gr.Row():
-                    with gr.According(label='检索参数设置', open=False):
-                        score_threshold = gr.Slider(
-                            minimum=0,
-                            maximum=1,
-                            value=0.5,
-                            step=0.01,
-                            interactive=True,
-                            label="score_threshold",
-                            info="相似性检索阈值，值越大表示越相关,低于阈值不会被返回。"
-                        )
-                        top_k = gr.Slider(
-                            minimum=1,
-                            maximum=5,
-                            value=1,
-                            step=1,
-                            interactive=True,
-                            label="top_k",
-                            info="相似性检索返回条数"
-                        )
-                        show_type = gr.Radio(['对话结束后显示','检索框单独显示','不显示'], value="对话结束后显示", label="知识库文档匹配结果展示方式选择")
-            with gr.Column(scale=200):
-                with gr.Tabs():
-                    with gr.TabItem("对话窗口"):
-                        chat_type = gr.Radio(['RAG检索增强对话','仅大模型对话'], value="RAG检索增强对话", label="请选择对话模式？")
-                        chatbot = gr.Chatbot(height=550)
-                        with gr.Row():
-                            msg = gr.Textbox(placeholder="在此输入问题...", container=False)
-                        with gr.Row():
-                            send_btn = gr.Button(value="发送", variant="primary")
-                            clean_btn = gr.Button(value="清空历史")
-                    with gr.TabItem("检索窗口"):
-                        chatbot_r = gr.Chatbot(height=550)
-                        with gr.Row():
-                            msg_r = gr.Textbox(placeholder="在此输入问题...", container=False)
-                        with gr.Row():
-                            send_btn_r = gr.Button(value="文档检索", variant="primary")
-                            clean_btn_r = gr.Button(value="清空历史")
-            # RAG发送
-            send_btn.click(user_query, [msg, chatbot], [msg, chatbot], queue=False
-                           ).then(bot_response,
-                                  [chatbot, chatbot_r, max_tokens, temperature, top_p, history_n, score_threshold, top_k, chat_type, show_type, is_rewrite, set_knowledge_name],
-                                  [chatbot, chatbot_r])
-            # RAG清除历史
-            clean_btn.click(clear_history, chatbot, chatbot)
-            # 上传文件
-            files.change(file_change, [], [])
-            upload_btn.click(file_upload, [files, set_knowledge_name, chunk_size, chunk_overlap], files)
-            # 管理所有知识库
-            creat_knowledge_btn.click(creat_knowledge_db, [set_knowledge_name],[]).then(get_knowledge_db, [],[knowledge_name_output, knowledge_number_output])
-            show_knowledge_btn.click(get_knowledge_db, [], [knowledge_name_output, knowledge_number_output])
-            delete_knowledge_btn.click(delete_knowledge_db, [set_knowledge_name], [knowledge_name_output, knowledge_number_output])
-            # 管理知识库里文件
-            knowledge_file_out_btn.click(get_document, [set_knowledge_name], [knowledge_name, knowledge_file_output, knowledge_file_num_output])
-            knowledge_clear_btn.click(clear_document, [set_knowledge_name], [knowledge_name, knowledge_file_output, knowledge_file_num_output])
-            # 检索发送
-            send_btn_r.click(user_retriever, [msg_r, chatbot_r], [msg_r, chatbot_r], queue=False
-                             ).then(re_response, [chatbot_r, score_threshold, top_k, set_knowledge_name], chatbot_r)
-            # 检索清除历史
-            clean_btn_r.click(clear_history, chatbot_r, chatbot_r)
-    return demo
-
-
-def create_gradio(ports):
-    demo = build_demo()
-    demo.queue()
-    demo.launch(share=True, server_name="0.0.0.0", server_port=ports)
-
-
-if __name__ == '__main__':
+                    with gr.Row():
+                        upload_btn = gr.Button("上传文件")
+                    with gr.Row():
+                        with gr.TabItem(label='知识库情况'):
+                            set_knowledge_name = gr.Textbox(label='设置当前知识库',placeholder='在此输入知识库名称，默认使用test知识库')
+                            with gr.Row():
+                                creat_knowledge_btn = gr.Button('创建知识库')
+                                delete_knowledge_btn = gr.Button('删除知识库')
+                            knowledge_name_output = gr.Textbox(label='知识库列表')
+                            knowledge_number_output = gr.Textbox(label='知识库数量')
+                            with gr.Row():
+                                show_knowledge_btn = gr.Button('显示知识库')
+                        with gr.TabItem(label='文件情况'):
+                            knowledge_name = gr.Textbox('知识库名称')
+                            knowledge_file_output = gr.Textbox('知识库文件列表')
+                            knowledge_file_num_output = gr.Textbox('知识库文件数量')
+                            with gr.Row():
+                                knowledge_file_out_btn = gr.Button('显示文件情况')
+                                knowledge_clear_btn = gr.Button('清空知识库')
+                    with gr.Row():
+                        with gr.According(label='文档切分参数设置',open=False):
+                            chunk_size = gr.Slider(
+                                minimum=50,
+                                maximum=5000,
+                                value=750,
+                                step=50,
+                                interactive=True,
+                                label="chunk_size",
+                                info="文本切分长度"
+                            )
+                            chunk_overlap = gr.Slider(
+                                minimum=10,
+                                maximum=500,
+                                value=150,
+                                step=10,
+                                interactive=True,
+                                label="chunk_overlap",
+                                info="文本切分填充长度"
+                            )
+                    with gr.Row():
+                        with gr.According(label='大模型参数设置', open=False):
+                            temperature = gr.Slider(
+                                minimum=0.01,
+                                maximum=2,
+                                value=0.5,
+                                step=0.01,
+                                interactive=True,
+                                label="温度",
+                                info="Token生成的随机性"
+                            )
+                            top_p = gr.Slider(
+                                minimum=0.01,
+                                maximum=1,
+                                value=0.95,
+                                step=0.05,
+                                interactive=True,
+                                label="Top P",
+                                info="累计概率总和阈值"
+                            )
+                            max_tokens = gr.Slider(
+                                minimum=100,
+                                maximum=1024,
+                                value=512,
+                                step=1,
+                                interactive=True,
+                                label="最大tokens",
+                                info="输入+输出最多的tokens数"
+                            )
+                            is_rewrite = gr.Radio(['是','否'], value='否', label="是否根据历史提问重写问题？")
+                            history_n = gr.Slider(
+                                minimum=1,
+                                maximum=10,
+                                value=5,
+                                step=1,
+                                interactive=True,
+                                label="历史提问重新轮数",
+                                info="问题重写时所参考的历史提问轮数"
+                            )
+                    with gr.Row():
+                        with gr.According(label='检索参数设置', open=False):
+                            score_threshold = gr.Slider(
+                                minimum=0,
+                                maximum=1,
+                                value=0.5,
+                                step=0.01,
+                                interactive=True,
+                                label="score_threshold",
+                                info="相似性检索阈值，值越大表示越相关,低于阈值不会被返回。"
+                            )
+                            top_k = gr.Slider(
+                                minimum=1,
+                                maximum=5,
+                                value=1,
+                                step=1,
+                                interactive=True,
+                                label="top_k",
+                                info="相似性检索返回条数"
+                            )
+                            show_type = gr.Radio(['对话结束后显示','检索框单独显示','不显示'], value="对话结束后显示", label="知识库文档匹配结果展示方式选择")
+                with gr.Column(scale=200):
+                    with gr.Tabs():
+                        with gr.TabItem("对话窗口"):
+                            chat_type = gr.Radio(['RAG检索增强对话','仅大模型对话'], value="RAG检索增强对话", label="请选择对话模式？")
+                            chatbot = gr.Chatbot(height=550)
+                            with gr.Row():
+                                msg = gr.Textbox(placeholder="在此输入问题...", container=False)
+                            with gr.Row():
+                                send_btn = gr.Button(value="发送", variant="primary")
+                                clean_btn = gr.Button(value="清空历史")
+                        with gr.TabItem("检索窗口"):
+                            chatbot_r = gr.Chatbot(height=550)
+                            with gr.Row():
+                                msg_r = gr.Textbox(placeholder="在此输入问题...", container=False)
+                            with gr.Row():
+                                send_btn_r = gr.Button(value="文档检索", variant="primary")
+                                clean_btn_r = gr.Button(value="清空历史")
+                # RAG发送
+                send_btn.click(user_query, [msg, chatbot], [msg, chatbot], queue=False
+                               ).then(bot_response,
+                                      [chatbot, chatbot_r, max_tokens, temperature, top_p, history_n, score_threshold, top_k, chat_type, show_type, is_rewrite, set_knowledge_name],
+                                      [chatbot, chatbot_r])
+                # RAG清除历史
+                clean_btn.click(clear_history, chatbot, chatbot)
+                # 上传文件
+                files.change(file_change, [], [])
+                upload_btn.click(file_upload, [files, set_knowledge_name, chunk_size, chunk_overlap], files)
+                # 管理所有知识库
+                creat_knowledge_btn.click(creat_knowledge_db, [set_knowledge_name],[]).then(get_knowledge_db, [],[knowledge_name_output, knowledge_number_output])
+                show_knowledge_btn.click(get_knowledge_db, [], [knowledge_name_output, knowledge_number_output])
+                delete_knowledge_btn.click(delete_knowledge_db, [set_knowledge_name], [knowledge_name_output, knowledge_number_output])
+                # 管理知识库里文件
+                knowledge_file_out_btn.click(get_document, [set_knowledge_name], [knowledge_name, knowledge_file_output, knowledge_file_num_output])
+                knowledge_clear_btn.click(clear_document, [set_knowledge_name], [knowledge_name, knowledge_file_output, knowledge_file_num_output])
+                # 检索发送
+                send_btn_r.click(user_retriever, [msg_r, chatbot_r], [msg_r, chatbot_r], queue=False
+                                 ).then(re_response, [chatbot_r, score_threshold, top_k, set_knowledge_name], chatbot_r)
+                # 检索清除历史
+                clean_btn_r.click(clear_history, chatbot_r, chatbot_r)
+        return demo
+
+
+    def create_gradio(ports):
+        demo = build_demo()
+        demo.queue()
+        demo.launch(share=True, server_name="0.0.0.0", server_port=ports)
+
+    # 启动gradio
     create_gradio(port)
-- 
Gitee


From 92a719ec317bf677315665fd781f6e717ab82dc7 Mon Sep 17 00:00:00 2001
From: tanfeng <823018000@qq.com>
Date: Mon, 25 Nov 2024 11:52:18 +0800
Subject: [PATCH 3/6] =?UTF-8?q?=E9=80=82=E9=85=8D=E6=9C=80=E6=96=B0POC?=
 =?UTF-8?q?=E7=89=88=E6=9C=AC=E5=B1=95=E7=A4=BA=E9=A1=B5=E9=9D=A2=EF=BC=8C?=
 =?UTF-8?q?=E5=8A=9F=E8=83=BD=E5=A2=9E=E5=8A=A0?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 mxRAG/PocValidation/chat_with_ascend/app.py | 68 ++++++++++-----------
 1 file changed, 34 insertions(+), 34 deletions(-)

diff --git a/mxRAG/PocValidation/chat_with_ascend/app.py b/mxRAG/PocValidation/chat_with_ascend/app.py
index 5d6dea138..4eb7c03cc 100644
--- a/mxRAG/PocValidation/chat_with_ascend/app.py
+++ b/mxRAG/PocValidation/chat_with_ascend/app.py
@@ -24,7 +24,7 @@ from mx_rag.document.loader import DocxLoader, PdfLoader, ExcelLoader
 from mx_rag.knowledge import KnowledgeStore, KnowledgeDB, KnowledgeMgrStore, KnowledgeMgr
 from mx_rag.knowledge.handler import upload_files, LoaderMng
 
-#初始化数据库管理
+# 初始化数据库管理
 knowledge_mgr = KnowledgeMgr(KnowledgeMgrStore("./test_sql.db"))
 
 
@@ -125,7 +125,7 @@ def file_upload(files,
         knowledge_db_cls = creat_knowledge_db(knowledge_db_name, is_regist=False)
     else:
         knowledge_db_cls = creat_knowledge_db(knowledge_db_name)
-    # 注册文档加载器
+    # 注册文档处理器
     loader_mng = LoaderMng()
     # 加载文档加载器，可以使用mxrag自有的，也可以使用langchain的
     loader_mng.register_loader(loader_class=TextLoader, file_types=[".txt", ".md"])
@@ -207,8 +207,8 @@ def bot_response(history,
                  top_k: int = 1,
                  chat_type: str = "RAG检索增强对话",
                  show_type: str = "不显示",
-                 is_rewrite:str = "否",
-                 knowledge_name: str = 'test',
+                 is_rewrite: str = "否",
+                 knowledge_name: str = 'test'
                  ):
     chat_type_mapping = {"RAG检索增强对话":1,
                          "仅大模型对话":0}
@@ -218,15 +218,15 @@ def bot_response(history,
     is_rewrite_mapping = {"是":1,
                           "否":0}
     # 初始化检索器
-    retrieve_cls = creat_retriever(knowledge_name, top_k, score_threshold)
+    retriever_cls = creat_retriever(knowledge_name, top_k, score_threshold)
     # 初始化chain
-    text2text_chain = SingleText2TextChain(llm=llm, retriever=retrieve_cls, reranker=reranker)
+    text2text_chain = SingleText2TextChain(llm=llm, retriever=retriever_cls, reranker=reranker)
     # 历史问题改写
     if is_rewrite_mapping.get(is_rewrite) == 1:
         history = generate_question(history, llm, history_n)
     history[-1][1] = '推理错误'
     try:
-        # 仅使用大模型
+        # 仅使用大模型回答
         if chat_type_mapping.get(chat_type) == 0:
             response = llm.chat_streamly(query=history[-1][0],
                                          llm_config=LLMParameterConfig(max_tokens=max_tokens,
@@ -257,10 +257,10 @@ def bot_response(history,
                 # 有检索到信息
                 if len(q_docs) > 0:
                     history_last = ''
-                    for i ,source in enumerate(q_docs):
-                        sources = "\n====检索信息来源:" + str(i + 1) + ":" + "[数据库名]:" + str(knowledge_name) + \
+                    for i, source in enumerate(q_docs):
+                        sources = "\n====检索信息来源:" + str(i + 1) + "[数据库名]:" + str(knowledge_name) + \
                                   "[文件名]:" + source['metadata']['source'] + "====" + "\n" + \
-                                  "参考内容：" +source['page_content'] + "\n"
+                                  "参考内容：" + source['page_content'] + "\n"
                         history_last += sources
                     history += [[None, history_last]]
                 yield history, history_r
@@ -274,9 +274,9 @@ def bot_response(history,
                 if len(q_docs) > 0:
                     history_r_last = ''
                     for i ,source in enumerate(q_docs):
-                        sources = "\n====检索信息来源:" + str(i + 1) + ":" + "[数据库名]:" + str(knowledge_name) + \
+                        sources = "\n====检索信息来源:" + str(i + 1) + "[数据库名]:" + str(knowledge_name) + \
                                   "[文件名]:" + source['metadata']['source'] + "====" + "\n" + \
-                                  "参考内容：" +source['page_content'] + "\n"
+                                  "参考内容：" + source['page_content'] + "\n"
                         history_r_last += sources
                     history_r += [[history[-1][0], history_r_last]]
                 yield history, history_r
@@ -292,12 +292,12 @@ def re_response(history_r,
                 knowledge_name: str = 'test'
                 ):
     # 初始化检索器
-    retrieve_cls = creat_retriever(knowledge_name, top_k, score_threshold)
-    q_docs = retrieve_cls.invoke(history_r[-1][0])
+    retriever_cls = creat_retriever(knowledge_name, top_k, score_threshold)
+    q_docs = retriever_cls.invoke(history_r[-1][0])
     if len(q_docs) > 0:
         history_r_last = ''
         for i, source in enumerate(q_docs):
-            sources = "\n====检索信息来源:" + str(i + 1) + "[数据库名]:" + str(knowledge_name)  +\
+            sources = "\n====检索信息来源:" + str(i + 1) + "[数据库名]:" + str(knowledge_name) + \
                       "[文件名]:" + source.metadata['source'] + "====" + "\n" + \
                       "参考内容：" + source.page_content + "\n"
             history_r_last += sources
@@ -309,7 +309,7 @@ def re_response(history_r,
 
 # 检索信息
 def user_retriever(user_message, history_r):
-    return "",history_r + [[user_message, None]]
+    return "", history_r + [[user_message, None]]
 
 
 # 聊天信息
@@ -366,7 +366,7 @@ if __name__ == '__main__':
         text_emb = TEIEmbedding(url=embedding_url, client_param=ClientParam(use_http=True))
     else:
         text_emb = TextEmbedding(model_path=embedding_path, dev_id=dev)
-    # 配置rerank,请根据模型具体路径适配
+    # 配置reranker,请根据模型具体路径适配
     if tei_reranker:
         reranker = TEIReranker(url=reranker_url, client_param=ClientParam(use_http=True))
     elif reranker_path is not None:
@@ -381,10 +381,10 @@ if __name__ == '__main__':
             with gr.Row():
                 with gr.Column(scale=100):
                     with gr.Row():
-                        model_select = gr.Dropdown(choices=["Default"], value="Default", container=False, interactive=True)
+                        model_select = gr.Dropdown(choices=[model_name], value="Llama3-8B-Chinese-Chat", container=False, interactive=True)
                     with gr.Row():
-                        files = gr.Files(
-                            height=300,
+                        files = gr.components.File(
+                            height=100,
                             file_count="multiple",
                             file_types=[".docx", ".txt", ".md", ".pdf", ".xlsx", ".xls"],
                             interactive=True,
@@ -393,8 +393,8 @@ if __name__ == '__main__':
                     with gr.Row():
                         upload_btn = gr.Button("上传文件")
                     with gr.Row():
-                        with gr.TabItem(label='知识库情况'):
-                            set_knowledge_name = gr.Textbox(label='设置当前知识库',placeholder='在此输入知识库名称，默认使用test知识库')
+                        with gr.TabItem("知识库情况"):
+                            set_knowledge_name = gr.Textbox(label='设置当前知识库', placeholder="在此输入知识库名称,默认使用test知识库")
                             with gr.Row():
                                 creat_knowledge_btn = gr.Button('创建知识库')
                                 delete_knowledge_btn = gr.Button('删除知识库')
@@ -402,15 +402,15 @@ if __name__ == '__main__':
                             knowledge_number_output = gr.Textbox(label='知识库数量')
                             with gr.Row():
                                 show_knowledge_btn = gr.Button('显示知识库')
-                        with gr.TabItem(label='文件情况'):
-                            knowledge_name = gr.Textbox('知识库名称')
-                            knowledge_file_output = gr.Textbox('知识库文件列表')
-                            knowledge_file_num_output = gr.Textbox('知识库文件数量')
+                        with gr.TabItem("文件情况"):
+                            knowledge_name = gr.Textbox(label='知识库名称')
+                            knowledge_file_output = gr.Textbox(label='知识库文件列表')
+                            knowledge_file_num_output = gr.Textbox(label='知识库文件数量')
                             with gr.Row():
                                 knowledge_file_out_btn = gr.Button('显示文件情况')
                                 knowledge_clear_btn = gr.Button('清空知识库')
                     with gr.Row():
-                        with gr.According(label='文档切分参数设置',open=False):
+                        with gr.Accordion(label='文档切分参数设置', open=False):
                             chunk_size = gr.Slider(
                                 minimum=50,
                                 maximum=5000,
@@ -430,7 +430,7 @@ if __name__ == '__main__':
                                 info="文本切分填充长度"
                             )
                     with gr.Row():
-                        with gr.According(label='大模型参数设置', open=False):
+                        with gr.Accordion(label='大模型参数设置', open=False):
                             temperature = gr.Slider(
                                 minimum=0.01,
                                 maximum=2,
@@ -458,14 +458,14 @@ if __name__ == '__main__':
                                 label="最大tokens",
                                 info="输入+输出最多的tokens数"
                             )
-                            is_rewrite = gr.Radio(['是','否'], value='否', label="是否根据历史提问重写问题？")
+                            is_rewrite = gr.Radio(['是', '否'], value="否", label="是否根据历史提问重写问题？")
                             history_n = gr.Slider(
                                 minimum=1,
                                 maximum=10,
                                 value=5,
                                 step=1,
                                 interactive=True,
-                                label="历史提问重新轮数",
+                                label="历史提问重写轮数",
                                 info="问题重写时所参考的历史提问轮数"
                             )
                     with gr.Row():
@@ -477,7 +477,7 @@ if __name__ == '__main__':
                                 step=0.01,
                                 interactive=True,
                                 label="score_threshold",
-                                info="相似性检索阈值，值越大表示越相关,低于阈值不会被返回。"
+                                info="相似性检索阈值,值越大表示越相关,低于阈值不会被返回。"
                             )
                             top_k = gr.Slider(
                                 minimum=1,
@@ -488,11 +488,11 @@ if __name__ == '__main__':
                                 label="top_k",
                                 info="相似性检索返回条数"
                             )
-                            show_type = gr.Radio(['对话结束后显示','检索框单独显示','不显示'], value="对话结束后显示", label="知识库文档匹配结果展示方式选择")
+                            show_type = gr.Radio(['对话结束后显示', '检索框单独显示', '不显示'], value="对话结束后显示", label="知识库文档匹配结果展示方式选择")
                 with gr.Column(scale=200):
                     with gr.Tabs():
                         with gr.TabItem("对话窗口"):
-                            chat_type = gr.Radio(['RAG检索增强对话','仅大模型对话'], value="RAG检索增强对话", label="请选择对话模式？")
+                            chat_type = gr.Radio(['RAG检索增强对话', '仅大模型对话'], value="RAG检索增强对话", label="请选择对话模式？")
                             chatbot = gr.Chatbot(height=550)
                             with gr.Row():
                                 msg = gr.Textbox(placeholder="在此输入问题...", container=False)
@@ -517,7 +517,7 @@ if __name__ == '__main__':
                 files.change(file_change, [], [])
                 upload_btn.click(file_upload, [files, set_knowledge_name, chunk_size, chunk_overlap], files)
                 # 管理所有知识库
-                creat_knowledge_btn.click(creat_knowledge_db, [set_knowledge_name],[]).then(get_knowledge_db, [],[knowledge_name_output, knowledge_number_output])
+                creat_knowledge_btn.click(creat_knowledge_db, [set_knowledge_name], []).then(get_knowledge_db, [], [knowledge_name_output, knowledge_number_output])
                 show_knowledge_btn.click(get_knowledge_db, [], [knowledge_name_output, knowledge_number_output])
                 delete_knowledge_btn.click(delete_knowledge_db, [set_knowledge_name], [knowledge_name_output, knowledge_number_output])
                 # 管理知识库里文件
-- 
Gitee


From 3b002c8d0dd2a8b452425f7dc60d8b0d170e7b10 Mon Sep 17 00:00:00 2001
From: tanfeng <823018000@qq.com>
Date: Mon, 25 Nov 2024 12:02:25 +0800
Subject: [PATCH 4/6] =?UTF-8?q?=E9=80=82=E9=85=8D=E6=9C=80=E6=96=B0POC?=
 =?UTF-8?q?=E7=89=88=E6=9C=AC=E5=B1=95=E7=A4=BA=E9=A1=B5=E9=9D=A2=EF=BC=8C?=
 =?UTF-8?q?=E5=8A=9F=E8=83=BD=E5=A2=9E=E5=8A=A0?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 mxRAG/PocValidation/chat_with_ascend/app.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mxRAG/PocValidation/chat_with_ascend/app.py b/mxRAG/PocValidation/chat_with_ascend/app.py
index 4eb7c03cc..53414ec6c 100644
--- a/mxRAG/PocValidation/chat_with_ascend/app.py
+++ b/mxRAG/PocValidation/chat_with_ascend/app.py
@@ -469,7 +469,7 @@ if __name__ == '__main__':
                                 info="问题重写时所参考的历史提问轮数"
                             )
                     with gr.Row():
-                        with gr.According(label='检索参数设置', open=False):
+                        with gr.Accordion(label='检索参数设置', open=False):
                             score_threshold = gr.Slider(
                                 minimum=0,
                                 maximum=1,
-- 
Gitee


From d63937f97b13c0bc7f9f88b02afe25b9dddafcad Mon Sep 17 00:00:00 2001
From: tanfeng <823018000@qq.com>
Date: Mon, 25 Nov 2024 12:09:27 +0800
Subject: [PATCH 5/6] =?UTF-8?q?=E9=80=82=E9=85=8D=E6=9C=80=E6=96=B0POC?=
 =?UTF-8?q?=E7=89=88=E6=9C=AC=E5=B1=95=E7=A4=BA=E9=A1=B5=E9=9D=A2=EF=BC=8C?=
 =?UTF-8?q?=E5=8A=9F=E8=83=BD=E5=A2=9E=E5=8A=A0?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 mxRAG/PocValidation/chat_with_ascend/app.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/mxRAG/PocValidation/chat_with_ascend/app.py b/mxRAG/PocValidation/chat_with_ascend/app.py
index 53414ec6c..cd4d6c720 100644
--- a/mxRAG/PocValidation/chat_with_ascend/app.py
+++ b/mxRAG/PocValidation/chat_with_ascend/app.py
@@ -218,6 +218,7 @@ def bot_response(history,
     is_rewrite_mapping = {"是":1,
                           "否":0}
     # 初始化检索器
+    knowledge_name = get_knowledge_rename(knowledge_name)
     retriever_cls = creat_retriever(knowledge_name, top_k, score_threshold)
     # 初始化chain
     text2text_chain = SingleText2TextChain(llm=llm, retriever=retriever_cls, reranker=reranker)
-- 
Gitee


From 3f99338c0c819b0154429f8d2cdc22dacba7ddd6 Mon Sep 17 00:00:00 2001
From: tanfeng <823018000@qq.com>
Date: Mon, 25 Nov 2024 12:42:28 +0800
Subject: [PATCH 6/6] =?UTF-8?q?=E9=80=82=E9=85=8D=E6=9C=80=E6=96=B0POC?=
 =?UTF-8?q?=E7=89=88=E6=9C=AC=E5=B1=95=E7=A4=BA=E9=A1=B5=E9=9D=A2=EF=BC=8C?=
 =?UTF-8?q?=E5=8A=9F=E8=83=BD=E5=A2=9E=E5=8A=A0?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 mxRAG/PocValidation/chat_with_ascend/app.py | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/mxRAG/PocValidation/chat_with_ascend/app.py b/mxRAG/PocValidation/chat_with_ascend/app.py
index cd4d6c720..1d23eb9a3 100644
--- a/mxRAG/PocValidation/chat_with_ascend/app.py
+++ b/mxRAG/PocValidation/chat_with_ascend/app.py
@@ -274,7 +274,7 @@ def bot_response(history,
                 # 有检索到信息
                 if len(q_docs) > 0:
                     history_r_last = ''
-                    for i ,source in enumerate(q_docs):
+                    for i, source in enumerate(q_docs):
                         sources = "\n====检索信息来源:" + str(i + 1) + "[数据库名]:" + str(knowledge_name) + \
                                   "[文件名]:" + source['metadata']['source'] + "====" + "\n" + \
                                   "参考内容：" + source['page_content'] + "\n"
@@ -293,6 +293,7 @@ def re_response(history_r,
                 knowledge_name: str = 'test'
                 ):
     # 初始化检索器
+    knowledge_name = get_knowledge_rename(knowledge_name)
     retriever_cls = creat_retriever(knowledge_name, top_k, score_threshold)
     q_docs = retriever_cls.invoke(history_r[-1][0])
     if len(q_docs) > 0:
-- 
Gitee