From 703e9b1ddbf5b65520fe9d43d9cfa07ba382ac4b Mon Sep 17 00:00:00 2001 From: HuangHai <10402852@qq.com> Date: Mon, 24 Mar 2025 13:24:32 +0800 Subject: [PATCH] 'commit' --- AI/WxMini/Milvus/Config/MulvusConfig.py | 2 +- .../__pycache__/MulvusConfig.cpython-310.pyc | Bin 434 -> 434 bytes AI/WxMini/Milvus/T5_search_near_data.py | 16 +- .../Milvus/Utils/MilvusCollectionManager.py | 34 +-- .../MilvusCollectionManager.cpython-310.pyc | Bin 3879 -> 4570 bytes AI/WxMini/Start.py | 13 +- AI/WxMini/X_Start.py | 229 ++++++++++++++++++ 7 files changed, 267 insertions(+), 27 deletions(-) create mode 100644 AI/WxMini/X_Start.py diff --git a/AI/WxMini/Milvus/Config/MulvusConfig.py b/AI/WxMini/Milvus/Config/MulvusConfig.py index d5443b7d..64ae6614 100644 --- a/AI/WxMini/Milvus/Config/MulvusConfig.py +++ b/AI/WxMini/Milvus/Config/MulvusConfig.py @@ -1,5 +1,5 @@ # Milvus 服务器的主机地址 -MS_HOST = "10.10.14.101" +MS_HOST = "10.10.14.205" # Milvus 服务器的端口号 MS_PORT = "19530" diff --git a/AI/WxMini/Milvus/Config/__pycache__/MulvusConfig.cpython-310.pyc b/AI/WxMini/Milvus/Config/__pycache__/MulvusConfig.cpython-310.pyc index 40071efdce3e4e86dfe5d386bfb261cd6a558607..058f3a0886feb1833174f1c373769be1760f1c91 100644 GIT binary patch delta 26 gcmdnQyos4BpO=@50SI2bn#dK(Y-C`%F)4r%09dpJ$^ZZW delta 26 gcmdnQyos4BpO=@50SM+Eo5&T)Y-nJ(F)4r%08;=4Q~&?~ diff --git a/AI/WxMini/Milvus/T5_search_near_data.py b/AI/WxMini/Milvus/T5_search_near_data.py index e6a1b127..ac32201c 100644 --- a/AI/WxMini/Milvus/T5_search_near_data.py +++ b/AI/WxMini/Milvus/T5_search_near_data.py @@ -1,5 +1,4 @@ import time -import jieba # 导入 jieba 分词库 from WxMini.Milvus.Utils.MilvusCollectionManager import MilvusCollectionManager from WxMini.Milvus.Utils.MilvusConnectionPool import * from WxMini.Milvus.Config.MulvusConfig import * @@ -13,14 +12,11 @@ print(f"模型加载成功，词向量维度: {model.vector_size}") # 将文本转换为嵌入向量 def text_to_embedding(text): - words = jieba.lcut(text) # 使用 jieba 分词 - print(f"文本: {text}, 分词结果: {words}") - embeddings = [model[word] for word in words if word in model] - print(f"有效词向量数量: {len(embeddings)}") - if embeddings: - avg_embedding = sum(embeddings) / len(embeddings) - print(f"生成的平均向量: {avg_embedding[:5]}...") # 打印前 5 维 - return avg_embedding + # 直接使用全句进行向量计算 + if text in model: + embedding = model[text] + print(f"生成的全句向量: {embedding[:5]}...") # 打印前 5 维 + return embedding else: print("未找到有效词，返回零向量") return [0.0] * model.vector_size @@ -74,4 +70,4 @@ print(f"查询耗时: {end_time - start_time:.4f} 秒") milvus_pool.release_connection(connection) # 12. 关闭连接池 -milvus_pool.close() +milvus_pool.close() \ No newline at end of file diff --git a/AI/WxMini/Milvus/Utils/MilvusCollectionManager.py b/AI/WxMini/Milvus/Utils/MilvusCollectionManager.py index eefad47e..c68ef130 100644 --- a/AI/WxMini/Milvus/Utils/MilvusCollectionManager.py +++ b/AI/WxMini/Milvus/Utils/MilvusCollectionManager.py @@ -81,22 +81,28 @@ class MilvusCollectionManager: except Exception as e: print(f"查询失败: {e}") return None - def search(self, query_embedding, search_params, limit=2): + def search(self, data, search_params, expr=None, limit=5): """ - 查询数据 - :param query_embedding: 查询向量 - :param search_params: 查询参数 - :param limit: 返回结果数量 - :return: 查询结果 + 在集合中搜索与输入向量最相似的数据 + :param data: 输入向量 + :param search_params: 搜索参数 + :param expr: 过滤条件（可选） + :param limit: 返回结果的数量 + :return: 搜索结果 """ - if self.collection is None: - raise Exception("集合未加载，请检查集合是否存在。") - return self.collection.search( - data=[query_embedding], - anns_field="embedding", - param=search_params, - limit=limit - ) + try: + # 构建搜索参数 + search_result = self.collection.search( + data=[data], # 输入向量 + anns_field="embedding", # 向量字段名称 + param=search_params, # 搜索参数 + limit=limit, # 返回结果的数量 + expr=expr # 过滤条件 + ) + return search_result + except Exception as e: + print(f"搜索失败: {e}") + return None def query_text_by_id(self, id): """ diff --git a/AI/WxMini/Milvus/Utils/__pycache__/MilvusCollectionManager.cpython-310.pyc b/AI/WxMini/Milvus/Utils/__pycache__/MilvusCollectionManager.cpython-310.pyc index a88defeefbcc6cb8afbb98a60fbd1ffb43becc00..941f0ea4313b339351180d32f660b275d4733c6b 100644 GIT binary patch delta 1332 zcmZ8hTWs4@81^}_PaG%lv}@NDhD*69fEWdVM9YMLiOWk~iiVhkqO8Z1jON0zLuutK z1zNUAl%}#KB-*Y}tpbGB4RxTR77(vI^JEFZL+v)a^o1uTCjQ^f($Zu3^Y8z5exL7? zzdY{s=;d@;lVJV*+n+;I2WIpJIX$!ROm~{=6VjOkKg{>@0~6$o#t-sC;7Rg*15)OQ z*WvcYJ64k9R*ZOVTzP_cEB(qZ1YZ7Scatu+E51!S-K2UyIp+4NO|ZKX4}<+uqF5iK zxM6F_@WcULfZO5#V;zokPC7?8amcJRPlqH<7YUC|(HeEAIP6d^cS^5G18{R?K$0b! zj`k9%79-M_ZpS!tVp#VfA&~JpNveU)&oj8>gs7$4@VvuQ&DqtB^>v^vxO@Zj(XMUDfamQz2qCO-pVuCEI&9-le*o!w zbCDy+{rN($b~Tv)Vq@+5?e&ZP^!4WWrS0`guBG*yiuM7~?bP0pQ#9b_nj2i1Ye&6= zSv+fUUMLNLd}N%)bAQ)5>*66e#EAQ7q89gA;5g)-S!P}gzO{?|ctrl=Q0S$jXfD*c z6*d!3?hZua5RZayk2=It;Cm1`wC)0+2}Wf?TNk^<;=6=#9EGt`OELm~hl!L)-gW;< zP4|l)Y@0zq4QE(5ay058^d3QA2#;V~*^#aLYr3Dj?K1sEEkm>7F$|~N)B1^fj$(Qj uf`M=^!U+VNS3K)3=-u7eT$NO!CR9yLsY!Syw1cS*fm+SN3A=yjFaHOH`Go%f delta 656 zcmZ8eO=}ZT6wRHlzL|VZQ#*-Z6HOHd7hSlrT112*E@IW)fP}GcOh=k&`Z6Q61QaQ_ zl2{)u1hG&M7YY`$6b1jrtlSvQ${(N$@0&Ccy^lF_xF6@8xf9pk%;h$+StEsU_3hW@ zi`;JRDb#oOE|iNjNAo-2=(Ir1os?tH>6?Vzpt)E*tnFt`rIW62 zy+?!R2OP4s2lyU7?oCt{nHNSZ7>|7a$Oo@pPO5cg;gya)k+zjsp>F6ph8qN(#vR{V zr?HxRe5|hfEkCqnHia{+gb-^T&tyAg%$C>%NN zFhDI^`|Fg=V1gYHpdrL+vQk@3e8-Yc`Mtt 1: # 过滤过短的关键词 + # 使用前缀匹配 + keyword_conditions.append(f"user_input like '{keyword}%'") + if keyword_conditions: + expr += " and (" + " or ".join(keyword_conditions) + ")" + return expr + +# 初始化 Word2Vec 模型 +model_path = MS_MODEL_PATH +model = KeyedVectors.load_word2vec_format(model_path, binary=False, limit=MS_MODEL_LIMIT) +print(f"模型加载成功，词向量维度: {model.vector_size}") + +# 初始化 Milvus 连接池 +milvus_pool = MilvusConnectionPool(host=MS_HOST, port=MS_PORT, max_connections=MS_MAX_CONNECTIONS) + +# 初始化集合管理器 +collection_name = MS_COLLECTION_NAME +collection_manager = MilvusCollectionManager(collection_name) + + +# 将文本转换为嵌入向量 +def text_to_embedding(text): + words = jieba.lcut(text) # 使用 jieba 分词 + print(f"文本: {text}, 分词结果: {words}") + embeddings = [model[word] for word in words if word in model] + print(f"有效词向量数量: {len(embeddings)}") + if embeddings: + avg_embedding = sum(embeddings) / len(embeddings) + print(f"生成的平均向量: {avg_embedding[:5]}...") # 打印前 5 维 + return avg_embedding + else: + print("未找到有效词，返回零向量") + return [0.0] * model.vector_size + + +# 使用 Lifespan Events 处理应用启动和关闭逻辑 +@asynccontextmanager +async def lifespan(app: FastAPI): + # 应用启动时加载集合到内存 + collection_manager.load_collection() + print(f"集合 '{collection_name}' 已加载到内存。") + yield + # 应用关闭时释放连接池 + milvus_pool.close() + print("Milvus 连接池已关闭。") + + +# 初始化 FastAPI 应用 +app = FastAPI(lifespan=lifespan) + +# 初始化 OpenAI 客户端 +client = OpenAI( + api_key=MODEL_API_KEY, + base_url="https://dashscope.aliyuncs.com/compatible-mode/v1", +) + +# 设置相似度阈值 +SIMILARITY_THRESHOLD = 0.5 # 距离小于 0.5 的结果被认为是高相似度 + +# 在 /reply 接口中优化查询逻辑和提示词构建方式 +# 在 /reply 接口中优化提示词传递和系统提示词 +# 在 /reply 接口中优化提示词传递和系统提示词 +# 在 /reply 接口中优化查询逻辑和提示词构建方式 +@app.post("/reply") +async def reply(session_id: str = Form(...), prompt: str = Form(...)): + """ + 接收用户输入的 prompt，调用大模型并返回结果 + :param session_id: 用户会话 ID + :param prompt: 用户输入的 prompt + :return: 大模型的回复 + """ + try: + # 从连接池中获取一个连接 + connection = milvus_pool.get_connection() + + # 将用户输入转换为嵌入向量 + current_embedding = text_to_embedding(prompt) + + # 提取用户输入的关键词 + keywords = extract_keywords(prompt) + print(f"提取的关键词: {keywords}") + + # 构建查询条件 + expr = f"session_id == '{session_id}'" # 只过滤 session_id + print(f"查询条件: {expr}") + + # 查询与当前对话最相关的五条历史交互 + search_params = { + "metric_type": "L2", # 使用 L2 距离度量方式 + "params": {"nprobe": MS_NPROBE} # 设置 IVF_FLAT 的 nprobe 参数 + } + start_time = time.time() + results = collection_manager.search( + data=current_embedding, # 输入向量 + search_params=search_params, # 搜索参数 + expr=expr, # 查询条件 + limit=5 # 返回 5 条结果 + ) + end_time = time.time() + + # 调试：输出查询结果 + print(f"查询结果: {results}") + + # 构建历史交互提示词 + history_prompt = "" + if results: + for hits in results: + for hit in hits: + try: + # 过滤低相似度的结果 + if hit.distance > SIMILARITY_THRESHOLD: + print(f"跳过低相似度结果，距离: {hit.distance}") + continue + + # 查询非向量字段 + record = collection_manager.query_by_id(hit.id) + if record: + print(f"查询到的记录: {record}") + # 只添加与当前问题高度相关的历史交互 + if any(keyword in record['user_input'] for keyword in keywords): + history_prompt += f"用户: {record['user_input']}\n大模型: {record['model_response']}\n" + except Exception as e: + print(f"查询失败: {e}") + + print(f"历史交互提示词: {history_prompt}") + + # 调用大模型，将历史交互作为提示词 + response = client.chat.completions.create( + model=MODEL_NAME, + messages=[ + {"role": "system", "content": "你是一个私人助理，负责回答用户的问题。请根据用户的历史对话和当前问题，提供准确且简洁的回答。不要提及你是通义千问或其他无关信息。"}, + {"role": "user", "content": f"{history_prompt}用户: {prompt}"} # 将历史交互和当前输入一起发送 + ], + max_tokens=500 + ) + + # 提取生成的回复 + if response.choices and response.choices[0].message.content: + result = response.choices[0].message.content.strip() + + # 记录用户输入和大模型反馈到向量数据库 + timestamp = time.strftime("%Y-%m-%d %H:%M:%S", time.localtime()) + entities = [ + [session_id], # session_id + [prompt], # user_input + [result], # model_response + [timestamp], # timestamp + [current_embedding] # embedding + ] + collection_manager.insert_data(entities) + print("用户输入和大模型反馈已记录到向量数据库。") + + # 调用tts进行生成mp3 + uuid_str = str(uuid.uuid4()) + tts_file = "audio/" + uuid_str + ".mp3" + t = TTS(tts_file) + t.start(result) + # 文件上传到oss + upload_mp3_to_oss(tts_file, tts_file) + # 删除临时文件 + try: + os.remove(tts_file) + print(f"临时文件 {tts_file} 已删除") + except Exception as e: + print(f"删除临时文件失败: {e}") + # 完整的url + url = 'https://ylt.oss-cn-hangzhou.aliyuncs.com/' + tts_file + return { + "success": True, + "url": url, + "search_time": end_time - start_time, # 返回查询耗时 + "response": result # 返回大模型的回复 + } + else: + raise HTTPException(status_code=500, detail="大模型未返回有效结果") + except Exception as e: + raise HTTPException(status_code=500, detail=f"调用大模型失败: {str(e)}") + finally: + # 释放连接 + milvus_pool.release_connection(connection) + + +# 运行 FastAPI 应用 +if __name__ == "__main__": + import uvicorn + + uvicorn.run(app, host="0.0.0.0", port=5600)