From d6434aa5be9b2770789861bb32f817b8786bca20 Mon Sep 17 00:00:00 2001
From: HuangHai <10402852@qq.com>
Date: Mon, 24 Mar 2025 10:42:57 +0800
Subject: [PATCH] 'commit'

---
 AI/WxMini/Milvus/Config/MulvusConfig.py       |  27 ++++++++++++------
 .../__pycache__/MulvusConfig.cpython-310.pyc  | Bin 269 -> 434 bytes
 AI/WxMini/Milvus/T1_create_collection.py      |   2 +-
 AI/WxMini/Milvus/T3_insert_data.py            |   8 ++++--
 AI/WxMini/Milvus/T5_search_near_word.py       |  13 +++++----
 5 files changed, 32 insertions(+), 18 deletions(-)

diff --git a/AI/WxMini/Milvus/Config/MulvusConfig.py b/AI/WxMini/Milvus/Config/MulvusConfig.py
index ef42f924..d5443b7d 100644
--- a/AI/WxMini/Milvus/Config/MulvusConfig.py
+++ b/AI/WxMini/Milvus/Config/MulvusConfig.py
@@ -1,12 +1,23 @@
+# Milvus 服务器的主机地址
 MS_HOST = "10.10.14.101"
+
+# Milvus 服务器的端口号
 MS_PORT = "19530"
+
+# Milvus 集合的名称
 MS_COLLECTION_NAME = "ds_collection"
+
+# Milvus 连接池的最大连接数
 MS_MAX_CONNECTIONS = 5
-# "dimension": 128
-# "index_file_size": 1024
-# "metric_type": "L2"
-# "nlist": 1024
-# "nprobe": 16
-# "topk": 10
-# "batch_size": 100
-# "index_type": "IVF_FLAT"
+
+# 腾讯 AI Lab 中文词向量模型的路径
+MS_MODEL_PATH = "D:/Tencent_AILab_ChineseEmbedding/Tencent_AILab_ChineseEmbedding.txt"
+
+# 加载词向量模型时限制的词汇数量
+MS_MODEL_LIMIT = 10000
+
+# 词向量的维度（腾讯 AI Lab 中文词向量模型的维度为 200）
+MS_DIMENSION = 200
+
+# Milvus 搜索时的 nprobe 参数，用于控制搜索的精度和性能
+MS_NPROBE = 100
\ No newline at end of file
diff --git a/AI/WxMini/Milvus/Config/__pycache__/MulvusConfig.cpython-310.pyc b/AI/WxMini/Milvus/Config/__pycache__/MulvusConfig.cpython-310.pyc
index 65b8dae5ed9deab6b0b626ab2bae7d4f4565a213..40071efdce3e4e86dfe5d386bfb261cd6a558607 100644
GIT binary patch
delta 289
zcmeBW+QiJ4&&$ij00eW7JxK3nn#d=kr2^!oFhnt=Fh((^Fhwz?Fh?<`utc$>utu?_
zutl+@ut%|{a0D}Ga!z#7tgmu$vC<Do%}Y+rD~WgX^hr#Lch1PnOD#@y%}q*8Ny*Gh
zCm^j?Qc;pApw7VX@&wT4mnlHPPm|*oyKiv3hktMgm=WL~6mm-lDB|qz<KyZa;_2@f
z@8{_2dTZizVaF(5poFi#i>pt3fMbY96d#=H<LT=ea*GGZck%Rf^$P}Si{b?G`~rgf
com^Kk6oI^6#5HlVpNIgE#lpzKjDQ@>0Gahr8UO$Q

delta 105
zcmdnQ+{?t5&&$ij00gTRK1dg2oX965B?9E7Fhnt=Fh((^Fhwz?Fh?<`umm$`vQ7-q
z6!X($xy9}q9Pi;D90Fzp_y>hVO?;uwxssuX4XCw<ZE^sk9}fqR$-=<G$im3M2msk<
B7CQg{

diff --git a/AI/WxMini/Milvus/T1_create_collection.py b/AI/WxMini/Milvus/T1_create_collection.py
index 2921fa96..1231995e 100644
--- a/AI/WxMini/Milvus/T1_create_collection.py
+++ b/AI/WxMini/Milvus/T1_create_collection.py
@@ -24,7 +24,7 @@ if utility.has_collection(collection_name):
 fields = [
     FieldSchema(name="id", dtype=DataType.INT64, is_primary=True, auto_id=True),  # 主键字段，自动生成 ID
     FieldSchema(name="text", dtype=DataType.VARCHAR, max_length=500),  # 存储对话文本
-    FieldSchema(name="embedding", dtype=DataType.FLOAT_VECTOR, dim=200)  # 向量字段，维度为 200
+    FieldSchema(name="embedding", dtype=DataType.FLOAT_VECTOR, dim=MS_DIMENSION)  # 向量字段，维度为 200
 ]
 schema_description = "Simple demo collection"
 
diff --git a/AI/WxMini/Milvus/T3_insert_data.py b/AI/WxMini/Milvus/T3_insert_data.py
index 7a5f7a77..fc431dcf 100644
--- a/AI/WxMini/Milvus/T3_insert_data.py
+++ b/AI/WxMini/Milvus/T3_insert_data.py
@@ -5,10 +5,11 @@ from gensim.models import KeyedVectors
 import jieba
 
 # 加载预训练的 Word2Vec 模型
-model_path = "D:/Tencent_AILab_ChineseEmbedding/Tencent_AILab_ChineseEmbedding.txt"  # 替换为你的 Word2Vec 模型路径
-model = KeyedVectors.load_word2vec_format(model_path, binary=False, limit=10000)
+model_path = MS_MODEL_PATH  # 替换为你的 Word2Vec 模型路径
+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}")
 
+
 # 将文本转换为嵌入向量
 def text_to_embedding(text):
     words = jieba.lcut(text)  # 使用 jieba 分词
@@ -23,6 +24,7 @@ def text_to_embedding(text):
         print("未找到有效词，返回零向量")
         return [0.0] * model.vector_size
 
+
 # 1. 使用连接池管理 Milvus 连接
 milvus_pool = MilvusConnectionPool(host=MS_HOST, port=MS_PORT, max_connections=MS_MAX_CONNECTIONS)
 
@@ -55,4 +57,4 @@ print("数据插入成功。")
 milvus_pool.release_connection(connection)
 
 # 7. 关闭连接池
-milvus_pool.close()
\ No newline at end of file
+milvus_pool.close()
diff --git a/AI/WxMini/Milvus/T5_search_near_word.py b/AI/WxMini/Milvus/T5_search_near_word.py
index d25057ae..64c82593 100644
--- a/AI/WxMini/Milvus/T5_search_near_word.py
+++ b/AI/WxMini/Milvus/T5_search_near_word.py
@@ -1,4 +1,3 @@
-
 import time
 import jieba  # 导入 jieba 分词库
 from WxMini.Milvus.Utils.MilvusCollectionManager import MilvusCollectionManager
@@ -7,10 +6,11 @@ from WxMini.Milvus.Config.MulvusConfig import *
 from gensim.models import KeyedVectors
 
 # 加载预训练的 Word2Vec 模型
-model_path = "D:/Tencent_AILab_ChineseEmbedding/Tencent_AILab_ChineseEmbedding.txt"  # 替换为你的 Word2Vec 模型路径
-model = KeyedVectors.load_word2vec_format(model_path, binary=False, limit=10000)
+model_path = MS_MODEL_PATH  # 替换为你的 Word2Vec 模型路径
+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}")
 
+
 # 将文本转换为嵌入向量
 def text_to_embedding(text):
     words = jieba.lcut(text)  # 使用 jieba 分词
@@ -25,6 +25,7 @@ def text_to_embedding(text):
         print("未找到有效词，返回零向量")
         return [0.0] * model.vector_size
 
+
 # 1. 使用连接池管理 Milvus 连接
 milvus_pool = MilvusConnectionPool(host=MS_HOST, port=MS_PORT, max_connections=MS_MAX_CONNECTIONS)
 
@@ -46,8 +47,8 @@ current_embedding = text_to_embedding(input_text)
 
 # 7. 查询与当前对话最相关的历史对话
 search_params = {
-    "metric_type": "L2",       # 使用 L2 距离度量方式
-    "params": {"nprobe": 100}  # 设置 IVF_FLAT 的 nprobe 参数
+    "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(current_embedding, search_params, limit=2)  # 返回 2 条结果
@@ -73,4 +74,4 @@ print(f"查询耗时: {end_time - start_time:.4f} 秒")
 milvus_pool.release_connection(connection)
 
 # 11. 关闭连接池
-milvus_pool.close()
\ No newline at end of file
+milvus_pool.close()