Merge pull request #614 from Bennu-Li/update_qa

update qa-chatbot to Milvus2.0.0-rc5

solutions/question_answering_system/question_answering.ipynb

@@ -5,7 +5,9 @@
    "metadata": {},
    "source": [
     "# Question Answering System\n",
-    "In this example we will be going over the code used to build a question answering system. This example uses a modified BERT model to extract features from questions and Milvus to search for similar questions and answers. "
+    "In this example we will be going over the code used to build a question answering system. This example uses a modified BERT model to extract features from questions and Milvus to search for similar questions and answers. \n",
+    "\n",
+    "> This project based Milvus 2.0.0-rc5."
    ]
   },
   {
@@ -29,7 +31,7 @@
     "\n",
     "|  Packages                  |  Servers      |\n",
     "|-                           | -             |   \n",
-    "| pymilvus-orm==2.0.0rc1     | milvus 2.0.0  |\n",
+    "| pymilvus==2.0.0rc5     | milvus 2.0.0-rc5  |\n",
     "| sentence_transformers      |               |\n",
     "| psmysql                    | mysql         |\n",
     "| pandas                     |               |\n",
@@ -146,10 +148,7 @@
    "outputs": [],
    "source": [
     "#Connectings to Milvus, BERT and Postgresql\n",
-    "from pymilvus_orm import connections, utility\n",
-    "from pymilvus_orm.types import DataType\n",
-    "from pymilvus_orm.schema import FieldSchema, CollectionSchema\n",
-    "from pymilvus_orm.collection import Collection\n",
+    "from pymilvus import connections, FieldSchema, CollectionSchema, DataType, Collection, utility\n",
     "import pymysql\n",
     "\n",
     "connections.connect(host='localhost', port='19537')\n",
@@ -454,4 +453,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 5
-}
+}

solutions/question_answering_system/quick_deploy/README.md

@@ -3,6 +3,8 @@
 
 This project combines Milvus and BERT to build a question and answer system. This aims to provide a solution to achieve semantic similarity matching with Milvus combined with AI models.
 
+> This project based Milvus2.0.0-rc5
+
 ## Data description
 
 The dataset needed for this system is a CSV format file which needs to contain a column of questions and a column of answers. 
@@ -163,4 +165,4 @@ After starting the service, Please visit `127.0.0.1:8000/docs` in your browser
 
   iii. **Obtain answer**: Click any of the similar questions obtained in the previous step, and you'll get the answer.
 
-
+

solutions/question_answering_system/quick_deploy/server/requirements.txt

@@ -1,7 +1,7 @@
 PyMySQL
 uvicorn
 numpy
-pymilvus-orm==2.0.0rc1
+pymilvus==2.0.0rc5
 fastapi
 python-multipart
 pandas

solutions/question_answering_system/quick_deploy/server/src/milvus_helpers.py

@@ -1,13 +1,11 @@
 import sys
-from pymilvus_orm import connections
-from pymilvus_orm.types import DataType
-from pymilvus_orm.schema import FieldSchema, CollectionSchema
-from pymilvus_orm.collection import Collection
+from pymilvus import connections, FieldSchema, CollectionSchema, DataType, Collection, utility
+
 from config import MILVUS_HOST, MILVUS_PORT, VECTOR_DIMENSION, METRIC_TYPE
-from pymilvus_orm import utility
 from logs import LOGGER
 
 
+
 class MilvusHelper:
     def __init__(self):
         try:
@@ -25,15 +23,15 @@ def set_collection(self, collection_name):
             else:
                 raise Exception("There has no collection named:{}".format(collection_name))
         except Exception as e:
-            LOGGER.error("Failed to load data to Milvus: {}".format(e))
+            LOGGER.error("ERROR: {}".format(e))
             sys.exit(1)
 
     # Return if Milvus has the collection
     def has_collection(self, collection_name):
         try:
             return utility.has_collection(collection_name)
         except Exception as e:
-            LOGGER.error("Failed to load data to Milvus: {}".format(e))
+            LOGGER.error("Failed: {}".format(e))
             sys.exit(1)
 
     # Create milvus collection if not exists
@@ -46,6 +44,8 @@ def create_collection(self, collection_name):
                 schema = CollectionSchema(fields=[field1, field2], description="collection description")
                 self.collection = Collection(name=collection_name, schema=schema)
                 LOGGER.debug("Create Milvus collection: {}".format(self.collection))
+            else:
+                self.collection = Collection(name=collection_name)
             return "OK"
         except Exception as e:
             LOGGER.error("Failed to load data to Milvus: {}".format(e))
@@ -55,7 +55,7 @@ def create_collection(self, collection_name):
     def insert(self, collection_name, vectors):
         try:
             self.create_collection(collection_name)
-            self.collection = Collection(name=collection_name)
+            # self.collection = Collection(name=collection_name)
             data = [vectors]
             mr = self.collection.insert(data)
             ids = mr.primary_keys

solutions/question_answering_system/requirements.txt

@@ -1,5 +1,5 @@
 PyMySQL
 numpy
-pymilvus-orm==2.0.0rc1
+pymilvus==2.0.0rc5
 pandas
 sentence_transformers==1.2.0