update PDF data source

Author: davidhou17

ai-integrations/langchain.ipynb

@@ -53,7 +53,7 @@
     "from langchain.text_splitter import RecursiveCharacterTextSplitter\n",
     "\n",
     "# Load the PDF\n",
-    "loader = PyPDFLoader(\"https://query.prod.cms.rt.microsoft.com/cms/api/am/binary/RE4HkJP\")\n",
+    "loader = PyPDFLoader(\"https://investors.mongodb.com/node/13176/pdf\")\n",
     "data = loader.load()\n",
     "\n",
     "# Split PDF into documents\n",
@@ -77,7 +77,7 @@
     "vector_store = MongoDBAtlasVectorSearch.from_connection_string(\n",
     "  connection_string = ATLAS_CONNECTION_STRING,\n",
     "  namespace = \"langchain_db.test\",\n",
-    "  embedding =  OpenAIEmbeddings(disallowed_special=()),\n",
+    "  embedding =  OpenAIEmbeddings(model=\"text-embedding-3-large\"),\n",
     "  index_name = \"vector_index\"\n",
     ")\n",
     "\n",
@@ -91,11 +91,16 @@
    "metadata": {},
    "outputs": [],
    "source": [
+    "import time\n",
+    "\n",
     "# Use helper method to create the vector search index\n",
     "vector_store.create_vector_search_index(\n",
-    "   dimensions = 1536, # The dimensions of the vector embeddings to be indexed\n",
-    "   filters = [ \"page\" ]\n",
-    ")"
+    "   dimensions = 3072, # The dimensions of the vector embeddings to be indexed\n",
+    "   filters = [ \"page_label\" ]\n",
+    ")\n",
+    "\n",
+    "# Wait for the index to build (this can take around a minute)\n",
+    "time.sleep(60)"
    ]
   },
   {
@@ -113,7 +118,7 @@
    "source": [
     "import pprint\n",
     "\n",
-    "query = \"MongoDB Atlas security\"\n",
+    "query = \"MongoDB acquisition\"\n",
     "results = vector_store.similarity_search(query)\n",
     "\n",
     "pprint.pprint(results)"
@@ -132,7 +137,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "query = \"MongoDB Atlas security\"\n",
+    "query = \"MongoDB acquisition\"\n",
     "results = vector_store.similarity_search_with_score(\n",
     "   query = query, k = 3\n",
     ")\n",
@@ -153,12 +158,12 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "query = \"MongoDB Atlas security\"\n",
+    "query = \"MongoDB acquisition\"\n",
     "\n",
     "results = vector_store.similarity_search_with_score(\n",
     "   query = query,\n",
     "   k = 3,\n",
-    "   pre_filter = { \"page\": { \"$eq\": 17 } }\n",
+    "   pre_filter = { \"page_label\": { \"$eq\": 2 } }\n",
     ")\n",
     "\n",
     "pprint.pprint(results)"
@@ -200,7 +205,7 @@
     "\"\"\"\n",
     "custom_rag_prompt = PromptTemplate.from_template(template)\n",
     "\n",
-    "llm = ChatOpenAI()\n",
+    "llm = ChatOpenAI(model=\"gpt-4o\")\n",
     "\n",
     "def format_docs(docs):\n",
     "   return \"\\n\\n\".join(doc.page_content for doc in docs)\n",
@@ -214,7 +219,7 @@
     ")\n",
     "\n",
     "# Prompt the chain\n",
-    "question = \"How can I secure my MongoDB Atlas cluster?\"\n",
+    "question = \"What was MongoDB's latest acquisition?\"\n",
     "answer = rag_chain.invoke(question)\n",
     "\n",
     "print(\"Question: \" + question)\n",
@@ -245,7 +250,7 @@
     "   search_kwargs = {\n",
     "      \"k\": 10,\n",
     "      \"score_threshold\": 0.75,\n",
-    "      \"pre_filter\": { \"page\": { \"$eq\": 17 } }\n",
+    "      \"pre_filter\": { \"page_label\": { \"$eq\": 2 } }\n",
     "   }\n",
     ")\n",
     "\n",
@@ -261,7 +266,7 @@
     "\"\"\"\n",
     "custom_rag_prompt = PromptTemplate.from_template(template)\n",
     "\n",
-    "llm = ChatOpenAI()\n",
+    "llm = ChatOpenAI(model=\"gpt-4o\")\n",
     "\n",
     "def format_docs(docs):\n",
     "   return \"\\n\\n\".join(doc.page_content for doc in docs)\n",
@@ -275,7 +280,7 @@
     ")\n",
     "\n",
     "# Prompt the chain\n",
-    "question = \"How can I secure my MongoDB Atlas cluster?\"\n",
+    "question = \"What was MongoDB's latest acquisition?\"\n",
     "answer = rag_chain.invoke(question)\n",
     "\n",
     "print(\"Question: \" + question)\n",