Fix openai issues

singlestore-labs · Nov 21, 2023 · 4a67aa6 · 4a67aa6
1 parent 0f4f2ef
commit 4a67aa6
Showing 1 changed file with 17 additions and 12 deletions.
diff --git a/notebooks/semantic-search-with-openai-embedding-creation/notebook.ipynb b/notebooks/semantic-search-with-openai-embedding-creation/notebook.ipynb
@@ -85,15 +85,13 @@
       "metadata": {},
       "outputs": [],
       "source": [
-        "!pip3 install openai matplotlib plotly pandas scipy scikit-learn requests --quiet\n",
+        "!pip3 install openai==1.3.3 requests --quiet\n",
         "\n",
         "import json\n",
         "import os\n",
         "\n",
-        "import openai\n",
-        "import requests\n",
-        "from openai.embeddings_utils import get_embedding\n",
-        "from openai.embeddings_utils import get_embeddings"
+        "from openai import OpenAI\n",
+        "import requests"
       ]
     },
     {
@@ -129,7 +127,9 @@
       "source": [
         "import getpass\n",
         "\n",
-        "openai.api_key = getpass.getpass('OpenAI API Key: ')"
+        "os.environ[\"OPENAI_API_KEY\"] = getpass.getpass('OpenAI API Key: ')\n",
+        "\n",
+        "client = OpenAI()"
       ]
     },
     {
@@ -174,8 +174,7 @@
       "metadata": {},
       "outputs": [],
       "source": [
-        "url = 'https://raw.githubusercontent.com/singlestore-labs/singlestoredb-samples/main/' + \\\n",
-        "      'Tutorials/ai-powered-semantic-search/hr_sample_data.sql'"
+        "url = 'https://raw.githubusercontent.com/singlestore-labs/singlestoredb-samples/main/Tutorials/ai-powered-semantic-search/hr_sample_data.sql'"
       ]
     },
     {
@@ -217,12 +216,18 @@
       "source": [
         "%sql ALTER TABLE reviews ADD embeddings BLOB;\n",
         "\n",
-        "reviews = %sql SELECT review FROM reviews;\n",
+        "from typing import List\n",
         "\n",
+        "reviews = %sql SELECT review FROM reviews;\n",
         "reviews = [x.review for x in reviews]\n",
-        "embeddings = get_embeddings(reviews, engine='text-embedding-ada-002')\n",
         "\n",
-        "for review, embedding in zip(reviews, embeddings):\n",
+        "def get_embeddings(inputs: List[str], model: str = 'text-embedding-ada-002') -> List[str]:\n",
+        "    \"\"\"Return list of embeddings.\"\"\"\n",
+        "    return [x.embedding for x in client.embeddings.create(input=inputs, model=model).data]\n",
+        "\n",
+        "embeddings = get_embeddings(reviews)\n",
+        "\n",
+        "for embedding, review in zip(embeddings, reviews):\n",
         "    %sql UPDATE reviews SET embeddings = JSON_ARRAY_PACK('{{json.dumps(embedding)}}') WHERE review='{{review}}';"
       ]
     },
@@ -245,7 +250,7 @@
       "source": [
         "searchstring = input('Please enter a search string: ')\n",
         "\n",
-        "search_embedding = json.dumps(get_embedding(searchstring, engine='text-embedding-ada-002'))\n",
+        "search_embedding = json.dumps(get_embeddings([searchstring])[0])\n",
         "\n",
         "results = %sql SELECT review, DOT_PRODUCT(embeddings, JSON_ARRAY_PACK('{{search_embedding}}')) AS score FROM reviews ORDER BY score DESC LIMIT 5;\n",
         "\n",