Update app.py

app.py

@@ -47,10 +47,58 @@ for i, row in tqdm(df.iterrows(), total=df.shape[0]):
         index.upsert(prepped)
         prepped = []
 
+#Use ada embedding model to create vector embeddings for input articles. 
+def get_embeddings(articles, model="text-embedding-ada-002"):
+   return openai_client.embeddings.create(input = articles, model=model)
 
-def run_query(query):
-    return
+def create_prompt(query):
+    #Create embedding for input query. 
+    embed = get_embeddings([query])
+    
+    #Search match in Pinecone. 
+    res = index.query(vector=embed.data[0].embedding, top_k=3, include_metadata=True)
+    
+    #Extract Metadata and Text from the matches object returned by Pinecone. 
+    contexts = [
+        x['metadata']['text'] for x in res['matches']
+    ]
+    
+    prompt_start = (
+        "Answer the question based on the context below.\n\n"+
+        "Context:\n"
+    )
+    
+    prompt_end = (
+        f"\n\nQuestion: {query}\nAnswer:"
+    )
+    
+    #Create a Prompt for OpenAI and provide results of pinecone as part of the prompt. 
+    prompt = (
+        prompt_start + "\n\n---\n\n".join(contexts) + 
+        prompt_end
+    )
+    return prompt
+
+def summarize(prompt):
+    #Use completions API with GPT 3.5 Turbo, and prepared prompt to call OpenAI. 
+    res = openai_client.completions.create(
+        model="gpt-3.5-turbo-instruct",
+        prompt=prompt,
+        temperature=0,
+        max_tokens=636,
+        top_p=1,
+        frequency_penalty=0,
+        presence_penalty=0,
+        stop=None
+    )
 
+    return res.choices[0].text
+    
+
+def run_query(query):
+    prompt=create_prompt(query)
+    result=summarize(prompt)
+    return result
 
 # Create the Gradio interface
 demo = gr.Interface(fn=run_query, inputs=gr.Textbox(label="User Input", placeholder="Type your question here..."), outputs=gr.Textbox(label="Matching Questions from Vector Database"))