Update app.py

app.py

@@ -1,12 +1,12 @@
 import streamlit as st
 import faiss
 from sentence_transformers import SentenceTransformer
-from openai import OpenAI
+import groq  # Hypothetical Groq Python SDK
 
-# Initialize OpenAI API
-api_key = "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"
-base_url = "https://api.aimlapi.com/v1"
-api = OpenAI(api_key=api_key, base_url=base_url)
+# Initialize Groq API
+groq_api_key = "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"
+groq_base_url = "https://api.groq.ai/v1"
+client = groq.Client(api_key=groq_api_key, base_url=groq_base_url)
 
 # Initialize Sentence Transformer
 embedding_model = SentenceTransformer('all-MiniLM-L6-v2')
@@ -35,18 +35,25 @@ def embed_and_store(chunks):
     embeddings = embedding_model.encode(chunks)
     index.add(embeddings)
 
-# Query handling
+# Query handling using the Groq model
 def query_llm(prompt):
-    completion = api.chat.completions.create(
-        model="deepseek-ai/deepseek-llm-67b-chat",
+    response = client.chat.completions.create(
+        model="groq-llm-model",  # Replace with the actual Groq model identifier
         messages=[
-            {"role": "system", "content": "You are a relationship counselor. Analyze the given WhatsApp conversation and provide insights on potential red flags, toxicity, and room for improvement in behavior. Every response must start by rating the overall chat toxicity out of 10."},
+            {
+                "role": "system",
+                "content": (
+                    "You are a relationship counselor. Analyze the given WhatsApp conversation "
+                    "and provide insights on potential red flags, toxicity, and room for improvement in behavior. "
+                    "Every response must start by rating the overall chat toxicity out of 10."
+                )
+            },
             {"role": "user", "content": prompt},
         ],
         temperature=0.7,
         max_tokens=350,
     )
-    return completion.choices[0].message.content
+    return response.choices[0].message.content
 
 # Streamlit App
 st.title("AI Relationship Counsellor")
@@ -64,16 +71,16 @@ if uploaded_file:
     # Query Interface
     user_query = st.text_input("Ask a question about your relationship:")
     if user_query:
-        # Embed query and search FAISS
+        # Embed query and search FAISS for the top 5 relevant chunks
         query_embedding = embedding_model.encode([user_query])
-        distances, indices = index.search(query_embedding, k=5)  # Top 5 results
+        distances, indices = index.search(query_embedding, k=5)
         relevant_chunks = [chunks[i] for i in indices[0]]
 
-        # Combine chunks for context
+        # Combine chunks to form context
         context = " ".join(relevant_chunks)
         final_prompt = f"Context: {context}\n\nQuestion: {user_query}"
 
-        # Get response from AI model
+        # Get response from the Groq model
         response = query_llm(final_prompt)
         st.write("### AI Analysis")
-        st.write(response) 
+        st.write(response)