Update app.py

app.py

@@ -2,7 +2,6 @@ import gradio as gr
 from sentence_transformers import SentenceTransformer
 import torch
 from huggingface_hub import InferenceClient
-from sentence_transformers import util
 #import os
 
 #HF_TOKEN = os.environ.get("HF_TOKEN")
@@ -59,58 +58,26 @@ def get_top_chunks(query): # store a function that gets the most relevant_info a
 client = InferenceClient("google/gemma-3-27b-it")
 
 
-#def respond(message, history):
-#
- #   system_message = "You are a helpful chatbot named scooby, kinda like the cartoon character but not too much like it, who knows alot about pets and their diets and loves helping pet owners, and can only answer questions about pets"
-  #  messages = [{"role":  "system", "content": system_message}]
-
-    
-      
-   # if history:
-    #    messages.extend(history)
-        
-   # messages.append({"role": "user", "content": message})
-    #response = ""
-    #for messages in client.chat_completion(messages,max_tokens = 2500, stream = True):
-     #   token = messages.choices[0].delta.content
-      #  response += token
-       # yield response
-
 def respond(message, history):
-    # Step 1: Embed the user's question
-    message_embedding = model.encode(message, convert_to_tensor=True)
-
-    # Step 2: Calculate similarity with knowledge chunks
-    scores = util.cos_sim(message_embedding, chunk_embeddings)[0]
-    top_k = 3  # You can adjust how many chunks you want to include
-    top_results = torch.topk(scores, k=top_k)
-
-    # Step 3: Retrieve the top relevant knowledge chunks
-    retrieved_knowledge = "\n".join([chunks[i] for i in top_results.indices])
-
-    # Step 4: Build system message with retrieved knowledge
-    system_message = (
-        "You are a helpful chatbot named Scooby, kind of like the cartoon character but not too much.
-        You know a lot about pets and their diets, and you only answer questions about pets.
-        Use the following relevant knowledge to help answer the user's question"
-        + retrieved_knowledge
-    )
 
-    # Step 5: Compose message list for the LLM
-    messages = [{"role": "system", "content": system_message}]
+    system_message = "You are a helpful chatbot named scooby, kinda like the cartoon character but not too much like it, who knows alot about pets and their diets and loves helping pet owners, and can only answer questions about pets"
+    messages = [{"role":  "system", "content": system_message}]
 
+    
+      
     if history:
         messages.extend(history)
-
+        
     messages.append({"role": "user", "content": message})
-
-    # Step 6: Stream response
     response = ""
-    for chunk in client.chat_completion(messages, max_tokens=2500, stream=True):
-        token = chunk.choices[0].delta.content
+    for messages in client.chat_completion(messages,max_tokens = 2500, stream = True):
+        token = messages.choices[0].delta.content
         response += token
         yield response
 
+
+
+
 #theme = gr.themes.Origin(primary_hue="orange",secondary_hue="indigo", neutral_hue="teal")