Update app.py

app.py

@@ -1,29 +1,59 @@
+import torch
 import gradio as gr
-import random
-from huggingface_hub import InferenceClient 
+from huggingface_hub import InferenceClient
+from sentence_transformers import SentenceTransformer
+
+with open("travel_knowledge_base.txt", "r", encoding="utf-8") as file:
+    travel_text = file.read()
+
+def preprocess_text(text):
+    cleaned_text = text.strip()
+    chunks = cleaned_text.split("\n")
+    cleaned_chunks = []
+    for chunk in chunks:
+        chunk = chunk.strip()
+        if chunk != "":
+            cleaned_chunks.append(chunk)
+    return cleaned_chunks
+
+cleaned_chunks = preprocess_text(travel_text)
+
+
+model = SentenceTransformer('all-MiniLM-L6-v2')
+
+def create_embeddings(text_chunks):
+    chunk_embeddings = model.encode(text_chunks, convert_to_tensor=True)
+    return chunk_embeddings
+
+chunk_embeddings = create_embeddings(cleaned_chunks)
+
+def get_top_chunks(query, chunk_embeddings, text_chunks):
+    query_embedding = model.encode(query, convert_to_tensor=True)
+    query_embedding_normalized = query_embedding / query_embedding.norm()
+    chunk_embeddings_normalized = chunk_embeddings / chunk_embeddings.norm(dim=1, keepdim=True)
+    similarities = torch.matmul(chunk_embeddings_normalized, query_embedding_normalized)
+    top_indices = torch.topk(similarities, k=3).indices
+    top_chunks = []
+    for index in top_indices:
+        top_chunks.append(text_chunks[index])
+    return top_chunks
 
 client = InferenceClient("Qwen/Qwen2.5-7B-Instruct")
 
 def respond(message, history):
-    messages = [{"role": "system", "content": "You are a travel advisor helping people plan trips. You are able to give info on anything about flights, food, places, landmarks, and more."}]
+    top_chunks = get_top_chunks(message, chunk_embeddings, cleaned_chunks)
+    context = "\n\n".join(top_chunks)
+    messages = [{"role": "system", "content": f"You are a travel advisor. Give information about transportation, food, hotels, landmarks, and more. \n{context}"}]
     if history:
         messages.extend(history)
     messages.append({"role": "user", "content": message})
 
-    # response = client.chat_completion(
-    #     messages,
-    #     max_tokens=100,
-    #     temperature = 1.5,
-    #     top_p = 0.34
-    # )
     response = ""
     for msg in client.chat_completion(messages, stream=True):
         token = msg.choices[0].delta.content
         if token is not None:
             response += token
-            yield response #response.choices[0].message.content.strip()
-    
-chatbot = gr.ChatInterface(respond, title="my bot")
-
+            yield response
 
+chatbot = gr.ChatInterface(respond, type="messages")
 chatbot.launch(debug=True)