Cursor Merge

app.py

@@ -1,20 +1,14 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
-import os
-
-# Load token from environment variable
-token = os.getenv("HUGGINGFACEHUB_API_TOKEN")
-if not token:
-    raise ValueError("HUGGINGFACEHUB_API_TOKEN environment variable not set.")
-
-# Instantiate the inference client with token from env
-client = InferenceClient(
-    "HuggingFaceH4/zephyr-7b-beta",
-    token=token
-)
+from huggingface_hub.inference._client import InferenceClient
+
+"""
+For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
+"""
+client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
+
 
 def respond(
-    message: str,
+    message,
     history: list[tuple[str, str]],
     system_message,
     max_tokens,
@@ -23,37 +17,48 @@ def respond(
 ):
     messages = [{"role": "system", "content": system_message}]
 
-    # Convert Gradio history tuples into OpenAI-style messages
-    for user_msg, assistant_msg in history:
-        messages.append({"role": "user", "content": user_msg})
-        messages.append({"role": "assistant", "content": assistant_msg})
+    for val in history:
+        if val[0]:
+            messages.append({"role": "user", "content": val[0]})
+        if val[1]:
+            messages.append({"role": "assistant", "content": val[1]})
 
-    # Append latest user message
     messages.append({"role": "user", "content": message})
 
-    try:
-        result = client.chat_completion(
-            messages=messages,
-            max_tokens=max_tokens,
-            stream=False,
-            temperature=temperature,
-            top_p=top_p,
-        )
-        yield result.choices[0].message.content
-    except Exception as e:
-        yield f"Error: {str(e)}"
-
-# Build the Gradio interface
+    response = ""
+
+    for message in client.chat_completion(
+        messages,
+        max_tokens=max_tokens,
+        stream=True,
+        temperature=temperature,
+        top_p=top_p,
+    ):
+        token = message.choices[0].delta.content
+        if token:
+            response += token
+            yield response
+
+
+"""
+For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
+"""
 demo = gr.ChatInterface(
     respond,
     additional_inputs=[
         gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
         gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
         gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)"),
+        gr.Slider(
+            minimum=0.1,
+            maximum=1.0,
+            value=0.95,
+            step=0.05,
+            label="Top-p (nucleus sampling)",
+        ),
     ],
-    type="messages"
 )
 
+
 if __name__ == "__main__":
     demo.launch()

requirements.txt

@@ -1,2 +1,11 @@
-huggingface_hub==0.25.2
-gradio
+gradio==4.31.0
+langchain==0.2.1
+langgraph==0.0.62
+huggingface_hub==0.23.0
+langchain-huggingface==0.0.3
+faiss-cpu==1.8.0
+langchain_community==0.2.1
+python-dotenv==1.0.1
+tiktoken==0.7.0
+langchain-text-splitters==0.2.0
+sentence-transformers==2.7.0

simple_turbot.py

@@ -0,0 +1,146 @@
+#!/usr/bin/env python3
+"""
+Simple TurBot implementation that works with Hugging Face.
+"""
+
+import os
+from typing import List
+from dotenv import load_dotenv
+from huggingface_hub.inference._client import InferenceClient
+from langchain_community.vectorstores import FAISS
+from langchain_huggingface import HuggingFaceEmbeddings
+
+# Load environment variables
+load_dotenv()
+
+# Initialize the LLM
+client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
+
+# Load the vector store
+DB_FAISS_PATH = "vectorstore/db_faiss"
+
+def search_travel_packages(query: str) -> str:
+    """
+    Search for travel packages based on the user's query.
+    """
+    try:
+        # Load the vector store
+        embeddings = HuggingFaceEmbeddings(
+            model_name='sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2',
+            model_kwargs={'device': 'cpu'}
+        )
+        
+        if os.path.exists(DB_FAISS_PATH):
+            vectorstore = FAISS.load_local(DB_FAISS_PATH, embeddings)
+            
+            # Search for relevant documents
+            docs = vectorstore.similarity_search(query, k=3)
+            
+            # Combine the relevant information
+            context = "\n\n".join([doc.page_content for doc in docs])
+            return f"Found relevant travel information:\n\n{context}"
+        else:
+            return "Travel package database not found. Please ensure the vector store has been created."
+    except Exception as e:
+        return f"Error searching travel packages: {str(e)}"
+
+def should_search_travel_info(message: str) -> bool:
+    """Decide whether to search for travel information based on the message content."""
+    message_lower = message.lower()
+    
+    # Keywords that indicate we should search for travel information
+    travel_keywords = [
+        "destinacija", "let", "hotel", "putovanje", "mediteran", "grčka", "italija", 
+        "turska", "budžet", "porodica", "all inclusive", "preporučite", "ponuda",
+        "destination", "flight", "travel", "mediterranean", "greece", "italy", 
+        "turkey", "budget", "family", "recommend", "offer"
+    ]
+    
+    return any(keyword in message_lower for keyword in travel_keywords)
+
+def chat_with_turbot(message: str, history: List[List[str]]) -> str:
+    """
+    Main function to chat with TurBot.
+    """
+    # Convert history to the format expected by our LLM
+    messages = []
+    for human, ai in history:
+        if human:
+            messages.append({"role": "user", "content": human})
+        if ai:
+            messages.append({"role": "assistant", "content": ai})
+    
+    # Add the current message
+    messages.append({"role": "user", "content": message})
+    
+    # Check if we should search for travel information
+    context = ""
+    if should_search_travel_info(message):
+        context = search_travel_packages(message)
+    
+    # Prepare the system message
+    if context:
+        system_message = f"""Ti si TurBot, digitalni asistent za turističku agenciju. 
+        Koristi sledeće informacije o putovanjima da odgovoriš na pitanje korisnika:
+        
+        {context}
+        
+        Odgovori na srpskom jeziku i budi koristan i prijateljski nastrojen."""
+    else:
+        system_message = """Ti si TurBot, digitalni asistent za turističku agenciju. 
+        Odgovori na srpskom jeziku i budi koristan i prijateljski nastrojen. 
+        Ako ne znaš odgovor na pitanje, preporuči korisniku da pita o putovanjima, letovima, hotelima ili destinacijama."""
+    
+    # Add system message to the beginning
+    messages.insert(0, {"role": "system", "content": system_message})
+    
+    try:
+        # Generate response
+        response = ""
+        for message_chunk in client.chat_completion(
+            messages,
+            stream=True,
+            temperature=0.7,
+            top_p=0.95,
+        ):
+            token = message_chunk.choices[0].delta.content
+            if token:
+                response += token
+        
+        return response
+    except Exception as e:
+        return f"Izvinjavam se, došlo je do greške: {str(e)}"
+
+def test_turbot():
+    """Test the TurBot agent with some sample questions."""
+    
+    print("🤖 Testing TurBot - Digitalni Asistent za Turističku Agenciju")
+    print("=" * 60)
+    
+    test_questions = [
+        "Zdravo! Ko ste vi?",
+        "Koje letnje ponude imate za Mediteran?",
+        "Tražim putovanje u Grčku za porodicu od četiri osobe.",
+        "Koji all-inclusive paketi za Tursku su dostupni?"
+    ]
+    
+    history = []
+    
+    for question in test_questions:
+        print(f"\n👤 Korisnik: {question}")
+        print("-" * 40)
+        
+        try:
+            response = chat_with_turbot(question, history)
+            print(f"🤖 TurBot: {response}")
+            
+            # Add to history for next iteration
+            history.append([question, response])
+            
+        except Exception as e:
+            print(f"❌ Greška: {str(e)}")
+        
+        print("-" * 40)
+
+if __name__ == "__main__":
+    test_turbot() 

test_agent.py

@@ -0,0 +1,40 @@
+#!/usr/bin/env python3
+"""
+Simple test script for TurBot agent.
+"""
+
+from src.agent import chat_with_turbot
+
+def test_turbot():
+    """Test the TurBot agent with some sample questions."""
+    
+    print("🤖 Testing TurBot - Digitalni Asistent za Turističku Agenciju")
+    print("=" * 60)
+    
+    test_questions = [
+        "Zdravo! Ko ste vi?",
+        "Koje letnje ponude imate za Mediteran?",
+        "Tražim putovanje u Grčku za porodicu od četiri osobe.",
+        "Koji all-inclusive paketi za Tursku su dostupni?"
+    ]
+    
+    history = []
+    
+    for question in test_questions:
+        print(f"\n👤 Korisnik: {question}")
+        print("-" * 40)
+        
+        try:
+            response = chat_with_turbot(question, history)
+            print(f"🤖 TurBot: {response}")
+            
+            # Add to history for next iteration
+            history.append([question, response])
+            
+        except Exception as e:
+            print(f"❌ Greška: {str(e)}")
+        
+        print("-" * 40)
+
+if __name__ == "__main__":
+    test_turbot()