Update app.py

app.py

@@ -1,14 +1,13 @@
-pip install torch torchvision torchaudio
 import streamlit as st
 import pandas as pd
+import torch  # Ensure PyTorch is imported
 from transformers import AutoTokenizer, AutoModelForCausalLM
 import speech_recognition as sr
 
 # Load the Netflix dataset from CSV
 @st.cache_data
 def load_data():
-    url = "https://huggingface.co/spaces/mfraz/Netflix-data/resolve/main/netflix_titles.csv"
-    return pd.read_csv(url)
+    return pd.read_csv("https://huggingface.co/spaces/mfraz/Netflix-data/resolve/main/netflix_titles.csv")
 
 # Load DialoGPT model and tokenizer
 @st.cache_resource
@@ -17,32 +16,32 @@ def load_model():
     model = AutoModelForCausalLM.from_pretrained("microsoft/DialoGPT-medium")
     return tokenizer, model
 
-# Function to search movie details
+# Function to search the dataset for movie details
 def search_movie_details(query, data):
     query = query.lower()
-    search_columns = ["title", "cast", "director"]
-    
-    results = data.dropna(subset=search_columns)  # Remove NaN values for safe searching
-    results = results[results[search_columns].apply(lambda x: x.astype(str).str.lower().str.contains(query).any(), axis=1)]
-    
+    filtered_data = data.dropna(subset=["title", "cast", "director"])  # Remove NaN values
+    results = filtered_data[
+        filtered_data["title"].str.lower().str.contains(query, na=False) |
+        filtered_data["cast"].str.lower().str.contains(query, na=False) |
+        filtered_data["director"].str.lower().str.contains(query, na=False)
+    ]
     return results
 
 # Function to convert voice to text
 def voice_to_text():
     recognizer = sr.Recognizer()
     with sr.Microphone() as source:
-        st.write("🎙 Speak now...")
+        st.write("Speak now...")
+        audio = recognizer.listen(source)
         try:
-            audio = recognizer.listen(source, timeout=5)
-            text = recognizer.recognize_google(audio)
-            return text
+            return recognizer.recognize_google(audio)
         except sr.UnknownValueError:
             return "Sorry, I could not understand the audio."
         except sr.RequestError:
             return "Sorry, the speech service is down."
 
 # Streamlit App
-st.title("🎬 Netflix Movie Search Chatbot")
+st.title("Netflix Movie Details Chatbot 🎬")
 
 # Load dataset and model
 data = load_data()
@@ -55,29 +54,28 @@ user_input = ""
 if input_option == "Text":
     user_input = st.text_input("Enter the movie name, director, or cast:")
 elif input_option == "Voice":
-    if st.button("🎤 Start Recording"):
+    if st.button("Start Recording"):
         user_input = voice_to_text()
-        st.write(f"🗣 You said: **{user_input}**")
+        st.write(f"You said: {user_input}")
 
 # Generate response
 if user_input:
-    # Search for movie details
     movie_results = search_movie_details(user_input, data)
     
     if not movie_results.empty:
-        st.write("🎥 **Here are the matching results:**")
+        st.write("Here are the matching results:")
         for _, row in movie_results.iterrows():
-            st.write(f"**📌 Title:** {row.get('title', 'N/A')}")
-            st.write(f"**🎭 Type:** {row.get('type', 'N/A')}")
-            st.write(f"**🎬 Director:** {row.get('director', 'N/A')}")
-            st.write(f"**👥 Cast:** {row.get('cast', 'N/A')}")
-            st.write(f"**📅 Release Year:** {row.get('release_year', 'N/A')}")
-            st.write(f"**⭐ Rating:** {row.get('rating', 'N/A')}")
-            st.write(f"**📝 Description:** {row.get('description', 'N/A')}")
+            st.write(f"**Title:** {row['title']}")
+            st.write(f"**Type:** {row['type']}")
+            st.write(f"**Director:** {row['director']}")
+            st.write(f"**Cast:** {row['cast']}")
+            st.write(f"**Release Year:** {row['release_year']}")
+            st.write(f"**Rating:** {row['rating']}")
+            st.write(f"**Description:** {row['description']}")
             st.write("---")
     else:
         # Use DialoGPT for general conversation
         inputs = tokenizer.encode(user_input, return_tensors="pt")
         outputs = model.generate(inputs, max_length=100, num_return_sequences=1)
         response = tokenizer.decode(outputs[0], skip_special_tokens=True)
-        st.write(f"🤖 **Chatbot:** {response}")
+        st.write(f"Chatbot: {response}")