pip install torch torchvision torchaudio import streamlit as st import pandas as pd 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) # Load DialoGPT model and tokenizer @st.cache_resource def load_model(): tokenizer = AutoTokenizer.from_pretrained("microsoft/DialoGPT-medium") model = AutoModelForCausalLM.from_pretrained("microsoft/DialoGPT-medium") return tokenizer, model # Function to search 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)] return results # Function to convert voice to text def voice_to_text(): recognizer = sr.Recognizer() with sr.Microphone() as source: st.write("🎙 Speak now...") try: audio = recognizer.listen(source, timeout=5) text = recognizer.recognize_google(audio) return text 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") # Load dataset and model data = load_data() tokenizer, model = load_model() # Input options: Text or Voice input_option = st.radio("Choose input method:", ("Text", "Voice")) 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"): user_input = voice_to_text() 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:**") 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("---") 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}")