Spaces:
Sleeping
Sleeping
| import streamlit as st | |
| import PyPDF2 | |
| import openai | |
| import faiss | |
| import os | |
| import numpy as np | |
| from sklearn.feature_extraction.text import TfidfVectorizer | |
| from sklearn.metrics.pairwise import cosine_similarity | |
| # Function to extract text from a PDF file | |
| def extract_text_from_pdf(pdf_file): | |
| reader = PyPDF2.PdfReader(pdf_file) | |
| text = "" | |
| for page in reader.pages: | |
| text += page.extract_text() | |
| return text | |
| # Function to generate embeddings for a piece of text | |
| def get_embeddings(text, model="text-embedding-ada-002"): | |
| response = openai.Embedding.create(input=[text], model=model) | |
| return response['data'][0]['embedding'] | |
| # Function to search for similar content | |
| def search_similar(query_embedding, index, stored_texts, top_k=3): | |
| distances, indices = index.search(np.array([query_embedding]), top_k) | |
| results = [(stored_texts[i], distances[0][idx]) for idx, i in enumerate(indices[0])] | |
| return results | |
| # Function to generate code based on a prompt | |
| def generate_code_from_prompt(prompt, model="gpt-4o-mini"): | |
| response = openai.ChatCompletion.create( | |
| model=model, | |
| messages=[{"role": "user", "content": prompt}] | |
| ) | |
| return response['choices'][0]['message']['content'] | |
| # Streamlit app starts here | |
| st.title("AI Assistance") | |
| # Input OpenAI API key | |
| openai_api_key = st.text_input("Enter your OpenAI API key:", type="password") | |
| if openai_api_key: | |
| openai.api_key = openai_api_key | |
| # Sidebar to toggle between Course Query Assistant and Code Generator | |
| st.sidebar.title("Select Mode") | |
| mode = st.sidebar.radio("Choose an option", ("Course Query Assistant", "Code Generator")) | |
| if mode == "Course Query Assistant": | |
| st.header("Course Query Assistant") | |
| # Upload course materials | |
| uploaded_files = st.file_uploader("Upload Course Materials (PDFs)", type=["pdf"], accept_multiple_files=True) | |
| if uploaded_files: | |
| st.write("Processing uploaded course materials...") | |
| # Extract text and generate embeddings for all uploaded PDFs | |
| course_texts = [] | |
| for uploaded_file in uploaded_files: | |
| text = extract_text_from_pdf(uploaded_file) | |
| course_texts.append(text) | |
| # Combine all course materials into one large text | |
| combined_text = " ".join(course_texts) | |
| # Split combined text into smaller chunks for embedding (max tokens ~1000) | |
| chunks = [combined_text[i:i+1000] for i in range(0, len(combined_text), 1000)] | |
| # Generate embeddings for all chunks | |
| embeddings = [get_embeddings(chunk) for chunk in chunks] | |
| # Convert the list of embeddings into a NumPy array (shape: [num_chunks, embedding_size]) | |
| embeddings_np = np.array(embeddings).astype("float32") | |
| # Create a FAISS index for similarity search | |
| index = faiss.IndexFlatL2(len(embeddings_np[0])) # Use the length of the embedding vectors for the dimension | |
| index.add(embeddings_np) | |
| st.write("Course materials have been processed and indexed.") | |
| # User query | |
| query = st.text_input("Enter your question about the course materials:") | |
| if query: | |
| # Generate embedding for the query | |
| query_embedding = get_embeddings(query) | |
| # Search for similar chunks in the FAISS index | |
| results = search_similar(query_embedding, index, chunks) | |
| # Create the context for the GPT prompt | |
| context = "\n".join([result[0] for result in results]) | |
| modified_prompt = f"Context: {context}\n\nQuestion: {query}\n\nProvide a detailed answer based on the context." | |
| # Get the GPT-4 response | |
| response = openai.ChatCompletion.create( | |
| model="gpt-4o-mini", # Update to GPT-4 (or your desired model) | |
| messages=[{"role": "user", "content": modified_prompt}] | |
| ) | |
| # Get the response content | |
| response_content = response['choices'][0]['message']['content'] | |
| # Display the response in Streamlit (Intelligent Reply) | |
| st.write("### Intelligent Reply:") | |
| st.write(response_content) | |
| elif mode == "Code Generator": | |
| st.header("Code Generator") | |
| # Code generation prompt input | |
| code_prompt = st.text_area("Describe the code you want to generate:", | |
| "e.g., Write a Python program that generates Fibonacci numbers.") | |
| if st.button("Generate Code"): | |
| if code_prompt: | |
| with st.spinner("Generating code..."): | |
| # Generate code using GPT-4 | |
| generated_code = generate_code_from_prompt(code_prompt) | |
| # Display the generated code | |
| st.write("### Generated Code:") | |
| st.code(generated_code, language="python") | |
| else: | |
| st.error("Please provide a prompt to generate the code.") | |