Upload app.py

app.py

@@ -0,0 +1,132 @@
+import gradio as gr
+from langchain.document_loaders import PyPDFLoader
+from langchain.text_splitter import CharacterTextSplitter
+from langchain.embeddings import SentenceTransformerEmbeddings
+from langchain.vectorstores import FAISS
+from langchain.memory import ConversationBufferMemory
+from groq import Groq
+import requests
+from bs4 import BeautifulSoup
+import time  
+
+
+client = Groq(api_key="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")
+embedding_model = SentenceTransformerEmbeddings(model_name="sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2")
+
+memory = ConversationBufferMemory(memory_key="chat_history", return_messages=True)
+
+def process_pdf_with_langchain(pdf_path, progress_callback):
+    # progress_callback("Initializing PDF processing... 0%")
+    time.sleep(0.5)
+    loader = PyPDFLoader(pdf_path)
+    # progress_callback("Loading PDF... 20%")
+    documents = loader.load()
+    time.sleep(0.5)
+    # progress_callback("Splitting documents... 50%")
+    text_splitter = CharacterTextSplitter(chunk_size=500, chunk_overlap=50)
+    split_documents = text_splitter.split_documents(documents)
+    time.sleep(0.5)
+    # progress_callback("Creating vector store... 80%")
+    vectorstore = FAISS.from_documents(split_documents, embedding_model)
+    retriever = vectorstore.as_retriever(search_kwargs={"k": 3})
+    progress_callback("Processing complete! 100%")
+    return retriever
+
+
+def scrape_google_search(query, num_results=3):
+
+    headers = {"User-Agent": "Mozilla/5.0"}
+    search_url = f"https://www.google.com/search?q={query}"
+    response = requests.get(search_url, headers=headers)
+    soup = BeautifulSoup(response.text, "html.parser")
+
+    results = []
+    for g in soup.find_all('div', class_='tF2Cxc')[:num_results]:
+        title = g.find('h3').text
+        link = g.find('a')['href']
+        results.append(f"{title}: {link}")
+    return "\n".join(results)
+
+
+def generate_response(query, retriever=None, use_web_search=False):
+
+    knowledge = ""
+
+    if retriever:
+        relevant_docs = retriever.get_relevant_documents(query)
+        knowledge += "\n".join([doc.page_content for doc in relevant_docs])
+
+    if use_web_search:
+        web_results = scrape_google_search(query)
+        knowledge += f"\n\nWeb Search Results:\n{web_results}"
+
+    chat_history = memory.load_memory_variables({}).get("chat_history", "")
+    context = (
+        f"This is a conversation with ParvizGPT, an AI model designed by Amir Mahdi Parviz from Kermanshah University of Technology (KUT), "
+        f"to help with tasks like answering questions in Persian, providing recommendations, and decision-making."
+    )
+    if knowledge:
+        context += f"\n\nRelevant Knowledge:\n{knowledge}"
+    if chat_history:
+        context += f"\n\nChat History:\n{chat_history}"
+
+    context += f"\n\nYou: {query}\nParvizGPT:"
+
+    chat_completion = client.chat.completions.create(
+        messages=[{"role": "user", "content": context}],
+        model="llama-3.3-70b-versatile",
+    )
+    response = chat_completion.choices[0].message.content.strip()
+
+    memory.save_context({"input": query}, {"output": response})
+    return response
+
+def upload_and_process(file, progress_display):
+    try:
+        global retriever
+        progress_updates = []
+        
+        retriever = process_pdf_with_langchain(file.name, lambda msg: progress_updates.append(msg))
+
+        return "\n".join(progress_updates), "File uploaded and processed successfully."
+    except Exception as e:
+        return "", f"Error processing file: {e}"
+
+def gradio_interface(user_message, chat_box, enable_web_search=False):
+    global retriever
+    response = generate_response(user_message, retriever=retriever, use_web_search=enable_web_search)
+    chat_box.append(("You", user_message))
+    chat_box.append(("ParvizGPT", response))
+    return chat_box
+
+def clear_memory():
+    memory.clear()
+    return []
+
+retriever = None
+with gr.Blocks() as interface:
+    gr.Markdown("## ParvizGPT")
+    with gr.Row():
+        chat_box = gr.Chatbot(label="Chat History", value=[])
+    with gr.Row():
+        user_message = gr.Textbox(
+            label="Your Message",
+            placeholder="Type your message here and press Enter...",
+            lines=1,
+            interactive=True,
+        )
+    with gr.Row():
+        clear_memory_btn = gr.Button("Clear Memory", interactive=True)
+        enable_web_search = gr.Checkbox(label="🌐Enable Web Search", value=False, interactive=True)
+    with gr.Row():
+        pdf_upload = gr.UploadButton(label="📄 Upload Your PDF", file_types=[".pdf"])
+        progress_display = gr.Textbox(label="Progress", placeholder="Progress updates will appear here", interactive=True)
+    with gr.Row():
+        submit_btn = gr.Button("Submit")
+    pdf_upload.upload(upload_and_process, inputs=[pdf_upload, progress_display], outputs=[progress_display])
+
+    submit_btn.click(gradio_interface, inputs=[user_message, chat_box, enable_web_search], outputs=chat_box)
+    user_message.submit(gradio_interface, inputs=[user_message, chat_box, enable_web_search], outputs=chat_box)
+    clear_memory_btn.click(clear_memory, inputs=[], outputs=chat_box)
+
+interface.launch()