Spaces:
Sleeping
Sleeping
File size: 4,102 Bytes
59bf8b2 0af6850 59bf8b2 0af6850 59bf8b2 0af6850 4fd36e5 0af6850 59bf8b2 0af6850 59bf8b2 0af6850 59bf8b2 0af6850 59bf8b2 0af6850 59bf8b2 0af6850 59bf8b2 0af6850 59bf8b2 0af6850 59bf8b2 0af6850 59bf8b2 0af6850 59bf8b2 0af6850 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 |
import time
import logging
import gradio as gr
from langchain.document_loaders import PyPDFLoader
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.embeddings import HuggingFaceEmbeddings
from langchain.vectorstores import FAISS
from langchain_core.vectorstores import InMemoryVectorStore
from groq import Groq
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)
client = Groq(api_key="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")
embeddings = HuggingFaceEmbeddings(model_name="heydariAI/persian-embeddings")
vector_store = InMemoryVectorStore(embeddings)
def process_pdf_with_langchain(pdf_path):
try:
loader = PyPDFLoader(pdf_path)
documents = loader.load()
text_splitter = RecursiveCharacterTextSplitter(chunk_size=500, chunk_overlap=50)
split_documents = text_splitter.split_documents(documents)
vectorstore = FAISS.from_documents(split_documents, embedding_model)
retriever = vectorstore.as_retriever(search_kwargs={"k": 3})
return retriever
except Exception as e:
logger.error(f"Error processing PDF: {e}")
raise
def generate_response(query, retriever=None):
try:
knowledge = ""
if retriever:
relevant_docs = retriever.get_relevant_documents(query)
knowledge += "\n".join([doc.page_content for doc in relevant_docs])
chat_history = memory.load_memory_variables({}).get("chat_history", "")
context = "This is a conversation with ParvizGPT, an AI model designed by Amir Mahdi Parviz from KUT."
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:"
# ابتدا یک پیام موقت نمایش داده شود
response = "در حال پردازش..."
retries = 3
for attempt in range(retries):
try:
chat_completion = client.chat.completions.create(
messages=[{"role": "user", "content": context}],
model="deepseek-r1-distill-llama-70b"
)
response = chat_completion.choices[0].message.content.strip()
memory.save_context({"input": query}, {"output": response})
break
except Exception as e:
logger.error(f"Attempt {attempt + 1} failed: {e}")
time.sleep(2)
return response
except Exception as e:
logger.error(f"Error generating response: {e}")
return f"Error: {e}"
def gradio_interface(user_message, chat_box, pdf_file=None):
global retriever
if pdf_file is not None:
try:
retriever = process_pdf_with_langchain(pdf_file.name)
except Exception as e:
return chat_box + [("Error", f"Error processing PDF: {e}")]
chat_box.append(("ParvizGPT", "در حال پردازش..."))
response = generate_response(user_message, retriever=retriever)
chat_box[-1] = ("ParvizGPT", response)
chat_box.append(("You", user_message))
return chat_box
def clear_memory():
memory.clear()
return []
retriever = None
with gr.Blocks() as interface:
gr.Markdown("## ParvizGPT")
chat_box = gr.Chatbot(label="Chat History", value=[])
user_message = gr.Textbox(label="Your Message", placeholder="Type your message here and press Enter...", lines=1, interactive=True)
clear_memory_btn = gr.Button("Clear Memory", interactive=True)
pdf_file = gr.File(label="Upload PDF for Context (Optional)", type="filepath", interactive=True, scale=1)
submit_btn = gr.Button("Submit")
submit_btn.click(gradio_interface, inputs=[user_message, chat_box, pdf_file], outputs=chat_box)
user_message.submit(gradio_interface, inputs=[user_message, chat_box, pdf_file], outputs=chat_box)
clear_memory_btn.click(clear_memory, inputs=[], outputs=chat_box)
interface.launch()
|