AI_pdf_reader / app.py
memoona-nz's picture
Upload 3 files
909ffed verified
Raw History Blame Contribute Delete
3.96 kB
import gradio as gr
from langchain_text_splitters import RecursiveCharacterTextSplitter
from langchain_community.document_loaders import PyPDFLoader
from langchain_community.embeddings import HuggingFaceEmbeddings
from langchain_community.vectorstores import FAISS
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
import torch
# MODEL
model_name = "google/flan-t5-base"
model = AutoModelForSeq2SeqLM.from_pretrained(model_name)
tokenizer = AutoTokenizer.from_pretrained(model_name)
# 1. PROCESS PDF
def process_pdfs(files):
if not files:
return None, "No files uploaded."
all_docs = []
for file in files:
loader = PyPDFLoader(file.name)
docs = loader.load()
all_docs.extend(docs)
splitter = RecursiveCharacterTextSplitter(
chunk_size=500,
chunk_overlap=100
)
split_docs = splitter.split_documents(all_docs)
embeddings = HuggingFaceEmbeddings(
model_name="sentence-transformers/all-MiniLM-L6-v2"
)
vectorstore = FAISS.from_documents(split_docs, embeddings)
return vectorstore, f"βœ… {len(split_docs)} chunks processed"
# 2. ASK QUESTION
def ask_question(question, vectorstore):
if vectorstore is None:
return "❌ Please upload PDF files first."
docs = vectorstore.similarity_search(question, k=5)
context = "\n".join([d.page_content for d in docs])
prompt = f"""
You are a university professor teaching Machine Learning.
You MUST give a detailed, long and structured answer.
Format your answer EXACTLY like this:
1. Definition:
2. Explanation:
3. Intuition (why it is used):
4. Step-by-step working (if applicable):
5. Example:
6. Summary:
Do NOT give short answers.
Context:
{context}
Question:
{question}
Now write a complete educational answer:
"""
#3. generated output
inputs = tokenizer(prompt, return_tensors="pt", truncation=True)
with torch.no_grad():
outputs = model.generate(
**inputs,
max_new_tokens=300,
temperature=0.8,
top_p=0.9,
do_sample=True,
repetition_penalty=1.2
)
return tokenizer.decode(outputs[0], skip_special_tokens=True)
# GRADIO UI (FIX IS HERE)
import gradio as gr
with gr.Blocks(theme=gr.themes.Soft(), css="""
#title {
text-align: center;
font-size: 28px;
font-weight: bold;
color: #4A90E2;
}
.container {
max-width: 900px;
margin: auto;
}
.chatbox {
height: 400px;
overflow-y: auto;
border-radius: 12px;
}
button {
border-radius: 10px !important;
}
""") as demo:
gr.Markdown("<h1 id='title'>πŸ“„ AI PDF Q&A Chatbot (RAG System)</h1>")
with gr.Row():
with gr.Column(scale=1):
gr.Markdown("### πŸ“‚ Upload PDF")
pdf_input = gr.File(file_count="multiple", file_types=[".pdf"])
upload_btn = gr.Button(" Process PDFs", variant="primary")
upload_output = gr.Textbox(label="Status")
with gr.Column(scale=2):
gr.Markdown("### πŸ’¬ Ask Questions")
question_input = gr.Textbox(
placeholder="Ask anything from your PDF...",
label="Your Question"
)
ask_btn = gr.Button(" Ask AI", variant="primary")
answer_output = gr.Textbox(
label="AI Answer",
lines=10
)
# -------------------------
# STATE
# -------------------------
vector_state = gr.State(None)
# -------------------------
# EVENTS
# -------------------------
upload_btn.click(
process_pdfs,
inputs=pdf_input,
outputs=[vector_state, upload_output]
)
ask_btn.click(
ask_question,
inputs=[question_input, vector_state],
outputs=answer_output
)
demo.launch()