Spaces:
Sleeping
Sleeping
File size: 3,959 Bytes
909ffed | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 | 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"β
{len(split_docs)} chunks processed"
# 2. ASK QUESTION
def ask_question(question, vectorstore):
if vectorstore is None:
return "β 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'>π AI PDF Q&A Chatbot (RAG System)</h1>")
with gr.Row():
with gr.Column(scale=1):
gr.Markdown("### π 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("### π¬ 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() |