File size: 4,905 Bytes
472fc23 175d074 3f6ce03 175d074 414b472 175d074 3f6ce03 7a4c79a 3f6ce03 9c26bdd 3f6ce03 f4d4afc 414b472 175d074 3f6ce03 175d074 414b472 175d074 3f6ce03 175d074 3f6ce03 175d074 414b472 175d074 d6a5a31 175d074 3f6ce03 175d074 d6a5a31 175d074 d6a5a31 175d074 3f6ce03 |
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 |
import spaces
import gradio as gr
from datasets import load_dataset
import os
from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer, BitsAndBytesConfig
import torch
from threading import Thread
from sentence_transformers import SentenceTransformer
import faiss
import fitz # PyMuPDF
# ํ๊ฒฝ ๋ณ์์์ Hugging Face ํ ํฐ ๊ฐ์ ธ์ค๊ธฐ
token = os.environ.get("HF_TOKEN")
# ์๋ฒ ๋ฉ ๋ชจ๋ธ ๋ก๋
ST = SentenceTransformer("jhgan/ko-sroberta-multitask")
# PDF์์ ํ
์คํธ ์ถ์ถ
def extract_text_from_pdf(pdf_path):
doc = fitz.open(pdf_path)
text = ""
for page in doc:
text += page.get_text()
return text
# ๋ฒ๋ฅ ๋ฌธ์ PDF ๊ฒฝ๋ก ์ง์ ๋ฐ ํ
์คํธ ์ถ์ถ
pdf_path = "laws.pdf" # ์ฌ๊ธฐ์ ์ค์ PDF ๊ฒฝ๋ก๋ฅผ ์
๋ ฅํ์ธ์.
law_text = extract_text_from_pdf(pdf_path)
# ๋ฒ๋ฅ ๋ฌธ์ ํ
์คํธ๋ฅผ ๋ฌธ์ฅ ๋จ์๋ก ๋๋๊ณ ์๋ฒ ๋ฉ
law_sentences = law_text.split('\n') # Adjust splitting based on your PDF structure
law_embeddings = ST.encode(law_sentences)
# FAISS ์ธ๋ฑ์ค ์์ฑ ๋ฐ ์๋ฒ ๋ฉ ์ถ๊ฐ
index = faiss.IndexFlatL2(law_embeddings.shape[1])
index.add(law_embeddings)
# Hugging Face์์ ๋ฒ๋ฅ ์๋ด ๋ฐ์ดํฐ์
๋ก๋
dataset = load_dataset("jihye-moon/LawQA-Ko")
data = dataset["train"]
# ์ง๋ฌธ ์ปฌ๋ผ์ ์๋ฒ ๋ฉํ์ฌ ์๋ก์ด ์ปฌ๋ผ์ ์ถ๊ฐ
data = data.map(lambda x: {"question_embedding": ST.encode(x["question"])}, batched=True)
data.add_faiss_index(column="question_embedding")
# LLaMA ๋ชจ๋ธ ์ค์
model_id = "google/gemma-2-27b-it"
bnb_config = BitsAndBytesConfig(
load_in_4bit=True, bnb_4bit_use_double_quant=True, bnb_4bit_quant_type="nf4", bnb_4bit_compute_dtype=torch.bfloat16
)
tokenizer = AutoTokenizer.from_pretrained(model_id, token=token)
model = AutoModelForCausalLM.from_pretrained(
model_id,
torch_dtype=torch.bfloat16,
device_map="auto",
quantization_config=bnb_config,
token=token
)
SYS_PROMPT = """You are an assistant for answering legal questions.
You are given the extracted parts of legal documents and a question. Provide a conversational answer.
If you don't know the answer, just say "I do not know." Don't make up an answer.
you must answer korean."""
# ๋ฒ๋ฅ ๋ฌธ์ ๊ฒ์
@spaces.Gpu
def search_law(query, k=5):
query_embedding = ST.encode([query])
D, I = index.search(query_embedding, k)
return [(law_sentences[i], D[0][idx]) for idx, i in enumerate(I[0])]
# ๋ฒ๋ฅ ์๋ด ๋ฐ์ดํฐ ๊ฒ์ ํจ์
@spaces.Gpu
def search_qa(query, k=3):
scores, retrieved_examples = data.get_nearest_examples(
"question_embedding", ST.encode(query), k=k
)
return [retrieved_examples["answer"][i] for i in range(k)]
# ์ต์ข
ํ๋กฌํํธ ์์ฑ
def format_prompt(prompt, law_docs, qa_docs):
PROMPT = f"Question: {prompt}\n\nLegal Context:\n"
for doc in law_docs:
PROMPT += f"{doc[0]}\n" # Assuming doc[0] contains the relevant text
PROMPT += "\nLegal QA:\n"
for doc in qa_docs:
PROMPT += f"{doc}\n"
return PROMPT
# ์ฑ๋ด ์๋ต ํจ์
@spaces.Gpu
def talk(prompt, history):
law_results = search_law(prompt, k=3)
qa_results = search_qa(prompt, k=3)
retrieved_law_docs = [result[0] for result in law_results]
formatted_prompt = format_prompt(prompt, retrieved_law_docs, qa_results)
formatted_prompt = formatted_prompt[:2000] # GPU ๋ฉ๋ชจ๋ฆฌ ๋ถ์กฑ์ ํผํ๊ธฐ ์ํด ํ๋กฌํํธ ์ ํ
# Adjust the message roles
messages = [{"role": "user", "content": SYS_PROMPT + "\n" + formatted_prompt}]
input_ids = tokenizer.apply_chat_template(
messages,
add_generation_prompt=True,
return_tensors="pt"
).to(model.device)
streamer = TextIteratorStreamer(
tokenizer, timeout=10.0, skip_prompt=True, skip_special_tokens=True
)
generate_kwargs = dict(
input_ids=input_ids,
streamer=streamer,
max_new_tokens=1024,
do_sample=True,
top_p=0.95,
temperature=0.2,
eos_token_id=tokenizer.eos_token_id,
)
t = Thread(target=model.generate, kwargs=generate_kwargs)
t.start()
outputs = []
for text in streamer:
outputs.append(text)
yield "".join(outputs)
# Gradio ์ธํฐํ์ด์ค ์ค์
TITLE = "Legal RAG Chatbot"
DESCRIPTION = """A chatbot that uses Retrieval-Augmented Generation (RAG) for legal consultation.
This chatbot can search legal documents and previous legal QA pairs to provide answers."""
demo = gr.ChatInterface(
fn=talk,
chatbot=gr.Chatbot(
show_label=True,
show_share_button=True,
show_copy_button=True,
likeable=True,
layout="bubble",
bubble_full_width=False,
),
theme="Soft",
examples=[["What are the regulations on data privacy?"]],
title=TITLE,
description=DESCRIPTION,
)
# Gradio ๋ฐ๋ชจ ์คํ
demo.launch(debug=True) |