File size: 2,488 Bytes
bc23008 293840d f6afc8f bc23008 d8fad44 5a87c2f 293840d d8fad44 19a84e6 d8fad44 19a84e6 bc23008 293840d 9dd451f e30807f 5264200 f5b923e bc23008 293840d bc23008 5264200 bc23008 293840d bc23008 5264200 293840d 5264200 293840d |
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 |
import streamlit as st
from hazm import Normalizer, SentenceTokenizer
import os
import docx
from langchain.chat_models import ChatOpenAI
# LLM setup
llm = ChatOpenAI(
base_url="https://api.together.xyz/v1",
api_key='0291f33aee03412a47fa5d8e562e515182dcc5d9aac5a7fb5eefdd1759005979',
model="deepseek-ai/DeepSeek-R1"
)
def rewrite_answer_with_llm(answer, user_input):
prompt = f"پاسخی که باید بازنویسی شود:\n{answer}\n\nلطفاً این پاسخ را با در نظر گرفتن محتوای سوال زیر و لحن آن بازنویسی کن:\n\nسوال: {user_input}"
response = llm([{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": prompt}])
return response.strip()
folder_path = '46'
texts = []
for filename in os.listdir(folder_path):
if filename.endswith(".docx"):
full_path = os.path.join(folder_path, filename)
doc = docx.Document(full_path)
file_text = "\n".join([para.text for para in doc.paragraphs])
if file_text.strip():
texts.append(file_text)
normalizer = Normalizer()
sentence_tokenizer = SentenceTokenizer()
all_sentences = []
for text in texts:
normalized = normalizer.normalize(text)
sentences = sentence_tokenizer.tokenize(normalized)
all_sentences.extend(sentences)
query = st.text_input("🔎 کلمه یا عبارت موردنظر خود را وارد کنید:")
# ✅ نمایش جمله و ۵ جمله بعدی + بازنویسی با LLM
if query:
found = False
for idx, sentence in enumerate(all_sentences):
if query in sentence:
st.success("✅ جمله یافت شد:")
st.write(sentence)
next_sentences = []
st.markdown("📌 پنج جمله بعدی:")
for i in range(1, 6):
if idx + i < len(all_sentences):
st.write(all_sentences[idx + i])
next_sentences.append(all_sentences[idx + i])
# ↪️ آماده‌سازی برای بازنویسی
total_text = sentence + " " + " ".join(next_sentences)
rewritten = rewrite_answer_with_llm(total_text, query)
st.markdown("🎨 **بازنویسی شده با LLM:**")
st.write(rewritten)
found = True
break
if not found:
st.warning("عبارت موردنظر در متن یافت نشد.")
|