Update app.py

app.py

@@ -2,118 +2,119 @@
 
 import os
 import re
-import tempfile
-import gc  # garbage collector 추가
-from collections.abc import Iterator
-from threading import Thread
 import json
 import requests
-import cv2
+from collections.abc import Iterator
+from threading import Thread
+
 import gradio as gr
-import spaces
-import torch
 from loguru import logger
-from PIL import Image
-from transformers import AutoProcessor, Gemma3ForConditionalGeneration, TextIteratorStreamer
-
-# CSV/TXT 분석
 import pandas as pd
-# PDF 텍스트 추출
 import PyPDF2
 
 ##############################################################################
-# 메모리 정리 함수 추가
+# API Configuration
 ##############################################################################
-def clear_cuda_cache():
-    """CUDA 캐시를 명시적으로 비웁니다."""
-    if torch.cuda.is_available():
-        torch.cuda.empty_cache()
-        gc.collect()
+FRIENDLI_TOKEN = os.environ.get("FRIENDLI_TOKEN")
+if not FRIENDLI_TOKEN:
+    raise ValueError("Please set FRIENDLI_TOKEN environment variable")
+
+FRIENDLI_MODEL_ID = "dep89a2fld32mcm"
+FRIENDLI_API_URL = "https://api.friendli.ai/dedicated/v1/chat/completions"
+
+# SERPHouse API key
+SERPHOUSE_API_KEY = os.getenv("SERPHOUSE_API_KEY", "")
+if not SERPHOUSE_API_KEY:
+    logger.warning("SERPHOUSE_API_KEY not set. Web search functionality will be limited.")
 
 ##############################################################################
-# SERPHouse API key from environment variable
+# File Processing Constants
 ##############################################################################
-SERPHOUSE_API_KEY = os.getenv("SERPHOUSE_API_KEY", "")
+MAX_FILE_SIZE = 10 * 1024 * 1024  # 10MB
+MAX_CONTENT_CHARS = 2000
 
 ##############################################################################
-# 간단한 키워드 추출 함수 (한글 + 알파벳 + 숫자 + 공백 보존)
+# Improved Keyword Extraction
 ##############################################################################
 def extract_keywords(text: str, top_k: int = 5) -> str:
     """
-    1) 한글(가-힣), 영어(a-zA-Z), 숫자(0-9), 공백만 남김
-    2) 공백 기준 토큰 분리
-    3) 최대 top_k개만
+    Extract keywords: supports English and Korean
     """
+    stop_words = {'은', '는', '이', '가', '을', '를', '의', '에', '에서', 
+                  'the', 'is', 'at', 'on', 'in', 'a', 'an', 'and', 'or', 'but'}
+    
     text = re.sub(r"[^a-zA-Z0-9가-힣\s]", "", text)
     tokens = text.split()
-    key_tokens = tokens[:top_k]
+    
+    key_tokens = [
+        token for token in tokens 
+        if token.lower() not in stop_words and len(token) > 1
+    ][:top_k]
+    
     return " ".join(key_tokens)
 
 ##############################################################################
-# SerpHouse Live endpoint 호출
-# - 상위 20개 결과 JSON을 LLM에 넘길 때 link, snippet 등 모두 포함
+# File Size Validation
 ##############################################################################
-def do_web_search(query: str) -> str:
+def validate_file_size(file_path: str) -> bool:
+    """Check if file size is within limits"""
+    try:
+        file_size = os.path.getsize(file_path)
+        return file_size <= MAX_FILE_SIZE
+    except:
+        return False
+
+##############################################################################
+# Web Search Function
+##############################################################################
+def do_web_search(query: str, use_korean: bool = False) -> str:
     """
-    상위 20개 'organic' 결과 item 전체(제목, link, snippet 등)를
-    JSON 문자열 형태로 반환
+    Search web and return top 20 organic results
     """
+    if not SERPHOUSE_API_KEY:
+        return "Web search unavailable. API key not configured."
+        
     try:
         url = "https://api.serphouse.com/serp/live"
         
-        # 기본 GET 방식으로 파라미터 간소화하고 결과 수를 20개로 제한
         params = {
             "q": query,
             "domain": "google.com",
-            "serp_type": "web",  # 기본 웹 검색
+            "serp_type": "web",
             "device": "desktop",
-            "lang": "en",
-            "num": "20"  # 최대 20개 결과만 요청
+            "lang": "ko" if use_korean else "en",
+            "num": "20"
         }
         
         headers = {
             "Authorization": f"Bearer {SERPHOUSE_API_KEY}"
         }
         
-        logger.info(f"SerpHouse API 호출 중... 검색어: {query}")
-        logger.info(f"요청 URL: {url} - 파라미터: {params}")
+        logger.info(f"Calling SerpHouse API... Query: {query}")
         
-        # GET 요청 수행
-        response = requests.get(url, headers=headers, params=params, timeout=60)
+        response = requests.get(url, headers=headers, params=params, timeout=30)
         response.raise_for_status()
         
-        logger.info(f"SerpHouse API 응답 상태 코드: {response.status_code}")
         data = response.json()
         
-        # 다양한 응답 구조 처리
+        # Parse results
         results = data.get("results", {})
         organic = None
         
-        # 가능한 응답 구조 1
         if isinstance(results, dict) and "organic" in results:
             organic = results["organic"]
-        
-        # 가능한 응답 구조 2 (중첩된 results)
         elif isinstance(results, dict) and "results" in results:
             if isinstance(results["results"], dict) and "organic" in results["results"]:
                 organic = results["results"]["organic"]
-        
-        # 가능한 응답 구조 3 (최상위 organic)
         elif "organic" in data:
             organic = data["organic"]
             
         if not organic:
-            logger.warning("응답에서 organic 결과를 찾을 수 없습니다.")
-            logger.debug(f"응답 구조: {list(data.keys())}")
-            if isinstance(results, dict):
-                logger.debug(f"results 구조: {list(results.keys())}")
-            return "No web search results found or unexpected API response structure."
+            return "No search results found or unexpected API response structure."
 
-        # 결과 수 제한 및 컨텍스트 길이 최적화
         max_results = min(20, len(organic))
         limited_organic = organic[:max_results]
         
-        # 결과 형식 개선 - 마크다운 형식으로 출력하여 가독성 향상
         summary_lines = []
         for idx, item in enumerate(limited_organic, start=1):
             title = item.get("title", "No title")
@@ -121,104 +122,132 @@ def do_web_search(query: str) -> str:
             snippet = item.get("snippet", "No description")
             displayed_link = item.get("displayed_link", link)
             
-            # 마크다운 형식 (링크 클릭 가능)
             summary_lines.append(
                 f"### Result {idx}: {title}\n\n"
                 f"{snippet}\n\n"
-                f"**출처**: [{displayed_link}]({link})\n\n"
+                f"**Source**: [{displayed_link}]({link})\n\n"
                 f"---\n"
             )
         
-        # 모델에게 명확한 지침 추가
         instructions = """
-# 웹 검색 결과
-아래는 검색 결과입니다. 질문에 답변할 때 이 정보를 활용하세요:
-1. 각 결과의 제목, 내용, 출처 링크를 참고하세요
-2. 답변에 관련 정보의 출처를 명시적으로 인용하세요 (예: "X 출처에 따르면...")
-3. 응답에 실제 출처 링크를 포함하세요
-4. 여러 출처의 정보를 종합하여 답변하세요
+# Web Search Results
+Below are the search results. Use this information when answering questions:
+1. Reference the title, content, and source links
+2. Explicitly cite sources in your answer (e.g., "According to source X...")
+3. Include actual source links in your response
+4. Synthesize information from multiple sources
 """
         
         search_results = instructions + "\n".join(summary_lines)
-        logger.info(f"검색 결과 {len(limited_organic)}개 처리 완료")
         return search_results
     
+    except requests.exceptions.Timeout:
+        logger.error("Web search timeout")
+        return "Web search timed out. Please try again."
+    except requests.exceptions.RequestException as e:
+        logger.error(f"Web search network error: {e}")
+        return "Network error during web search."
     except Exception as e:
         logger.error(f"Web search failed: {e}")
         return f"Web search failed: {str(e)}"
 
-
-##############################################################################
-# 모델/프로세서 로딩
-##############################################################################
-MAX_CONTENT_CHARS = 2000
-MAX_INPUT_LENGTH = 2096  # 최대 입력 토큰 수 제한 추가
-model_id = os.getenv("MODEL_ID", "VIDraft/Gemma-3-R1984-27B")
-
-processor = AutoProcessor.from_pretrained(model_id, padding_side="left")
-model = Gemma3ForConditionalGeneration.from_pretrained(
-    model_id,
-    device_map="auto",
-    torch_dtype=torch.bfloat16,
-    attn_implementation="eager"  # 가능하다면 "flash_attention_2"로 변경
-)
-MAX_NUM_IMAGES = int(os.getenv("MAX_NUM_IMAGES", "5"))
-
-
 ##############################################################################
-# CSV, TXT, PDF 분석 함수
+# File Analysis Functions
 ##############################################################################
 def analyze_csv_file(path: str) -> str:
-    """
-    CSV 파일을 전체 문자열로 변환. 너무 길 경우 일부만 표시.
-    """
+    """Analyze CSV file with size validation and encoding handling"""
+    if not validate_file_size(path):
+        return f"⚠️ Error: File size exceeds {MAX_FILE_SIZE/1024/1024:.1f}MB limit."
+        
     try:
-        df = pd.read_csv(path)
-        if df.shape[0] > 50 or df.shape[1] > 10:
-            df = df.iloc[:50, :10]
+        encodings = ['utf-8', 'cp949', 'euc-kr', 'latin-1']
+        df = None
+        
+        for encoding in encodings:
+            try:
+                df = pd.read_csv(path, encoding=encoding, nrows=50)
+                break
+            except UnicodeDecodeError:
+                continue
+                
+        if df is None:
+            return f"Failed to read CSV: Unsupported encoding"
+            
+        total_rows = len(pd.read_csv(path, encoding=encoding, usecols=[0]))
+        
+        if df.shape[1] > 10:
+            df = df.iloc[:, :10]
+            
+        summary = f"**Data size**: {total_rows} rows x {df.shape[1]} columns\n"
+        summary += f"**Showing**: Top {min(50, total_rows)} rows\n"
+        summary += f"**Columns**: {', '.join(df.columns)}\n\n"
+        
         df_str = df.to_string()
         if len(df_str) > MAX_CONTENT_CHARS:
             df_str = df_str[:MAX_CONTENT_CHARS] + "\n...(truncated)..."
-        return f"**[CSV File: {os.path.basename(path)}]**\n\n{df_str}"
+            
+        return f"**[CSV File: {os.path.basename(path)}]**\n\n{summary}{df_str}"
     except Exception as e:
-        return f"Failed to read CSV ({os.path.basename(path)}): {str(e)}"
-
+        logger.error(f"CSV read error: {e}")
+        return f"Failed to read CSV file ({os.path.basename(path)}): {str(e)}"
 
 def analyze_txt_file(path: str) -> str:
-    """
-    TXT 파일 전문 읽기. 너무 길면 일부만 표시.
-    """
-    try:
-        with open(path, "r", encoding="utf-8") as f:
-            text = f.read()
-        if len(text) > MAX_CONTENT_CHARS:
-            text = text[:MAX_CONTENT_CHARS] + "\n...(truncated)..."
-        return f"**[TXT File: {os.path.basename(path)}]**\n\n{text}"
-    except Exception as e:
-        return f"Failed to read TXT ({os.path.basename(path)}): {str(e)}"
-
+    """Analyze text file with automatic encoding detection"""
+    if not validate_file_size(path):
+        return f"⚠️ Error: File size exceeds {MAX_FILE_SIZE/1024/1024:.1f}MB limit."
+        
+    encodings = ['utf-8', 'cp949', 'euc-kr', 'latin-1', 'utf-16']
+    
+    for encoding in encodings:
+        try:
+            with open(path, "r", encoding=encoding) as f:
+                text = f.read()
+            
+            file_size = os.path.getsize(path)
+            size_info = f"**File size**: {file_size/1024:.1f}KB\n\n"
+            
+            if len(text) > MAX_CONTENT_CHARS:
+                text = text[:MAX_CONTENT_CHARS] + "\n...(truncated)..."
+                
+            return f"**[TXT File: {os.path.basename(path)}]**\n\n{size_info}{text}"
+        except UnicodeDecodeError:
+            continue
+    
+    return f"Failed to read text file ({os.path.basename(path)}): Unsupported encoding"
 
 def pdf_to_markdown(pdf_path: str) -> str:
-    """
-    PDF 텍스트를 Markdown으로 변환. 페이지별로 간단히 텍스트 추출.
-    """
+    """Convert PDF to markdown with improved error handling"""
+    if not validate_file_size(pdf_path):
+        return f"⚠️ Error: File size exceeds {MAX_FILE_SIZE/1024/1024:.1f}MB limit."
+        
     text_chunks = []
     try:
         with open(pdf_path, "rb") as f:
             reader = PyPDF2.PdfReader(f)
-            max_pages = min(5, len(reader.pages))
+            total_pages = len(reader.pages)
+            max_pages = min(5, total_pages)
+            
+            text_chunks.append(f"**Total pages**: {total_pages}")
+            text_chunks.append(f"**Showing**: First {max_pages} pages\n")
+            
             for page_num in range(max_pages):
-                page = reader.pages[page_num]
-                page_text = page.extract_text() or ""
-                page_text = page_text.strip()
-                if page_text:
-                    if len(page_text) > MAX_CONTENT_CHARS // max_pages:
-                        page_text = page_text[:MAX_CONTENT_CHARS // max_pages] + "...(truncated)"
-                    text_chunks.append(f"## Page {page_num+1}\n\n{page_text}\n")
-            if len(reader.pages) > max_pages:
-                text_chunks.append(f"\n...(Showing {max_pages} of {len(reader.pages)} pages)...")
+                try:
+                    page = reader.pages[page_num]
+                    page_text = page.extract_text() or ""
+                    page_text = page_text.strip()
+                    
+                    if page_text:
+                        if len(page_text) > MAX_CONTENT_CHARS // max_pages:
+                            page_text = page_text[:MAX_CONTENT_CHARS // max_pages] + "...(truncated)"
+                        text_chunks.append(f"## Page {page_num+1}\n\n{page_text}\n")
+                except Exception as e:
+                    text_chunks.append(f"## Page {page_num+1}\n\nFailed to read page: {str(e)}\n")
+                    
+            if total_pages > max_pages:
+                text_chunks.append(f"\n...({max_pages}/{total_pages} pages shown)...")
     except Exception as e:
-        return f"Failed to read PDF ({os.path.basename(pdf_path)}): {str(e)}"
+        logger.error(f"PDF read error: {e}")
+        return f"Failed to read PDF file ({os.path.basename(pdf_path)}): {str(e)}"
 
     full_text = "\n".join(text_chunks)
     if len(full_text) > MAX_CONTENT_CHARS:
@@ -226,365 +255,256 @@ def pdf_to_markdown(pdf_path: str) -> str:
 
     return f"**[PDF File: {os.path.basename(pdf_path)}]**\n\n{full_text}"
 
-
 ##############################################################################
-# 이미지/비디오 업로드 제한 검사
+# File Type Check Functions
 ##############################################################################
-def count_files_in_new_message(paths: list[str]) -> tuple[int, int]:
-    image_count = 0
-    video_count = 0
-    for path in paths:
-        if path.endswith(".mp4"):
-            video_count += 1
-        elif re.search(r"\.(png|jpg|jpeg|gif|webp)$", path, re.IGNORECASE):
-            image_count += 1
-    return image_count, video_count
-
-
-def count_files_in_history(history: list[dict]) -> tuple[int, int]:
-    image_count = 0
-    video_count = 0
-    for item in history:
-        if item["role"] != "user" or isinstance(item["content"], str):
-            continue
-        if isinstance(item["content"], list) and len(item["content"]) > 0:
-            file_path = item["content"][0]
-            if isinstance(file_path, str):
-                if file_path.endswith(".mp4"):
-                    video_count += 1
-                elif re.search(r"\.(png|jpg|jpeg|gif|webp)$", file_path, re.IGNORECASE):
-                    image_count += 1
-    return image_count, video_count
-
-
-def validate_media_constraints(message: dict, history: list[dict]) -> bool:
-    media_files = []
-    for f in message["files"]:
-        if re.search(r"\.(png|jpg|jpeg|gif|webp)$", f, re.IGNORECASE) or f.endswith(".mp4"):
-            media_files.append(f)
-
-    new_image_count, new_video_count = count_files_in_new_message(media_files)
-    history_image_count, history_video_count = count_files_in_history(history)
-    image_count = history_image_count + new_image_count
-    video_count = history_video_count + new_video_count
-
-    if video_count > 1:
-        gr.Warning("Only one video is supported.")
-        return False
-    if video_count == 1:
-        if image_count > 0:
-            gr.Warning("Mixing images and videos is not allowed.")
-            return False
-        if "<image>" in message["text"]:
-            gr.Warning("Using <image> tags with video files is not supported.")
-            return False
-    if video_count == 0 and image_count > MAX_NUM_IMAGES:
-        gr.Warning(f"You can upload up to {MAX_NUM_IMAGES} images.")
-        return False
-    
-    if "<image>" in message["text"]:
-        image_files = [f for f in message["files"] if re.search(r"\.(png|jpg|jpeg|gif|webp)$", f, re.IGNORECASE)]
-        image_tag_count = message["text"].count("<image>")
-        if image_tag_count != len(image_files):
-            gr.Warning("The number of <image> tags in the text does not match the number of image files.")
-            return False
-
-    return True
-
-
-##############################################################################
-# 비디오 처리 - 임시 파일 추적 코드 추가
-##############################################################################
-def downsample_video(video_path: str) -> list[tuple[Image.Image, float]]:
-    vidcap = cv2.VideoCapture(video_path)
-    fps = vidcap.get(cv2.CAP_PROP_FPS)
-    total_frames = int(vidcap.get(cv2.CAP_PROP_FRAME_COUNT))
-    frame_interval = max(int(fps), int(total_frames / 10))
-    frames = []
-
-    for i in range(0, total_frames, frame_interval):
-        vidcap.set(cv2.CAP_PROP_POS_FRAMES, i)
-        success, image = vidcap.read()
-        if success:
-            image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
-            # 이미지 크기 줄이기 추가
-            image = cv2.resize(image, (0, 0), fx=0.5, fy=0.5)
-            pil_image = Image.fromarray(image)
-            timestamp = round(i / fps, 2)
-            frames.append((pil_image, timestamp))
-            if len(frames) >= 5:
-                break
-
-    vidcap.release()
-    return frames
-
+def is_image_file(file_path: str) -> bool:
+    """Check if file is an image"""
+    return bool(re.search(r"\.(png|jpg|jpeg|gif|webp)$", file_path, re.IGNORECASE))
 
-def process_video(video_path: str) -> tuple[list[dict], list[str]]:
-    content = []
-    temp_files = []  # 임시 파일 추적을 위한 리스트
-    
-    frames = downsample_video(video_path)
-    for frame in frames:
-        pil_image, timestamp = frame
-        with tempfile.NamedTemporaryFile(delete=False, suffix=".png") as temp_file:
-            pil_image.save(temp_file.name)
-            temp_files.append(temp_file.name)  # 추적을 위해 경로 저장
-            content.append({"type": "text", "text": f"Frame {timestamp}:"})
-            content.append({"type": "image", "url": temp_file.name})
-    
-    return content, temp_files
+def is_video_file(file_path: str) -> bool:
+    """Check if file is a video"""
+    return bool(re.search(r"\.(mp4|avi|mov|mkv)$", file_path, re.IGNORECASE))
 
+def is_document_file(file_path: str) -> bool:
+    """Check if file is a document"""
+    return bool(re.search(r"\.(pdf|csv|txt)$", file_path, re.IGNORECASE))
 
 ##############################################################################
-# interleaved <image> 처리
+# Message Processing Functions
 ##############################################################################
-def process_interleaved_images(message: dict) -> list[dict]:
-    parts = re.split(r"(<image>)", message["text"])
-    content = []
-    image_index = 0
+def process_new_user_message(message: dict) -> str:
+    """Process user message and convert to text"""
+    content_parts = [message["text"]]
     
-    image_files = [f for f in message["files"] if re.search(r"\.(png|jpg|jpeg|gif|webp)$", f, re.IGNORECASE)]
+    if not message.get("files"):
+        return message["text"]
+
+    # Classify files
+    csv_files = []
+    txt_files = []
+    pdf_files = []
+    image_files = []
+    video_files = []
+    unknown_files = []
     
-    for part in parts:
-        if part == "<image>" and image_index < len(image_files):
-            content.append({"type": "image", "url": image_files[image_index]})
-            image_index += 1
-        elif part.strip():
-            content.append({"type": "text", "text": part.strip()})
+    for file_path in message["files"]:
+        if file_path.lower().endswith(".csv"):
+            csv_files.append(file_path)
+        elif file_path.lower().endswith(".txt"):
+            txt_files.append(file_path)
+        elif file_path.lower().endswith(".pdf"):
+            pdf_files.append(file_path)
+        elif is_image_file(file_path):
+            image_files.append(file_path)
+        elif is_video_file(file_path):
+            video_files.append(file_path)
         else:
-            if isinstance(part, str) and part != "<image>":
-                content.append({"type": "text", "text": part})
-    return content
-
-
-##############################################################################
-# PDF + CSV + TXT + 이미지/비디오
-##############################################################################
-def is_image_file(file_path: str) -> bool:
-    return bool(re.search(r"\.(png|jpg|jpeg|gif|webp)$", file_path, re.IGNORECASE))
-
-def is_video_file(file_path: str) -> bool:
-    return file_path.endswith(".mp4")
-
-def is_document_file(file_path: str) -> bool:
-    return (
-        file_path.lower().endswith(".pdf")
-        or file_path.lower().endswith(".csv")
-        or file_path.lower().endswith(".txt")
-    )
-
-
-def process_new_user_message(message: dict) -> tuple[list[dict], list[str]]:
-    temp_files = []  # 임시 파일 추적용 리스트
+            unknown_files.append(file_path)
     
-    if not message["files"]:
-        return [{"type": "text", "text": message["text"]}], temp_files
-
-    video_files = [f for f in message["files"] if is_video_file(f)]
-    image_files = [f for f in message["files"] if is_image_file(f)]
-    csv_files = [f for f in message["files"] if f.lower().endswith(".csv")]
-    txt_files = [f for f in message["files"] if f.lower().endswith(".txt")]
-    pdf_files = [f for f in message["files"] if f.lower().endswith(".pdf")]
-
-    content_list = [{"type": "text", "text": message["text"]}]
-
+    # Process document files
     for csv_path in csv_files:
         csv_analysis = analyze_csv_file(csv_path)
-        content_list.append({"type": "text", "text": csv_analysis})
+        content_parts.append(csv_analysis)
 
     for txt_path in txt_files:
         txt_analysis = analyze_txt_file(txt_path)
-        content_list.append({"type": "text", "text": txt_analysis})
+        content_parts.append(txt_analysis)
 
     for pdf_path in pdf_files:
         pdf_markdown = pdf_to_markdown(pdf_path)
-        content_list.append({"type": "text", "text": pdf_markdown})
-
+        content_parts.append(pdf_markdown)
+    
+    # Warning messages for unsupported files
+    if image_files:
+        image_names = [os.path.basename(f) for f in image_files]
+        content_parts.append(
+            f"\n⚠️ **Image files detected**: {', '.join(image_names)}\n"
+            "This demo currently does not support image analysis. "
+            "Please describe the image content in text if you need help with it."
+        )
+    
     if video_files:
-        video_content, video_temp_files = process_video(video_files[0])
-        content_list += video_content
-        temp_files.extend(video_temp_files)
-        return content_list, temp_files
-
-    if "<image>" in message["text"] and image_files:
-        interleaved_content = process_interleaved_images({"text": message["text"], "files": image_files})
-        if content_list and content_list[0]["type"] == "text":
-            content_list = content_list[1:]
-        return interleaved_content + content_list, temp_files
-    else:
-        for img_path in image_files:
-            content_list.append({"type": "image", "url": img_path})
-
-    return content_list, temp_files
-
+        video_names = [os.path.basename(f) for f in video_files]
+        content_parts.append(
+            f"\n⚠️ **Video files detected**: {', '.join(video_names)}\n"
+            "This demo currently does not support video analysis. "
+            "Please describe the video content in text if you need help with it."
+        )
+        
+    if unknown_files:
+        unknown_names = [os.path.basename(f) for f in unknown_files]
+        content_parts.append(
+            f"\n⚠️ **Unsupported file format**: {', '.join(unknown_names)}\n"
+            "Supported formats: PDF, CSV, TXT"
+        )
+    
+    return "\n\n".join(content_parts)
 
-##############################################################################
-# history -> LLM 메시지 변환
-##############################################################################
 def process_history(history: list[dict]) -> list[dict]:
+    """Convert conversation history to Friendli API format"""
     messages = []
-    current_user_content: list[dict] = []
+    
     for item in history:
         if item["role"] == "assistant":
-            if current_user_content:
-                messages.append({"role": "user", "content": current_user_content})
-                current_user_content = []
-            messages.append({"role": "assistant", "content": [{"type": "text", "text": item["content"]}]})
-        else:
+            messages.append({
+                "role": "assistant",
+                "content": item["content"]
+            })
+        else:  # user
             content = item["content"]
             if isinstance(content, str):
-                current_user_content.append({"type": "text", "text": content})
+                messages.append({
+                    "role": "user",
+                    "content": content
+                })
             elif isinstance(content, list) and len(content) > 0:
-                file_path = content[0]
-                if is_image_file(file_path):
-                    current_user_content.append({"type": "image", "url": file_path})
-                else:
-                    current_user_content.append({"type": "text", "text": f"[File: {os.path.basename(file_path)}]"})
-
-    if current_user_content:
-        messages.append({"role": "user", "content": current_user_content})
-        
+                # File processing
+                file_info = []
+                for file_path in content:
+                    if isinstance(file_path, str):
+                        file_info.append(f"[File: {os.path.basename(file_path)}]")
+                if file_info:
+                    messages.append({
+                        "role": "user",
+                        "content": " ".join(file_info)
+                    })
+    
     return messages
 
-
 ##############################################################################
-# 모델 생성 함수에서 OOM 캐치
+# Streaming Response Handler
 ##############################################################################
-def _model_gen_with_oom_catch(**kwargs):
-    """
-    별도 스레드에서 OutOfMemoryError를 잡아주기 위해
-    """
+def stream_friendli_response(messages: list[dict], max_tokens: int = 1000) -> Iterator[str]:
+    """Get streaming response from Friendli AI API"""
+    headers = {
+        "Authorization": f"Bearer {FRIENDLI_TOKEN}",
+        "Content-Type": "application/json"
+    }
+    
+    payload = {
+        "model": FRIENDLI_MODEL_ID,
+        "messages": messages,
+        "max_tokens": max_tokens,
+        "top_p": 0.8,
+        "temperature": 0.7,
+        "stream": True,
+        "stream_options": {
+            "include_usage": True
+        }
+    }
+    
     try:
-        model.generate(**kwargs)
-    except torch.cuda.OutOfMemoryError:
-        raise RuntimeError(
-            "[OutOfMemoryError] GPU 메모리가 부족합니다. "
-            "Max New Tokens을 줄이거나, 프롬프트 길이를 줄여주세요."
+        response = requests.post(
+            FRIENDLI_API_URL,
+            headers=headers,
+            json=payload,
+            stream=True,
+            timeout=60
         )
-    finally:
-        # 생성 완료 후 한번 더 캐시 비우기
-        clear_cuda_cache()
-
+        response.raise_for_status()
+        
+        full_response = ""
+        for line in response.iter_lines():
+            if line:
+                line_text = line.decode('utf-8')
+                if line_text.startswith("data: "):
+                    data_str = line_text[6:]
+                    if data_str == "[DONE]":
+                        break
+                    
+                    try:
+                        data = json.loads(data_str)
+                        if "choices" in data and len(data["choices"]) > 0:
+                            delta = data["choices"][0].get("delta", {})
+                            content = delta.get("content", "")
+                            if content:
+                                full_response += content
+                                yield full_response
+                    except json.JSONDecodeError:
+                        logger.warning(f"JSON parsing failed: {data_str}")
+                        continue
+                        
+    except requests.exceptions.Timeout:
+        yield "⚠️ Response timeout. Please try again."
+    except requests.exceptions.RequestException as e:
+        logger.error(f"Friendli API network error: {e}")
+        yield f"⚠️ Network error occurred: {str(e)}"
+    except Exception as e:
+        logger.error(f"Friendli API error: {str(e)}")
+        yield f"⚠️ API call error: {str(e)}"
 
 ##############################################################################
-# 메인 추론 함수 (web search 체크 시 자동 키워드추출->검색->결과 system msg)
+# Main Inference Function
 ##############################################################################
-@spaces.GPU(duration=120)
+
 def run(
     message: dict,
     history: list[dict],
-    system_prompt: str = "",
     max_new_tokens: int = 512,
     use_web_search: bool = False,
-    web_search_query: str = "",
+    use_korean: bool = False,
+    system_prompt: str = "",
 ) -> Iterator[str]:
-
-    if not validate_media_constraints(message, history):
-        yield ""
-        return
-
-    temp_files = []  # 임시 파일 추적용
     
     try:
-        combined_system_msg = ""
-
-        # 내부적으로만 사용 (UI에서는 보이지 않음)
+        # Prepare system message
+        messages = []
+        
+        if use_korean:
+            combined_system_msg = "너는 AI 어시스턴트 역할이다. 한국어로 친절하고 정확하게 답변해라."
+        else:
+            combined_system_msg = "You are an AI assistant. Please respond helpfully and accurately in English."
+        
         if system_prompt.strip():
-            combined_system_msg += f"[System Prompt]\n{system_prompt.strip()}\n\n"
-
+            combined_system_msg += f"\n\n{system_prompt.strip()}"
+        
+        # Web search processing
         if use_web_search:
-            user_text = message["text"]
-            ws_query = extract_keywords(user_text, top_k=5)
-            if ws_query.strip():
-                logger.info(f"[Auto WebSearch Keyword] {ws_query!r}")
-                ws_result = do_web_search(ws_query)
-                combined_system_msg += f"[Search top-20 Full Items Based on user prompt]\n{ws_result}\n\n"
-                # >>> 추가된 안내 문구 (검색 결과의 link 등 출처를 활용)
-                combined_system_msg += "[참고: 위 검색결과 내용과 link를 출처로 인용하여 답변해 주세요.]\n\n"
-                combined_system_msg += """
-[중요 지시사항]
-1. 답변에 검색 결과에서 찾은 정보의 출처를 반드시 인용하세요.
-2. 출처 인용 시 "[출처 제목](링크)" 형식의 마크다운 링크를 사용하세요.
-3. 여러 출처의 정보를 종합하여 답변하세요.
-4. 답변 마지막에 "참고 자료:" 섹션을 추가하고 사용한 주요 출처 링크를 나열하세요.
-"""
-            else:
-                combined_system_msg += "[No valid keywords found, skipping WebSearch]\n\n"
-
-        messages = []
-        if combined_system_msg.strip():
-            messages.append({
-                "role": "system",
-                "content": [{"type": "text", "text": combined_system_msg.strip()}],
-            })
-
-        messages.extend(process_history(history))
-
-        user_content, user_temp_files = process_new_user_message(message)
-        temp_files.extend(user_temp_files)  # 임시 파일 추적
+            user_text = message.get("text", "")
+            if user_text:
+                ws_query = extract_keywords(user_text, top_k=5)
+                if ws_query.strip():
+                    logger.info(f"[Auto web search keywords] {ws_query!r}")
+                    ws_result = do_web_search(ws_query, use_korean=use_korean)
+                    if not ws_result.startswith("Web search"):
+                        combined_system_msg += f"\n\n[Search Results]\n{ws_result}"
+                        if use_korean:
+                            combined_system_msg += "\n\n[중요: 답변에 검색 결과의 출처를 반드시 인용하세요]"
+                        else:
+                            combined_system_msg += "\n\n[Important: Always cite sources from search results in your answer]"
         
-        for item in user_content:
-            if item["type"] == "text" and len(item["text"]) > MAX_CONTENT_CHARS:
-                item["text"] = item["text"][:MAX_CONTENT_CHARS] + "\n...(truncated)..."
-        messages.append({"role": "user", "content": user_content})
-
-        inputs = processor.apply_chat_template(
-            messages,
-            add_generation_prompt=True,
-            tokenize=True,
-            return_dict=True,
-            return_tensors="pt",
-        ).to(device=model.device, dtype=torch.bfloat16)
+        messages.append({
+            "role": "system",
+            "content": combined_system_msg
+        })
         
-        # 입력 토큰 수 제한 추가
-        if inputs.input_ids.shape[1] > MAX_INPUT_LENGTH:
-            inputs.input_ids = inputs.input_ids[:, -MAX_INPUT_LENGTH:]
-            if 'attention_mask' in inputs:
-                inputs.attention_mask = inputs.attention_mask[:, -MAX_INPUT_LENGTH:]
+        # Add conversation history
+        messages.extend(process_history(history))
         
-        streamer = TextIteratorStreamer(processor, timeout=30.0, skip_prompt=True, skip_special_tokens=True)
-        gen_kwargs = dict(
-            inputs,
-            streamer=streamer,
-            max_new_tokens=max_new_tokens,
-        )
-
-        t = Thread(target=_model_gen_with_oom_catch, kwargs=gen_kwargs)
-        t.start()
-
-        output = ""
-        for new_text in streamer:
-            output += new_text
-            yield output
-
-    except Exception as e:
-        logger.error(f"Error in run: {str(e)}")
-        yield f"죄송합니다. 오류가 발생했습니다: {str(e)}"
-    
-    finally:
-        # 임시 파일 삭제
-        for temp_file in temp_files:
-            try:
-                if os.path.exists(temp_file):
-                    os.unlink(temp_file)
-                    logger.info(f"Deleted temp file: {temp_file}")
-            except Exception as e:
-                logger.warning(f"Failed to delete temp file {temp_file}: {e}")
+        # Process current message
+        user_content = process_new_user_message(message)
+        messages.append({
+            "role": "user",
+            "content": user_content
+        })
         
-        # 명시적 메모리 정리
-        try:
-            del inputs, streamer
-        except:
-            pass
+        # Debug log
+        logger.debug(f"Total messages: {len(messages)}")
         
-        clear_cuda_cache()
-
-
+        # Call Friendli API and stream
+        for response_text in stream_friendli_response(messages, max_new_tokens):
+            yield response_text
+            
+    except Exception as e:
+        logger.error(f"run function error: {str(e)}")
+        yield f"⚠️ Sorry, an error occurred: {str(e)}"
 
 ##############################################################################
-# 예시들 (모두 영어로)
+# Examples
 ##############################################################################
 examples = [
+    # PDF comparison example
     [
         {
             "text": "Compare the contents of the two PDF files.",
@@ -594,256 +514,253 @@ examples = [
             ],
         }
     ],
+    # CSV analysis example
     [
         {
             "text": "Summarize and analyze the contents of the CSV file.",
             "files": ["assets/additional-examples/sample-csv.csv"],
         }
     ],
+    # Web search example
     [
         {
-            "text": "Assume the role of a friendly and understanding girlfriend. Describe this video.",
-            "files": ["assets/additional-examples/tmp.mp4"],
-        }
-    ],
-    [
-        {
-            "text": "Describe the cover and read the text on it.",
-            "files": ["assets/additional-examples/maz.jpg"],
-        }
-    ],
-    [
-        {
-            "text": "I already have this supplement <image> and I plan to buy this product <image>. Are there any precautions when taking them together?",
-            "files": ["assets/additional-examples/pill1.png", "assets/additional-examples/pill2.png"],
+            "text": "Explain discord.gg/openfreeai",
+            "files": [],
         }
     ],
+    # Code generation example
     [
         {
-            "text": "Solve this integral.",
-            "files": ["assets/additional-examples/4.png"],
+            "text": "Write Python code to generate Fibonacci sequence.",
+            "files": [],
         }
     ],
-    [
-        {
-            "text": "When was this ticket issued, and what is its price?",
-            "files": ["assets/additional-examples/2.png"],
-        }
-    ],
-    [
-        {
-            "text": "Based on the sequence of these images, create a short story.",
-            "files": [
-                "assets/sample-images/09-1.png",
-                "assets/sample-images/09-2.png",
-                "assets/sample-images/09-3.png",
-                "assets/sample-images/09-4.png",
-                "assets/sample-images/09-5.png",
-            ],
-        }
-    ],
-    [
-        {
-            "text": "Write Python code using matplotlib to plot a bar chart that matches this image.",
-            "files": ["assets/additional-examples/barchart.png"],
-        }
-    ],
-    [
-        {
-            "text": "Read the text in the image and write it out in Markdown format.",
-            "files": ["assets/additional-examples/3.png"],
-        }
-    ],
-    [
-        {
-            "text": "What does this sign say?",
-            "files": ["assets/sample-images/02.png"],
-        }
-    ],
-    [
-        {
-            "text": "Compare the two images and describe their similarities and differences.",
-            "files": ["assets/sample-images/03.png"],
-        }
-    ], 
 ]
 
 ##############################################################################
-# Gradio UI (Blocks) 구성 (좌측 사이드 메뉴 없이 전체화면 채팅)
+# Gradio UI - CSS Styles (Removed blue colors)
 ##############################################################################
 css = """
-/* 1) UI를 처음부터 가장 넓게 (width 100%) 고정하여 표시 */
+/* Full width UI */
 .gradio-container {
-    background: rgba(255, 255, 255, 0.7); /* 배경 투명도 증가 */
+    background: rgba(255, 255, 255, 0.95);
     padding: 30px 40px;
-    margin: 20px auto;  /* 위아래 여백만 유지 */
+    margin: 20px auto;
     width: 100% !important;
-    max-width: none !important; /* 1200px 제한 제거 */
+    max-width: none !important;
+    border-radius: 12px;
+    box-shadow: 0 4px 6px rgba(0, 0, 0, 0.1);
 }
+
 .fillable {
     width: 100% !important; 
     max-width: 100% !important; 
 }
-/* 2) 배경을 완전히 투명하게 변경 */
+
+/* Background */
 body {
-    background: transparent; /* 완전 투명 배경 */
+    background: linear-gradient(135deg, #f5f7fa 0%, #e0e0e0 100%);
     margin: 0;
     padding: 0;
-    font-family: 'Helvetica Neue', Helvetica, Arial, sans-serif;
+    font-family: 'Segoe UI', 'Helvetica Neue', Arial, sans-serif;
     color: #333;
 }
-/* 버튼 색상 완전히 제거하고 투명하게 */
+
+/* Button styles - neutral gray */
 button, .btn {
-    background: transparent !important; /* 색상 완전히 제거 */
-    border: 1px solid #ddd; /* 경계선만 살짝 추가 */
-    color: #333;
-    padding: 12px 24px;
+    background: #6b7280 !important;
+    border: none;
+    color: white !important;
+    padding: 10px 20px;
     text-transform: uppercase;
-    font-weight: bold;
-    letter-spacing: 1px;
+    font-weight: 600;
+    letter-spacing: 0.5px;
     cursor: pointer;
+    border-radius: 6px;
+    transition: all 0.3s ease;
 }
+
 button:hover, .btn:hover {
-    background: rgba(0, 0, 0, 0.05) !important; /* 호버 시 아주 살짝 어둡게만 */
+    background: #4b5563 !important;
+    transform: translateY(-1px);
+    box-shadow: 0 2px 4px rgba(0, 0, 0, 0.2);
 }
 
-/* examples 관련 모든 색상 제거 */
+/* Examples section */
 #examples_container, .examples-container {
-    margin: auto;
+    margin: 20px auto;
     width: 90%;
-    background: transparent !important;
+    background: rgba(255, 255, 255, 0.8);
+    padding: 20px;
+    border-radius: 8px;
 }
+
 #examples_row, .examples-row {
     justify-content: center;
-    background: transparent !important;
 }
 
-/* examples 버튼 내부의 모든 색상 제거 */
+/* Example buttons */
 .gr-samples-table button,
-.gr-samples-table .gr-button,
-.gr-samples-table .gr-sample-btn,
 .gr-examples button,
-.gr-examples .gr-button,
-.gr-examples .gr-sample-btn,
-.examples button,
-.examples .gr-button,
-.examples .gr-sample-btn {
-    background: transparent !important;
-    border: 1px solid #ddd;
-    color: #333;
+.examples button {
+    background: #f0f2f5 !important;
+    border: 1px solid #d1d5db;
+    color: #374151 !important;
+    margin: 5px;
+    font-size: 14px;
 }
 
-/* examples 버튼 호버 시에도 색상 없게 */
 .gr-samples-table button:hover,
-.gr-samples-table .gr-button:hover,
-.gr-samples-table .gr-sample-btn:hover,
 .gr-examples button:hover,
-.gr-examples .gr-button:hover,
-.gr-examples .gr-sample-btn:hover,
-.examples button:hover,
-.examples .gr-button:hover,
-.examples .gr-sample-btn:hover {
-    background: rgba(0, 0, 0, 0.05) !important;
+.examples button:hover {
+    background: #e5e7eb !important;
+    border-color: #9ca3af;
 }
 
-/* 채팅 인터페이스 요소들도 투명하게 */
-.chatbox, .chatbot, .message {
-    background: transparent !important;
+/* Chat interface */
+.chatbox, .chatbot {
+    background: white !important;
+    border-radius: 8px;
+    box-shadow: 0 1px 3px rgba(0, 0, 0, 0.1);
 }
 
-/* 입력창 투명도 조정 */
-.multimodal-textbox, textarea, input {
-    background: rgba(255, 255, 255, 0.5) !important;
+.message {
+    padding: 15px;
+    margin: 10px 0;
+    border-radius: 8px;
 }
 
-/* 모든 컨테이너 요소에 배경색 제거 */
-.container, .wrap, .box, .panel, .gr-panel {
-    background: transparent !important;
+/* Input styles */
+.multimodal-textbox, textarea, input[type="text"] {
+    background: white !important;
+    border: 1px solid #d1d5db;
+    border-radius: 6px;
+    padding: 10px;
+    font-size: 16px;
 }
 
-/* 예제 섹션의 모든 요소에서 배경색 제거 */
-.gr-examples-container, .gr-examples, .gr-sample, .gr-sample-row, .gr-sample-cell {
-    background: transparent !important;
+.multimodal-textbox:focus, textarea:focus, input[type="text"]:focus {
+    border-color: #6b7280;
+    outline: none;
+    box-shadow: 0 0 0 3px rgba(107, 114, 128, 0.1);
 }
-"""
 
-title_html = """
-<h1 align="center" style="margin-bottom: 0.2em; font-size: 1.6em;"> 🤗 Gemma3-R1984-27B </h1>
-<p align="center" style="font-size:1.1em; color:#555;">
-    ✅Agentic AI Platform ✅Reasoning & Uncensored ✅Multimodal & VLM ✅Deep-Research & RAG <br>
-    Operates on an ✅'NVIDIA A100 GPU' as an independent local server, enhancing security and preventing information leakage.<br>
-    @Model Rpository: VIDraft/Gemma-3-R1984-27B, @Based by 'Google Gemma-3-27b', @Powered by 'MOUSE-II'(VIDRAFT)
-</p>
-"""
+/* Warning messages */
+.warning-box {
+    background: #fef3c7 !important;
+    border: 1px solid #f59e0b;
+    border-radius: 8px;
+    padding: 15px;
+    margin: 10px 0;
+    color: #92400e;
+}
 
+/* Headings */
+h1, h2, h3 {
+    color: #1f2937;
+}
 
-with gr.Blocks(css=css, title="Gemma3-R1984-27B") as demo:
-    gr.Markdown(title_html)
+/* Links - neutral gray */
+a {
+    color: #6b7280;
+    text-decoration: none;
+}
 
-    # Display the web search option (while the system prompt and token slider remain hidden)
-    web_search_checkbox = gr.Checkbox(
-        label="Deep Research",
-        value=False
-    )
+a:hover {
+    text-decoration: underline;
+    color: #4b5563;
+}
 
-    # Used internally but not visible to the user
-    system_prompt_box = gr.Textbox(
-        lines=3,
-        value="You are a deep thinking AI that may use extremely long chains of thought to thoroughly analyze the problem and deliberate using systematic reasoning processes to arrive at a correct solution before answering.",
-        visible=False  # hidden from view
-    )
-    
-    max_tokens_slider = gr.Slider(
-        label="Max New Tokens",
-        minimum=100,
-        maximum=8000,
-        step=50,
-        value=1000,
-        visible=False  # hidden from view
-    )
+/* Slider */
+.gr-slider {
+    margin: 15px 0;
+}
+
+/* Checkbox */
+input[type="checkbox"] {
+    width: 18px;
+    height: 18px;
+    margin-right: 8px;
+}
+
+/* Scrollbar */
+::-webkit-scrollbar {
+    width: 8px;
+    height: 8px;
+}
+
+::-webkit-scrollbar-track {
+    background: #f1f1f1;
+}
+
+::-webkit-scrollbar-thumb {
+    background: #888;
+    border-radius: 4px;
+}
+
+::-webkit-scrollbar-thumb:hover {
+    background: #555;
+}
+"""
+
+##############################################################################
+# Gradio UI Main
+##############################################################################
+with gr.Blocks(css=css, title="Gemma-3-R1984-27B Chatbot") as demo:
+    # Title
+    gr.Markdown("# 🤗 Gemma-3-R1984-27B Chatbot")
+    gr.Markdown("Community: [https://discord.gg/openfreeai](https://discord.gg/openfreeai)")
     
-    web_search_text = gr.Textbox(
-        lines=1,
-        label="(Unused) Web Search Query",
-        placeholder="No direct input needed",
-        visible=False  # hidden from view
-    )
+    # UI Components
+    with gr.Row():
+        with gr.Column(scale=2):
+            web_search_checkbox = gr.Checkbox(
+                label="🔍 Enable Deep Research (Web Search)",
+                value=False,
+                info="Check for questions requiring latest information"
+            )
+        with gr.Column(scale=1):
+            korean_checkbox = gr.Checkbox(
+                label="🇰🇷 한글 (Korean)",
+                value=False,
+                info="Check for Korean responses"
+            )
+        with gr.Column(scale=1):
+            max_tokens_slider = gr.Slider(
+                label="Max Tokens",
+                minimum=100,
+                maximum=8000,
+                step=50,
+                value=1000,
+                info="Adjust response length"
+            )
     
-    # Configure the chat interface
+    # Main chat interface
     chat = gr.ChatInterface(
         fn=run,
         type="messages",
-        chatbot=gr.Chatbot(type="messages", scale=1, allow_tags=["image"]),
+        chatbot=gr.Chatbot(type="messages", scale=1),
         textbox=gr.MultimodalTextbox(
             file_types=[
                 ".webp", ".png", ".jpg", ".jpeg", ".gif",
                 ".mp4", ".csv", ".txt", ".pdf"
             ],
             file_count="multiple",
-            autofocus=True
+            autofocus=True,
+            placeholder="Enter text or upload PDF, CSV, TXT files. (Images/videos not supported in this demo)"
         ),
         multimodal=True,
         additional_inputs=[
-            system_prompt_box,
             max_tokens_slider,
             web_search_checkbox,
-            web_search_text,
+            korean_checkbox,
         ],
         stop_btn=False,
-        title='<a href="https://discord.gg/openfreeai" target="_blank">https://discord.gg/openfreeai</a>',
         examples=examples,
         run_examples_on_click=False,
         cache_examples=False,
-        css_paths=None,
         delete_cache=(1800, 1800),
     )
 
-    # Example section - since examples are already set in ChatInterface, this is for display only
-    with gr.Row(elem_id="examples_row"):
-        with gr.Column(scale=12, elem_id="examples_container"):
-            gr.Markdown("### Example Inputs (click to load)")
-
-
 if __name__ == "__main__":
-    # Run locally
-    demo.launch()
+    demo.launch()