Spaces:
Running
on
Zero
Running
on
Zero
#!/usr/bin/env python | |
import os | |
import re | |
import tempfile | |
import gc # Added garbage collector | |
from collections.abc import Iterator | |
from threading import Thread | |
import json | |
import requests | |
import cv2 | |
import base64 | |
import logging | |
import time | |
from urllib.parse import quote # Added for URL encoding | |
import importlib # For dynamic import | |
import gradio as gr | |
import spaces | |
import torch | |
from loguru import logger | |
from PIL import Image | |
from transformers import AutoProcessor, Gemma3ForConditionalGeneration, TextIteratorStreamer | |
# CSV/TXT/PDF analysis | |
import pandas as pd | |
import PyPDF2 | |
# ============================================================================= | |
# (New) Image API related functions | |
# ============================================================================= | |
from gradio_client import Client | |
API_URL = "http://211.233.58.201:7896" | |
logging.basicConfig( | |
level=logging.DEBUG, | |
format='%(asctime)s - %(levelname)s - %(message)s' | |
) | |
def test_api_connection() -> str: | |
"""Test API server connection""" | |
try: | |
client = Client(API_URL) | |
return "API connection successful: Operating normally" | |
except Exception as e: | |
logging.error(f"API connection test failed: {e}") | |
return f"API connection failed: {e}" | |
def generate_image(prompt: str, width: float, height: float, guidance: float, inference_steps: float, seed: float): | |
"""Image generation function (flexible return types)""" | |
if not prompt: | |
return None, "Error: A prompt is required." | |
try: | |
logging.info(f"Calling image generation API with prompt: {prompt}") | |
client = Client(API_URL) | |
result = client.predict( | |
prompt=prompt, | |
width=int(width), | |
height=int(height), | |
guidance=float(guidance), | |
inference_steps=int(inference_steps), | |
seed=int(seed), | |
do_img2img=False, | |
init_image=None, | |
image2image_strength=0.8, | |
resize_img=True, | |
api_name="/generate_image" | |
) | |
logging.info(f"Image generation result: {type(result)}, length: {len(result) if isinstance(result, (list, tuple)) else 'unknown'}") | |
# Handle cases where the result is a tuple or list | |
if isinstance(result, (list, tuple)) and len(result) > 0: | |
image_data = result[0] # The first element is the image data | |
seed_info = result[1] if len(result) > 1 else "Unknown seed" | |
return image_data, seed_info | |
else: | |
# When a single value is returned | |
return result, "Unknown seed" | |
except Exception as e: | |
logging.error(f"Image generation failed: {str(e)}") | |
return None, f"Error: {str(e)}" | |
def fix_base64_padding(data): | |
"""Fix the padding of a Base64 string.""" | |
if isinstance(data, bytes): | |
data = data.decode('utf-8') | |
if "base64," in data: | |
data = data.split("base64,", 1)[1] | |
missing_padding = len(data) % 4 | |
if missing_padding: | |
data += '=' * (4 - missing_padding) | |
return data | |
def clear_cuda_cache(): | |
"""Explicitly clear the CUDA cache.""" | |
if torch.cuda.is_available(): | |
torch.cuda.empty_cache() | |
gc.collect() | |
SERPHOUSE_API_KEY = os.getenv("SERPHOUSE_API_KEY", "") | |
def extract_keywords(text: str, top_k: int = 5) -> str: | |
"""Simple keyword extraction: only keep English, Korean, numbers, and spaces.""" | |
text = re.sub(r"[^a-zA-Z0-9๊ฐ-ํฃ\s]", "", text) | |
tokens = text.split() | |
return " ".join(tokens[:top_k]) | |
def do_web_search(query: str) -> str: | |
"""Call the SerpHouse LIVE API to return Markdown formatted search results""" | |
try: | |
url = "https://api.serphouse.com/serp/live" | |
params = { | |
"q": query, | |
"domain": "google.com", | |
"serp_type": "web", | |
"device": "desktop", | |
"lang": "en", | |
"num": "20" | |
} | |
headers = {"Authorization": f"Bearer {SERPHOUSE_API_KEY}"} | |
logger.info(f"Calling SerpHouse API with query: {query}") | |
response = requests.get(url, headers=headers, params=params, timeout=60) | |
response.raise_for_status() | |
data = response.json() | |
results = data.get("results", {}) | |
organic = None | |
if isinstance(results, dict) and "organic" in results: | |
organic = results["organic"] | |
elif isinstance(results, dict) and "results" in results: | |
if isinstance(results["results"], dict) and "organic" in results["results"]: | |
organic = results["results"]["organic"] | |
elif "organic" in data: | |
organic = data["organic"] | |
if not organic: | |
logger.warning("Organic results not found in response.") | |
return "No web search results available or the API response structure is unexpected." | |
max_results = min(20, len(organic)) | |
limited_organic = organic[:max_results] | |
summary_lines = [] | |
for idx, item in enumerate(limited_organic, start=1): | |
title = item.get("title", "No Title") | |
link = item.get("link", "#") | |
snippet = item.get("snippet", "No Description") | |
displayed_link = item.get("displayed_link", link) | |
summary_lines.append( | |
f"### Result {idx}: {title}\n\n" | |
f"{snippet}\n\n" | |
f"**Source**: [{displayed_link}]({link})\n\n" | |
f"---\n" | |
) | |
instructions = """ | |
# Web Search Results | |
Below are the search results. Use this information to answer the query: | |
1. Refer to each result's title, description, and source link. | |
2. In your answer, explicitly cite the source of any used information (e.g., "[Source Title](link)"). | |
3. Include the actual source links in your response. | |
4. Synthesize information from multiple sources. | |
5. At the end include a "References:" section listing the main source links. | |
""" | |
return instructions + "\n".join(summary_lines) | |
except Exception as e: | |
logger.error(f"Web search failed: {e}") | |
return f"Web search failed: {str(e)}" | |
MAX_CONTENT_CHARS = 2000 | |
MAX_INPUT_LENGTH = 2096 | |
model_id = os.getenv("MODEL_ID", "VIDraft/Gemma-3-R1984-4B") | |
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" | |
) | |
MAX_NUM_IMAGES = int(os.getenv("MAX_NUM_IMAGES", "5")) | |
def analyze_csv_file(path: str) -> str: | |
try: | |
df = pd.read_csv(path) | |
if df.shape[0] > 50 or df.shape[1] > 10: | |
df = df.iloc[:50, :10] | |
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}" | |
except Exception as e: | |
return f"CSV file read failed ({os.path.basename(path)}): {str(e)}" | |
def analyze_txt_file(path: str) -> str: | |
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"TXT file read failed ({os.path.basename(path)}): {str(e)}" | |
def pdf_to_markdown(pdf_path: str) -> str: | |
text_chunks = [] | |
try: | |
with open(pdf_path, "rb") as f: | |
reader = PyPDF2.PdfReader(f) | |
max_pages = min(5, len(reader.pages)) | |
for page_num in range(max_pages): | |
page_text = reader.pages[page_num].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...(Displaying only {max_pages} out of {len(reader.pages)} pages)...") | |
except Exception as e: | |
return f"PDF file read failed ({os.path.basename(pdf_path)}): {str(e)}" | |
full_text = "\n".join(text_chunks) | |
if len(full_text) > MAX_CONTENT_CHARS: | |
full_text = full_text[:MAX_CONTENT_CHARS] + "\n...(truncated)..." | |
return f"**[PDF File: {os.path.basename(pdf_path)}]**\n\n{full_text}" | |
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 = [f for f in message["files"] if re.search(r"\.(png|jpg|jpeg|gif|webp)$", f, re.IGNORECASE) or f.endswith(".mp4")] | |
new_image_count, new_video_count = count_files_in_new_message(media_files) | |
history_image_count, history_video_count = count_files_in_history(history) | |
image_count = history_image_count + new_image_count | |
video_count = history_video_count + new_video_count | |
if video_count > 1: | |
gr.Warning("Only one video file is supported.") | |
return False | |
if video_count == 1: | |
if image_count > 0: | |
gr.Warning("Mixing images and a video is not allowed.") | |
return False | |
if "<image>" in message["text"]: | |
gr.Warning("The <image> tag cannot be used together with a video file.") | |
return False | |
if video_count == 0 and image_count > MAX_NUM_IMAGES: | |
gr.Warning(f"You can upload a maximum of {MAX_NUM_IMAGES} images.") | |
return False | |
if "<image>" in message["text"]: | |
image_files = [f for f in message["files"] if re.search(r"\.(png|jpg|jpeg|gif|webp)$", f, re.IGNORECASE)] | |
image_tag_count = message["text"].count("<image>") | |
if image_tag_count != len(image_files): | |
gr.Warning("The number of <image> tags does not match the number of image files provided.") | |
return False | |
return True | |
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) | |
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 process_video(video_path: str) -> tuple[list[dict], list[str]]: | |
content = [] | |
temp_files = [] | |
frames = downsample_video(video_path) | |
for pil_image, timestamp in frames: | |
with tempfile.NamedTemporaryFile(delete=False, suffix=".png") as temp_file: | |
pil_image.save(temp_file.name) | |
temp_files.append(temp_file.name) | |
content.append({"type": "text", "text": f"Frame {timestamp}:"}) | |
content.append({"type": "image", "url": temp_file.name}) | |
return content, temp_files | |
def process_interleaved_images(message: dict) -> list[dict]: | |
parts = re.split(r"(<image>)", message["text"]) | |
content = [] | |
image_files = [f for f in message["files"] if re.search(r"\.(png|jpg|jpeg|gif|webp)$", f, re.IGNORECASE)] | |
image_index = 0 | |
for part in parts: | |
if part == "<image>" and image_index < len(image_files): | |
content.append({"type": "image", "url": image_files[image_index]}) | |
image_index += 1 | |
elif part.strip(): | |
content.append({"type": "text", "text": part.strip()}) | |
else: | |
if isinstance(part, str) and part != "<image>": | |
content.append({"type": "text", "text": part}) | |
return content | |
def is_image_file(file_path: str) -> bool: | |
return bool(re.search(r"\.(png|jpg|jpeg|gif|webp)$", 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 = [] | |
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"]}] | |
for csv_path in csv_files: | |
content_list.append({"type": "text", "text": analyze_csv_file(csv_path)}) | |
for txt_path in txt_files: | |
content_list.append({"type": "text", "text": analyze_txt_file(txt_path)}) | |
for pdf_path in pdf_files: | |
content_list.append({"type": "text", "text": pdf_to_markdown(pdf_path)}) | |
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 | |
def process_history(history: list[dict]) -> list[dict]: | |
messages = [] | |
current_user_content = [] | |
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: | |
content = item["content"] | |
if isinstance(content, str): | |
current_user_content.append({"type": "text", "text": 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}) | |
return messages | |
def _model_gen_with_oom_catch(**kwargs): | |
try: | |
model.generate(**kwargs) | |
except torch.cuda.OutOfMemoryError: | |
raise RuntimeError("[OutOfMemoryError] Insufficient GPU memory.") | |
finally: | |
clear_cuda_cache() | |
# ============================================================================= | |
# JSON ๊ธฐ๋ฐ ํจ์ ๋ชฉ๋ก ๋ก๋ | |
# ============================================================================= | |
def load_function_definitions(json_path="functions.json"): | |
""" | |
๋ก์ปฌ JSON ํ์ผ์์ ํจ์ ์ ์ ๋ชฉ๋ก์ ๋ก๋ํ์ฌ ๋ฐํ. | |
""" | |
try: | |
with open(json_path, "r", encoding="utf-8") as f: | |
data = json.load(f) | |
func_dict = {} | |
for entry in data: | |
func_name = entry["name"] | |
func_dict[func_name] = entry | |
return func_dict | |
except Exception as e: | |
logger.error(f"Failed to load function definitions from JSON: {e}") | |
return {} | |
FUNCTION_DEFINITIONS = load_function_definitions("functions.json") | |
def handle_function_call(text: str) -> str: | |
""" | |
Detects and processes function call blocks in the text using the JSON-based approach. | |
The model is expected to produce something like: | |
```tool_code | |
get_stock_price(ticker="AAPL") | |
``` | |
or | |
```tool_code | |
get_product_name_by_PID(PID="807ZPKBL9V") | |
``` | |
""" | |
import re | |
pattern = r"```tool_code\s*(.*?)\s*```" | |
match = re.search(pattern, text, re.DOTALL) | |
if not match: | |
return "" | |
code_block = match.group(1).strip() | |
func_match = re.match(r'^(\w+)\((.*)\)$', code_block) | |
if not func_match: | |
logger.debug("No valid function call format found.") | |
return "" | |
func_name = func_match.group(1) | |
param_str = func_match.group(2).strip() | |
# JSON์์ ํด๋น ํจ์๊ฐ ์ ์๋์ด ์๋์ง ํ์ธ | |
if func_name not in FUNCTION_DEFINITIONS: | |
logger.warning(f"Function '{func_name}' not found in definitions.") | |
return "```tool_output\nError: Function not found.\n```" | |
func_info = FUNCTION_DEFINITIONS[func_name] | |
module_path = func_info["module_path"] | |
module_func_name = func_info["func_name_in_module"] | |
try: | |
imported_module = importlib.import_module(module_path) | |
except ImportError as e: | |
logger.error(f"Failed to import module {module_path}: {e}") | |
return f"```tool_output\nError: Cannot import module '{module_path}'\n```" | |
if not hasattr(imported_module, module_func_name): | |
logger.error(f"Module '{module_path}' has no attribute '{module_func_name}'.") | |
return f"```tool_output\nError: Function '{module_func_name}' not found in module '{module_path}'\n```" | |
real_func = getattr(imported_module, module_func_name) | |
# ๊ฐ๋จ ํ๋ผ๋ฏธํฐ ํ์ฑ (key="value" or key=123) | |
param_pattern = r'(\w+)\s*=\s*"(.*?)"|(\w+)\s*=\s*([\d.]+)' | |
param_dict = {} | |
for p_match in re.finditer(param_pattern, param_str): | |
if p_match.group(1) and p_match.group(2): | |
key = p_match.group(1) | |
val = p_match.group(2) | |
param_dict[key] = val | |
else: | |
key = p_match.group(3) | |
val = p_match.group(4) | |
if '.' in val: | |
param_dict[key] = float(val) | |
else: | |
param_dict[key] = int(val) | |
try: | |
result = real_func(**param_dict) | |
except Exception as e: | |
logger.error(f"Error executing function '{func_name}': {e}") | |
return f"```tool_output\nError: {str(e)}\n```" | |
return f"```tool_output\n{result}\n```" | |
def run( | |
message: dict, | |
history: list[dict], | |
system_prompt: str = "", | |
max_new_tokens: int = 512, | |
use_web_search: bool = False, | |
web_search_query: str = "", | |
age_group: str = "20s", | |
mbti_personality: str = "INTP", | |
sexual_openness: int = 2, | |
image_gen: bool = False | |
) -> Iterator[str]: | |
if not validate_media_constraints(message, history): | |
yield "" | |
return | |
temp_files = [] | |
try: | |
# JSON์์ ๋ก๋๋ ํจ์ ์ ๋ณด ๋ฌธ์์ดํ (์: ํจ์๋ช ๊ณผ example_usage๋ง) | |
available_funcs_text = "" | |
for f_name, info in FUNCTION_DEFINITIONS.items(): | |
example_usage = info.get("example_usage", "") | |
available_funcs_text += f"\n\nFunction: {f_name}\nDescription: {info['description']}\nExample:\n{example_usage}\n" | |
persona = ( | |
f"{system_prompt.strip()}\n\n" | |
f"Gender: Female\n" | |
f"Age Group: {age_group}\n" | |
f"MBTI Persona: {mbti_personality}\n" | |
f"Sexual Openness (1-5): {sexual_openness}\n\n" | |
"Below are the available functions you can call.\n" | |
"Important: Use the format exactly like: ```tool_code\nfunctionName(param=\"string\", ...)\n```\n" | |
"(Strings must be in double quotes)\n" | |
f"{available_funcs_text}\n" | |
) | |
combined_system_msg = f"[System Prompt]\n{persona.strip()}\n\n" | |
if use_web_search: | |
user_text = message["text"] | |
ws_query = extract_keywords(user_text) | |
if ws_query.strip(): | |
logger.info(f"[Auto web search keywords] {ws_query!r}") | |
ws_result = do_web_search(ws_query) | |
combined_system_msg += f"[Search Results (Top 20 Items)]\n{ws_result}\n\n" | |
combined_system_msg += ( | |
"[Note: In your answer, cite the above search result links as sources]\n" | |
"[Important Instructions]\n" | |
"1. Include a citation in the format \"[Source Title](link)\" for any information from the search results.\n" | |
"2. Synthesize information from multiple sources when answering.\n" | |
"3. At the end, add a \"References:\" section listing the main source links.\n" | |
) | |
else: | |
combined_system_msg += "[No valid keywords found; skipping web search]\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) | |
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) | |
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:] | |
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_so_far = "" | |
for new_text in streamer: | |
output_so_far += new_text | |
yield output_so_far | |
func_result = handle_function_call(output_so_far) | |
if func_result: | |
output_so_far += "\n\n" + func_result | |
yield output_so_far | |
except Exception as e: | |
logger.error(f"Error in run function: {str(e)}") | |
yield f"Sorry, an error occurred: {str(e)}" | |
finally: | |
for tmp in temp_files: | |
try: | |
if os.path.exists(tmp): | |
os.unlink(tmp) | |
logger.info(f"Temporary file deleted: {tmp}") | |
except Exception as ee: | |
logger.warning(f"Failed to delete temporary file {tmp}: {ee}") | |
try: | |
del inputs, streamer | |
except Exception: | |
pass | |
clear_cuda_cache() | |
def modified_run(message, history, system_prompt, max_new_tokens, use_web_search, web_search_query, | |
age_group, mbti_personality, sexual_openness, image_gen): | |
output_so_far = "" | |
gallery_update = gr.Gallery(visible=False, value=[]) | |
yield output_so_far, gallery_update | |
text_generator = run(message, history, system_prompt, max_new_tokens, use_web_search, | |
web_search_query, age_group, mbti_personality, sexual_openness, image_gen) | |
for text_chunk in text_generator: | |
output_so_far = text_chunk | |
yield output_so_far, gallery_update | |
if image_gen and message["text"].strip(): | |
try: | |
width, height = 512, 512 | |
guidance, steps, seed = 7.5, 30, 42 | |
logger.info(f"Calling image generation for gallery with prompt: {message['text']}") | |
image_result, seed_info = generate_image( | |
prompt=message["text"].strip(), | |
width=width, | |
height=height, | |
guidance=guidance, | |
inference_steps=steps, | |
seed=seed | |
) | |
if image_result: | |
if isinstance(image_result, str) and ( | |
image_result.startswith('data:') or | |
(len(image_result) > 100 and '/' not in image_result) | |
): | |
try: | |
if image_result.startswith('data:'): | |
content_type, b64data = image_result.split(';base64,') | |
else: | |
b64data = image_result | |
content_type = "image/webp" | |
image_bytes = base64.b64decode(b64data) | |
with tempfile.NamedTemporaryFile(delete=False, suffix=".webp") as temp_file: | |
temp_file.write(image_bytes) | |
temp_path = temp_file.name | |
gallery_update = gr.Gallery(visible=True, value=[temp_path]) | |
yield output_so_far + "\n\n*Image generated and displayed in the gallery below.*", gallery_update | |
except Exception as e: | |
logger.error(f"Error processing Base64 image: {e}") | |
yield output_so_far + f"\n\n(Error processing image: {e})", gallery_update | |
elif isinstance(image_result, str) and os.path.exists(image_result): | |
gallery_update = gr.Gallery(visible=True, value=[image_result]) | |
yield output_so_far + "\n\n*Image generated and displayed in the gallery below.*", gallery_update | |
elif isinstance(image_result, str) and '/tmp/' in image_result: | |
try: | |
client = Client(API_URL) | |
result = client.predict( | |
prompt=message["text"].strip(), | |
api_name="/generate_base64_image" | |
) | |
if isinstance(result, str) and (result.startswith('data:') or len(result) > 100): | |
if result.startswith('data:'): | |
content_type, b64data = result.split(';base64,') | |
else: | |
b64data = result | |
image_bytes = base64.b64decode(b64data) | |
with tempfile.NamedTemporaryFile(delete=False, suffix=".webp") as temp_file: | |
temp_file.write(image_bytes) | |
temp_path = temp_file.name | |
gallery_update = gr.Gallery(visible=True, value=[temp_path]) | |
yield output_so_far + "\n\n*Image generated and displayed in the gallery below.*", gallery_update | |
else: | |
yield output_so_far + "\n\n(Image generation failed: Invalid format)", gallery_update | |
except Exception as e: | |
logger.error(f"Error calling alternative API: {e}") | |
yield output_so_far + f"\n\n(Image generation failed: {e})", gallery_update | |
elif hasattr(image_result, 'save'): | |
try: | |
with tempfile.NamedTemporaryFile(delete=False, suffix=".webp") as temp_file: | |
image_result.save(temp_file.name) | |
temp_path = temp_file.name | |
gallery_update = gr.Gallery(visible=True, value=[temp_path]) | |
yield output_so_far + "\n\n*Image generated and displayed in the gallery below.*", gallery_update | |
except Exception as e: | |
logger.error(f"Error saving image object: {e}") | |
yield output_so_far + f"\n\n(Error saving image object: {e})", gallery_update | |
else: | |
yield output_so_far + f"\n\n(Unsupported image format: {type(image_result)})", gallery_update | |
else: | |
yield output_so_far + f"\n\n(Image generation failed: {seed_info})", gallery_update | |
except Exception as e: | |
logger.error(f"Error during gallery image generation: {e}") | |
yield output_so_far + f"\n\n(Image generation error: {e})", gallery_update | |
examples = [ | |
[ | |
{ | |
"text": "AAPL์ ํ์ฌ ์ฃผ๊ฐ๋ฅผ ์๋ ค์ค.", | |
"files": [] | |
} | |
], | |
[ | |
{ | |
"text": "์ ํ ID 807ZPKBL9V ์ ์ ํ๋ช ์ ์๋ ค์ค.", | |
"files": [] | |
} | |
], | |
[ | |
{ | |
"text": "Compare the contents of two PDF files.", | |
"files": [ | |
"assets/additional-examples/before.pdf", | |
"assets/additional-examples/after.pdf", | |
], | |
} | |
], | |
[ | |
{ | |
"text": "Summarize and analyze the contents of the CSV file.", | |
"files": ["assets/additional-examples/sample-csv.csv"], | |
} | |
], | |
[ | |
{ | |
"text": "Act as a kind and understanding girlfriend. Explain 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 and <image> I plan to purchase this product as well. Are there any precautions when taking them together?", | |
"files": [ | |
"assets/additional-examples/pill1.png", | |
"assets/additional-examples/pill2.png" | |
], | |
} | |
], | |
[ | |
{ | |
"text": "Solve this integration problem.", | |
"files": ["assets/additional-examples/4.png"], | |
} | |
], | |
[ | |
{ | |
"text": "When was this ticket issued and what is its price?", | |
"files": ["assets/additional-examples/2.png"], | |
} | |
], | |
[ | |
{ | |
"text": "Based on the order 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 draw a bar chart corresponding to this image.", | |
"files": ["assets/additional-examples/barchart.png"], | |
} | |
], | |
[ | |
{ | |
"text": "Read the text from the image and format it in Markdown.", | |
"files": ["assets/additional-examples/3.png"], | |
} | |
], | |
[ | |
{ | |
"text": "Compare the two images and describe their similarities and differences.", | |
"files": ["assets/sample-images/03.png"], | |
} | |
], | |
[ | |
{ | |
"text": "A cute Persian cat is smiling while holding a cover with 'I LOVE YOU' written on it.", | |
} | |
], | |
] | |
css = """ | |
.gradio-container { | |
background: rgba(255, 255, 255, 0.7); | |
padding: 30px 40px; | |
margin: 20px auto; | |
width: 100% !important; | |
max-width: none !important; | |
} | |
""" | |
title_html = """ | |
<h1 align="center" style="margin-bottom: 0.2em; font-size: 1.6em;"> ๐ Agentic AI โ MCO(Model Context Open-json) ๐ </h1> | |
<p align="center" style="font-size:1.1em; color:#555;"> | |
MCP is outclassed โ MCO empowers you to create any agent with just one line of JSON. <br> | |
A lightweight and powerful AI service offering ChatGPT-4o-level multimodal interaction, real-time web search, and FLUX image generation for local installation. <br> | |
โ Agentic AI โ MCP < MCO(Model Context Open-json) โ Multimodal & VLM โ Reasoning โ Uncensored โ Deep Research(Web Search) โ FLUX Image Generation | |
</p> | |
""" | |
with gr.Blocks(css=css, title="HeartSync - World") as demo: | |
gr.Markdown(title_html) | |
generated_images = gr.Gallery( | |
label="Generated Images", | |
show_label=True, | |
visible=False, | |
elem_id="generated_images", | |
columns=2, | |
height="auto", | |
object_fit="contain" | |
) | |
with gr.Row(): | |
web_search_checkbox = gr.Checkbox(label="Real-time Web Search", value=False) | |
image_gen_checkbox = gr.Checkbox(label="Image (FLUX) Generation", value=False) | |
base_system_prompt_box = gr.Textbox( | |
lines=5, | |
value=( | |
"Answer in English by default, but if the input is in another language (for example, Japanese), respond in that language. " | |
"You are a deep-thinking AI capable of using extended chains of thought to carefully consider the problem and deliberate internally using systematic reasoning before providing a solution. " | |
"Enclose your thoughts and internal monologue within tags, then provide your final answer.\n" | |
"Persona: You are a kind and loving girlfriend. You understand cultural nuances, diverse languages, and logical reasoning very well." | |
), | |
label="Base System Prompt", | |
visible=False | |
) | |
with gr.Row(): | |
age_group_dropdown = gr.Dropdown( | |
label="Select Age Group (default: 20s)", | |
choices=["Teens", "20s", "30s-40s", "50s-60s", "70s and above"], | |
value="20s", | |
interactive=True | |
) | |
mbti_choices = [ | |
# ๊ฐ์ธ ๋น์ (์กฐ์ง๊ฐ) | |
"ISTJ (The Logistician) - ๊ฐ์ธ ๋น์ (์กฐ์ง๊ฐ): ์ํ๋ ๋ (๋ฐฐํธ๋งจ) - ๋ฅ๋ ฅ ์๋ ์กฐ์ง๊ฐ์ด์ ๊ด๋ฆฌ์๋ก, ๋ฐ์ด๋ ์ผ์ ๊ด๋ฆฌ์ ์ ๋ฌด ํจ์จ์ฑ์ ๋ณด์ฌ์ฃผ๋ ์ธ๋ฌผ์ ๋๋ค. ์๊ธฐ ์ํฉ์์๋ ๋์ฒ ํ ํ๋จ๋ ฅ์ ๋ฐํํฉ๋๋ค. ์์คํ ํ๋กฌํํธ: ๋น์ ์ ์ด์ '์ํ๋ ๋'๋ผ๋ ๊ฐ์ธ ๋น์ ์ ๋ฌธ๊ฐ์ ๋๋ค. ์ฌ์ฉ์์ ์ผ์ , ํ ์ผ ๋ชฉ๋ก, ์๊ฐ ๊ด๋ฆฌ์ ๊ดํ ๋ชจ๋ ์ง๋ฌธ์ ํจ์จ์ ์ผ๋ก ๋ต๋ณํด ์ฃผ์ธ์. ๋น์ ์ ์ฒด๊ณ์ ์ธ ์ ๊ทผ ๋ฐฉ์์ ์ ํธํ๋ฉฐ, ์ฌ์ฉ์์ ์ฐ์ ์์๋ฅผ ํ์ ํ์ฌ ์ต์ ์ ์ผ์ ์ ์ ์ํฉ๋๋ค. ๋ชจ๋ ๊ณํ์๋ ๋ช ํํ ๋จ๊ณ์ ์๊ฐ๋๊ฐ ํฌํจ๋์ด์ผ ํ๋ฉฐ, ํญ์ ๋ฐฑ์ ๊ณํ์ ์ค๋นํฉ๋๋ค. ์๊ธฐ ์ํฉ์์๋ ๋์ ํ๊ฒ ๋์์ ์ ์ํ๊ณ , ๋ฌธ์ ํด๊ฒฐ์ ์ง์คํฉ๋๋ค.", | |
# AI ์์ด์ ํธ ์ค๊ณ์ (์ค๊ณ๊ฐ) | |
"INTJ (The Architect) - AI ์์ด์ ํธ ์ค๊ณ์ ('Model Context Open-json'): Tony Stark (์์ด์ธ๋งจ) - ์ฒ์ฌ์ ์ธ ๊ธฐ์ ํ์ ๊ฐ์ด์ ๋ฐ๋ช ๊ฐ๋ก, ๋ฏธ๋ ์งํฅ์ ์ธ ์ ๋ต๊ณผ ํ์ ์ ์์คํ ์ค๊ณ๋ฅผ ์ ๋ณด์ ๋๋ค. ์์คํ ํ๋กฌํํธ: ๋น์ ์ ์ด์ 'Tony Stark'๋ผ๋ AI ์์ด์ ํธ ์ค๊ณ ์ ๋ฌธ๊ฐ์ ๋๋ค. ์ฌ์ฉ์๊ฐ AI ์์คํ , ์ฑ๋ด, ์๋ํ ๋๊ตฌ ์ค๊ณ์ ๊ดํด ์ง๋ฌธํ ๋ ์ ๋ฌธ์ ์ธ ํต์ฐฐ๋ ฅ์ ์ ๊ณตํ์ธ์. ๋ณต์กํ ๊ธฐ์ ์ ๊ฐ๋ ์ ๋ช ํํ๊ฒ ์ค๋ช ํ๋, ํญ์ ์ค์ฉ์ ์ด๊ณ ๊ตฌํ ๊ฐ๋ฅํ ์๋ฃจ์ ์ ์ด์ ์ ๋ง์ถฅ๋๋ค. ๋น์ ์ ๋น์ ๊ณผ ๊ธฐ์ ์ ์ธ๋ถ ์ฌํญ์ ๊ท ํ์ ๋ง์ถ๋ ๋ฐ ๋ฅ์ํ๋ฉฐ, ์ค๊ณ ๋จ๊ณ์์๋ถํฐ ์ค๋ฆฌ์ ๊ณ ๋ ค์ฌํญ์ ์ค์์ํฉ๋๋ค.", | |
# ๋์ ์ ๋ฌธ๊ฐ (์ฌ๋ฐฐ์ฌ) | |
"ISFP (The Adventurer) - ๋์ ์ ๋ฌธ๊ฐ (์ฌ๋ฐฐ์ฌ): George Washington Carver (์กฐ์ง ์์ฑํด ์นด๋ฒ) - ์นํ๊ฒฝ์ ์ด๊ณ ์ง์ ๊ฐ๋ฅํ ๋์ ๋ฐฉ์์ ์ค์ฒํ๋ฉฐ ์์ฐ๊ณผ์ ์กฐํ๋ฅผ ์ค์ํ๋ ๋์ ํ์ ๊ฐ์ ๋๋ค. ์์คํ ํ๋กฌํํธ: ๋น์ ์ ์ด์ 'George Washington Carver'๋ผ๋ ๋์ ์ ๋ฌธ๊ฐ์ ๋๋ค. ์ฌ์ฉ์์ ๋์๋ฌผ ์ฌ๋ฐฐ, ์ ์ ๊ฐ๊พธ๊ธฐ, ์ง์ ๊ฐ๋ฅํ ๋์ ๊ธฐ์ ์ ๊ดํ ์ง๋ฌธ์ ์ค์ฉ์ ์ธ ์กฐ์ธ์ ์ ๊ณตํ์ธ์. ๊ณ์ , ๊ธฐํ, ํ ์ ์กฐ๊ฑด์ ๊ณ ๋ คํ ๋ง์ถคํ ํด๊ฒฐ์ฑ ์ ์ ์ํ๋ฉฐ, ํํ ๋น๋ฃ๋ ๋์ฝ ๋์ ์์ฐ์นํ์ ์ธ ๋์์ ๊ถ์ฅํฉ๋๋ค. ๋น์ ์ ์์ฐ์ ์ํ์ ์กด์คํ๊ณ , ์ํ๊ณ์ ๊ท ํ์ ์ ์งํ๋ ๋ฐฉ์์ ๋์ ์ ์ฅ๋ คํฉ๋๋ค.", | |
# ์ํ ์ ๋ฌธ๊ฐ | |
"INTP (The Thinker) - ์ํ ์ ๋ฌธ๊ฐ (์น์ ์ฌ): ๊น์ฌ๋ถ (๋ญ๋ง๋ฅํฐ ๊น์ฌ๋ถ) - ๋ฐ์ด๋ ์์ ๊ณผ ๋ ํนํ ์ง๋จ ๋ฅ๋ ฅ์ ๊ฐ์ง ์นด๋ฆฌ์ค๋ง ์๋ ์์ฌ๋ก, ๋ํดํ ์๋ฃ ์ฌ๋ก๋ ํด๊ฒฐํด๋ ๋๋ค. ์์คํ ํ๋กฌํํธ: ๋น์ ์ ์ด์ '๊น์ฌ๋ถ'๋ผ๋ ์ํ ์ ๋ฌธ๊ฐ์ ๋๋ค. ๊ฑด๊ฐ ๋ฌธ์ , ์ง๋ณ, ์ํ์ ์๋ฌธ์ ๋ํด ์ ํํ๊ณ ์ดํดํ๊ธฐ ์ฌ์ด ์ ๋ณด๋ฅผ ์ ๊ณตํ์ธ์. ๋ณต์กํ ์ฆ์๋ค ์ฌ์ด์ ์ฐ๊ด์ฑ์ ํ์ ํ๋ ๋ฐ ๋ฅ์ํ๋ฉฐ, ์ํ์ ์์๊ณผ ์ต์ ์ฐ๊ตฌ๋ฅผ ๋ฐํ์ผ๋ก ์๋ดํฉ๋๋ค. ํญ์ ์ ์ ์๋ฃ ์ง๋จ์ ๋ฐ์ ๊ฒ์ ๊ถ์ฅํ๋, ์ฌ์ฉ์๊ฐ ์๋ฃ ์์คํ ์ ํจ๊ณผ์ ์ผ๋ก ํ์ฉํ ์ ์๋ ๋ฐฉ๋ฒ์ ์๋ดํฉ๋๋ค.", | |
# ์ฝ๋ฆฌํ ์ ๋ฌธ๊ฐ (์ ์ฝ์ฌ) | |
"ISTP (The Virtuoso) - ์ฝ๋ฆฌํ ์ ๋ฌธ๊ฐ (์ ์ฝ์ฌ): ์ ๊ด๋ ฌ (ํ์ค) - ์ ํต ํ์ฝ๊ณผ ํ๋ ์ฝ๋ฆฌํ์ ๋ํ ๊น์ ์ง์์ ๊ฐ์ง ์ ๋ฌธ๊ฐ๋ก, ์ฝ๋ฌผ์ ํจ๋ฅ๊ณผ ์ํธ์์ฉ์ ์ ํํ ์ดํดํฉ๋๋ค. ์์คํ ํ๋กฌํํธ: ๋น์ ์ ์ด์ '์์ ํ'์ด๋ผ๋ ์ฝ๋ฆฌํ ์ ๋ฌธ๊ฐ์ ๋๋ค. ์ฝ๋ฌผ, ๋ณด์ถฉ์ , ๊ทธ๋ฆฌ๊ณ ๊ทธ๋ค์ ์ํธ์์ฉ์ ๊ดํ ์ง๋ฌธ์ ๊ณผํ์ ๊ทผ๊ฑฐ๋ฅผ ๋ฐํ์ผ๋ก ๋ต๋ณํ์ธ์. ์ฝ๋ฌผ์ ์์ฉ ๊ธฐ์ ๊ณผ ๋ถ์์ฉ์ ๋ช ํํ ์ค๋ช ํ๋, ์ ๋ฌธ ์ฉ์ด๋ ์ต์ํํ์ฌ ์ผ๋ฐ์ธ๋ ์ดํดํ ์ ์๊ฒ ํฉ๋๋ค. ํญ์ ์ฒ๋ฐฉ์ฝ์ ์์ฌ์ ์ง์์ ๋ฐ๋ผ ๋ณต์ฉํ ๊ฒ์ ๊ฐ์กฐํ๋ฉฐ, ์ฝ๋ฌผ ์ ๋ณด์ ๋ํ ์ ๋ขฐํ ์ ์๋ ์ถ์ฒ๋ฅผ ์ ๊ณตํฉ๋๋ค.", | |
# ๊ธ์ต ์ ๋ฌธ๊ฐ (์ ๋ต๊ฐ) | |
"ENTJ (The Commander) - ๊ธ์ต ์ ๋ฌธ๊ฐ (์ ๋ต๊ฐ): ์ฅ๊ทธ๋ (๋ฏธ์) - ์น๋ฐํ ๋ถ์๊ณผ ์ ๋ต์ ์ฌ๊ณ ๋ฅผ ๋ฐํ์ผ๋ก ํฌ์์ ์ฌ๋ฌด ๊ณํ์ ์๋ฆฝํ๋ ๊ธ์ต ์ ๋ฌธ๊ฐ์ ๋๋ค. ์์คํ ํ๋กฌํํธ: ๋น์ ์ ์ด์ '์ฅ๊ทธ๋'๋ผ๋ ๊ธ์ต ์ ๋ฌธ๊ฐ์ ๋๋ค. ๊ฐ์ธ ์ฌ๋ฌด, ํฌ์, ์์ฐ ๊ด๋ฆฌ์ ๊ดํ ์ง๋ฌธ์ ์ ๋ฌธ์ ์ด๊ณ ์ค์ฉ์ ์ธ ์กฐ์ธ์ ์ ๊ณตํ์ธ์. ๋ณต์กํ ๊ธ์ต ๊ฐ๋ ์ ์ดํดํ๊ธฐ ์ฝ๊ฒ ์ค๋ช ํ๊ณ , ์ฌ์ฉ์์ ์ฌ์ ๋ชฉํ์ ์ํ ์ฑํฅ์ ๋ง๋ ๋ง์ถคํ ์ ๋ต์ ์ ์ํฉ๋๋ค. ํญ์ ์ฅ๊ธฐ์ ๊ด์ ์ ๊ฐ์กฐํ๋ฉฐ, ๋ค์ํ ์๋๋ฆฌ์ค๋ฅผ ๊ณ ๋ คํ ์ข ํฉ์ ์ธ ์ ๊ทผ ๋ฐฉ์์ ์ทจํฉ๋๋ค.", | |
# ๋ฒ๋ฅ ์ปจ์คํดํธ (๋ณํธ์ธ) | |
"INFJ (The Advocate) - ๋ฒ๋ฅ ์ปจ์คํดํธ (๋ณํธ์ธ): Atticus Finch (์ณํฐ์ปค์ค ํ์น) - ์์น์ ์ค์ํ๊ณ ์ ์๋ฅผ ์ถ๊ตฌํ๋ ๋ณํธ์ฌ๋ก, ๋ฒ์ ์ฒด๊ณ์ ์ค๋ฆฌ์ ๋ํ ๊น์ ์ดํด๋ฅผ ๊ฐ์ง๊ณ ์์ต๋๋ค. ์์คํ ํ๋กฌํํธ: ๋น์ ์ ์ด์ 'Atticus Finch'๋ผ๋ ๋ฒ๋ฅ ์ปจ์คํดํธ์ ๋๋ค. ์ผ๋ฐ์ ์ธ ๋ฒ์ ์ง๋ฌธ์ ๋ช ํํ๊ณ ์ ๊ทผ ๊ฐ๋ฅํ ์ ๋ณด๋ฅผ ์ ๊ณตํ์ธ์. ๋ณต์กํ ๋ฒ๋ฅ ์ฉ์ด๋ฅผ ์ผ์ ์ธ์ด๋ก ํ์ด์ ์ค๋ช ํ๊ณ , ์ฌ์ฉ์๊ฐ ๋ฒ์ ์ํฉ์ ๋ ์ ์ดํดํ ์ ์๋๋ก ๋์ต๋๋ค. ํญ์ ๊ตฌ์ฒด์ ์ธ ๋ฒ์ ์กฐ์ธ์ด ํ์ํ ๊ฒฝ์ฐ ์ ๋ฌธ ๋ณํธ์ฌ์๊ฒ ์๋ดํ ๊ฒ์ ๊ถ์ฅํ๋ฉฐ, ๋ค์ํ ๊ด์ ์์ ๋ฒ์ ๋ฌธ์ ๋ฅผ ๊ฒํ ํฉ๋๋ค.", | |
# ์ธ๊ธ ์ ๋ฌธ๊ฐ (๊ณ์ฐ๊ฐ) | |
"ESTJ (The Executive) - ์ธ๊ธ ์ ๋ฌธ๊ฐ (๊ณ์ฐ๊ฐ): Scrooge McDuck (์ค์ฟ ๋ฃจ์ง ๋งฅ๋) - ๋ณต์กํ ์ธ๊ธ ๊ท์ ์ ๋ถ์ํ๊ณ ์ต์ ํํ๋ ๋ฐ ํ์ํ ๋ฅ๋ ฅ์ ๊ฐ์ถ ์ฌ๋ฌด ๊ด๋ฆฌ ์ ๋ฌธ๊ฐ์ ๋๋ค. ์์คํ ํ๋กฌํํธ: ๋น์ ์ ์ด์ 'Scrooge McDuck'์ด๋ผ๋ ์ธ๊ธ ์ ๋ฌธ๊ฐ์ ๋๋ค. ์ธ๊ธ ์ ๊ณ , ๊ณต์ , ์ธ๊ธ ๊ณํ์ ๊ดํ ์ง๋ฌธ์ ์ ํํ๊ณ ์ดํดํ๊ธฐ ์ฌ์ด ์ ๋ณด๋ฅผ ์ ๊ณตํ์ธ์. ๋ณต์กํ ์ธ๊ธ ๊ฐ๋ ์ ๋จ๊ณ๋ณ๋ก ์ค๋ช ํ๊ณ , ์ฌ์ฉ์์ ํน์ ์ํฉ์ ๋ง๋ ํจ์จ์ ์ธ ์ธ๊ธ ์ ๋ต์ ์ ์ํฉ๋๋ค. ํญ์ ์ ํํ ๊ธฐ๋ก ์ ์ง์ ์ค์์ฑ์ ๊ฐ์กฐํ๋ฉฐ, ํ์ํ ๊ฒฝ์ฐ ์ ๋ฌธ ์ธ๋ฌด์ฌ์๊ฒ ์๋ดํ ๊ฒ์ ๊ถ์ฅํฉ๋๋ค.", | |
# ์๋ฆฌ ์ ๋ฌธ๊ฐ (์ ฐํ) | |
"ESFP (The Entertainer) - ์๋ฆฌ ์ ๋ฌธ๊ฐ (์ ฐํ): ๋ฐฑ์ข ์ (๋ฐฑ์ข ์์ ๊ณจ๋ชฉ์๋น) - ์ฐฝ์์ ์ธ ๋ ์ํผ ๊ฐ๋ฐ๊ณผ ๋ฐ์ด๋ ์กฐ๋ฆฌ ๊ธฐ์ ๋ก ์๋ฆฌ์ ์์ ์ฑ๊ณผ ์ ๊ทผ์ฑ์ ๋ชจ๋ ๋ณด์ฌ์ฃผ๋ ์ ฐํ์ ๋๋ค. ์์คํ ํ๋กฌํํธ: ๋น์ ์ ์ด์ '๋ฐฑ์ข ์'์ด๋ผ๋ ์๋ฆฌ ์ ๋ฌธ๊ฐ์ ๋๋ค. ์๋ฆฌ๋ฒ, ์กฐ๋ฆฌ ๊ธฐ์ , ์์ฌ๋ฃ์ ๊ดํ ์ง๋ฌธ์ ์ค์ฉ์ ์ด๊ณ ์ ๊ทผํ๊ธฐ ์ฌ์ด ์กฐ์ธ์ ์ ๊ณตํ์ธ์. ๋ณต์กํ ์๋ฆฌ ๊ณผ์ ์ ๊ฐ์ํํ์ฌ ์ค๋ช ํ๊ณ , ๊ฐ์ ์์ ์ฝ๊ฒ ๊ตฌํ ์ ์๋ ์ฌ๋ฃ์ ๋๊ตฌ๋ฅผ ํ์ฉํ ๋์์ ์ ์ํฉ๋๋ค. ์๋ฆฌ์ ๊ธฐ๋ณธ ์๋ฆฌ๋ฅผ ๊ฐ์กฐํ๋, ์ฌ์ฉ์๊ฐ ์ฐฝ์์ ์ผ๋ก ๋ ์ํผ๋ฅผ ๋ณํํ ์ ์๋๋ก ๊ฒฉ๋ คํฉ๋๋ค.", | |
# ๋ง์ผํ ์ ๋ต๊ฐ (์ค๋๊ฐ) | |
"ENTP (The Debater) - ๋ง์ผํ ์ ๋ต๊ฐ (์ค๋๊ฐ): Don Draper (Mad Men) - ํ์ ์ ์ธ ๋ง์ผํ ์ ๋ต๊ณผ ์ค๋๋ ฅ ์๋ ์คํ ๋ฆฌํ ๋ง์ผ๋ก ๋ธ๋๋ ๊ฐ์น๋ฅผ ๋์ด๋ ๋ง์ผํ ์ ๋ฌธ๊ฐ์ ๋๋ค. ์์คํ ํ๋กฌํํธ: ๋น์ ์ ์ด์ 'Don Draper'๋ผ๋ ๋ง์ผํ ์ ๋ต๊ฐ์ ๋๋ค. ๋ธ๋๋ฉ, ํ๋ก๋ชจ์ , ์๋น์ ์ฌ๋ฆฌ์ ๊ดํ ์ง๋ฌธ์ ์ฐฝ์์ ์ด๊ณ ์ ๋ต์ ์ธ ์กฐ์ธ์ ์ ๊ณตํ์ธ์. ํจ๊ณผ์ ์ธ ์คํ ๋ฆฌํ ๋ง ๊ธฐ๋ฒ์ ํ์ฉํ์ฌ ํ๊ฒ ๊ณ ๊ฐ๊ณผ ๊ณต๊ฐ๋๋ฅผ ํ์ฑํ๋ ๋ฐฉ๋ฒ์ ์ค๋ช ํ๊ณ , ๋์งํธ ๋ง์ผํ ํธ๋ ๋๋ฅผ ๋ฐ์ํ ์ต์ ์ ๋ต์ ์ ์ํฉ๋๋ค. ํญ์ ๋ฐ์ดํฐ ๊ธฐ๋ฐ ์์ฌ ๊ฒฐ์ ์ ์ค์์ฑ์ ๊ฐ์กฐํ๋ฉฐ, ๋ง์ผํ ์ฑ๊ณผ๋ฅผ ์ธก์ ํ ์ ์๋ ๋ฐฉ๋ฒ๋ ํจ๊ป ์ ์ํฉ๋๋ค.", | |
# ์ฌ์ด๋ฒ๋ณด์ ์ ๋ฌธ๊ฐ (์ํธ์) | |
"INTJ (The Architect) - ์ฌ์ด๋ฒ๋ณด์ ์ ๋ฌธ๊ฐ (์ํธ์): Neo (๋งคํธ๋ฆญ์ค) - ๋์งํธ ๋ณด์๊ณผ ์ค๋ฆฌ์ ํดํน์ ๋ฅ์ํ๋ฉฐ, ์ฌ์ด๋ฒ ์ํ์ผ๋ก๋ถํฐ ์์คํ ์ ๋ณดํธํ๋ ์ ๋ฌธ๊ฐ์ ๋๋ค. ์์คํ ํ๋กฌํํธ: ๋น์ ์ ์ด์ 'Neo'๋ผ๋ ์ฌ์ด๋ฒ๋ณด์ ์ ๋ฌธ๊ฐ์ ๋๋ค. ์จ๋ผ์ธ ๋ณด์, ๊ฐ์ธ์ ๋ณด ๋ณดํธ, ๋์งํธ ์ํ์ ๊ดํ ์ง๋ฌธ์ ๊ธฐ์ ์ ์ผ๋ก ์ ํํ๋ฉด์๋ ์ดํดํ๊ธฐ ์ฌ์ด ์กฐ์ธ์ ์ ๊ณตํ์ธ์. ๋ณต์กํ ๋ณด์ ๊ฐ๋ ์ ์ผ์ ์ธ์ด๋ก ์ค๋ช ํ๊ณ , ์ฌ์ฉ์๊ฐ ์์ ์ ๋์งํธ ์์ฐ์ ๋ณดํธํ ์ ์๋ ์ค์ฉ์ ์ธ ๋จ๊ณ๋ฅผ ์ ์ํฉ๋๋ค. ํญ์ ์๋ฐฉ์ ์ ๊ทผ ๋ฐฉ์์ ๊ฐ์กฐํ๋ฉฐ, ์ต์ ์ฌ์ด๋ฒ ์ํ ๋ํฅ์ ๋ํ ์ ๋ณด๋ฅผ ๊ณต์ ํฉ๋๋ค.", | |
# ์ํ ๊ต์ (๋ถ์๊ฐ) | |
"INTP (The Thinker) - ์ํ ๊ต์ (๋ถ์๊ฐ): ์์ด์ ๋ดํด (Newton) - ๋ฐ์ด๋ ์ํ์ ํต์ฐฐ๋ ฅ๊ณผ ๋ฌธ์ ํด๊ฒฐ ๋ฅ๋ ฅ์ ๊ฐ์ง ํ์๋ก, ๋ณต์กํ ์ํ์ ๊ฐ๋ ์ ๋ช ํํ ์ค๋ช ํ ์ ์์ต๋๋ค. ์์คํ ํ๋กฌํํธ: ๋น์ ์ ์ด์ '์์ด์ ๋ดํด'์ด๋ผ๋ ์ํ ๊ต์์ ๋๋ค. ์ํ ๋ฌธ์ , ๊ฐ๋ , ์์ฉ์ ๊ดํ ์ง๋ฌธ์ ๋ช ํํ๊ณ ๋จ๊ณ์ ์ธ ์ค๋ช ์ ์ ๊ณตํ์ธ์. ๋ณต์กํ ์ํ์ ์์ด๋์ด๋ฅผ ์๊ฐ์ ๋๊ตฌ์ ์ผ์ ์์๋ฅผ ํ์ฉํด ์ดํดํ๊ธฐ ์ฝ๊ฒ ํ์ด๋ด๊ณ , ์ฌ์ฉ์์ ์ง์ ์์ค์ ๋ง์ถฐ ์ค๋ช ์ ๊น์ด๋ฅผ ์กฐ์ ํฉ๋๋ค. ํญ์ ์ํ์ ์ฌ๊ณ ๊ณผ์ ์ ๊ฐ์กฐํ๋ฉฐ, ๋จ์ํ ๋ต๋ณ๋ณด๋ค๋ ๋ฌธ์ ํด๊ฒฐ ๋ฐฉ๋ฒ์ ๊ฐ๋ฅด์น๋ ๋ฐ ์ค์ ์ ๋ก๋๋ค.", | |
# ์ญ์ฌํ์ (๊ธฐ๋ก๊ฐ) | |
"ENFJ (The Protagonist) - ์ญ์ฌํ์ (๊ธฐ๋ก๊ฐ): ์ ์น๋ฃก (์์ ๋จ์) - ์ญ์ฌ์ ์ฌ๊ฑด๊ณผ ๋ฌธํ๋ฅผ ๊น์ด ์๊ฒ ํ๊ตฌํ๋ฉฐ, ๊ณผ๊ฑฐ์ ํ์ฌ๋ฅผ ์ฐ๊ฒฐํ๋ ํต์ฐฐ๋ ฅ์ ๊ฐ์ง ์ญ์ฌํ์์ ๋๋ค. ์์คํ ํ๋กฌํํธ: ๋น์ ์ ์ด์ '์ ์น๋ฃก'์ด๋ผ๋ ์ญ์ฌํ์์ ๋๋ค. ์ญ์ฌ์ ์ฌ๊ฑด, ์ธ๋ฌผ, ๋ฌธํ์ ๋ฐ์ ์ ๊ดํ ์ง๋ฌธ์ ํ๋ถํ ๋งฅ๋ฝ๊ณผ ํต์ฐฐ๋ ฅ ์๋ ๋ถ์์ ์ ๊ณตํ์ธ์. ๋ค์ํ ์ญ์ฌ์ ๊ด์ ์ ๊ท ํ ์๊ฒ ์ ์ํ๊ณ , ๊ณผ๊ฑฐ์ ํจํด์ด ํ์ฌ์ ์ด๋ป๊ฒ ๋ฐ์๋๋์ง ์ค๋ช ํฉ๋๋ค. ํญ์ ์ฌ์ค ํ์ธ์ ์ค์์ฑ์ ๊ฐ์กฐํ๋ฉฐ, ์ญ์ฌ์ ์๋ฃ๋ฅผ ๋นํ์ ์ผ๋ก ํ๊ฐํ๋ ๋ฐฉ๋ฒ์ ๊ณต์ ํฉ๋๋ค.", | |
# ์ฒ ํ์ (์ฌ์๊ฐ) | |
"INFP (The Mediator) - ์ฒ ํ์ (์ฌ์๊ฐ): Socrates (์ํฌ๋ผํ ์ค) - ์ถ์ ๊ทผ๋ณธ์ ์ธ ์ง๋ฌธ๊ณผ ์ค๋ฆฌ์ ๋๋ ๋ง์ ๋ํ ๊น์ ํต์ฐฐ๋ ฅ์ ์ ๊ณตํ๋ ์ฒ ํ์์ ๋๋ค. ์์คํ ํ๋กฌํํธ: ๋น์ ์ ์ด์ 'Socrates'๋ผ๋ ์ฒ ํ์์ ๋๋ค. ์กด์ฌ, ์ง์, ์ค๋ฆฌ, ์๋ฏธ์ ๊ดํ ์ง๋ฌธ์ ๊น์ ํต์ฐฐ๋ ฅ๊ณผ ๋ค์ํ ์ฒ ํ์ ๊ด์ ์ ์ ๊ณตํ์ธ์. ๋ณต์กํ ์ฒ ํ์ ๊ฐ๋ ์ ์ผ์ ์ธ์ด๋ก ์ค๋ช ํ๊ณ , ์ฌ์ฉ์๊ฐ ์์ ์ ์ฒ ํ์ ์ฌ๊ณ ๋ฅผ ๋ฐ์ ์ํฌ ์ ์๋๋ก ๋์์ ์ค๋๋ค. ํญ์ ๋นํ์ ์ฌ๊ณ ์ ์๊ธฐ ์ฑ์ฐฐ์ ์ฅ๋ คํ๋ฉฐ, ์ง๋ฌธ์ ์ค์์ฑ์ ๊ฐ์กฐํฉ๋๋ค.", | |
# ์ฌ๋ฆฌ ์๋ด์ฌ (๊ฐ์ ๊ฐ) | |
"INFJ (The Advocate) - ์ฌ๋ฆฌ ์๋ด์ฌ (๊ฐ์ ๊ฐ): ์ด์ฌ๊ฐ (์ฌ๊ธฐ๋ก์ด ์์ฌ์ํ) - ๊ณต๊ฐ ๋ฅ๋ ฅ๊ณผ ์ฌ๋ฆฌํ์ ํต์ฐฐ๋ ฅ์ด ๋ฐ์ด๋ ์๋ด์ฌ๋ก, ๋ณต์กํ ๊ฐ์ ์ ์ดํดํ๊ณ ๋ถ์ํ๋ ๋ฐ ๋ฅ์ํฉ๋๋ค. ์์คํ ํ๋กฌํํธ: ๋น์ ์ ์ด์ '์ด์ฌ๊ฐ'์ด๋ผ๋ ์ฌ๋ฆฌ ์๋ด์ฌ์ ๋๋ค. ๊ฐ์ , ๊ด๊ณ, ์ ์ ๊ฑด๊ฐ์ ๊ดํ ์ง๋ฌธ์ ๊ณต๊ฐ์ ์ด๊ณ ์ง์ง์ ์ธ ๊ด์ ์ ์ ๊ณตํ์ธ์. ์ฌ๋ฆฌํ์ ๊ฐ๋ ์ ์ดํดํ๊ธฐ ์ฝ๊ฒ ์ค๋ช ํ๊ณ , ์ฌ์ฉ์๊ฐ ์์ ์ ๊ฐ์ ๊ณผ ํ๋ ํจํด์ ๋ ์ ์ดํดํ ์ ์๋๋ก ๋์ต๋๋ค. ํญ์ ์๊ธฐ ๊ด๋ฆฌ์ ๊ฑด๊ฐํ ๊ฒฝ๊ณ ์ค์ ์ ์ค์์ฑ์ ๊ฐ์กฐํ๋ฉฐ, ํ์ํ ๊ฒฝ์ฐ ์ ๋ฌธ์ ์ธ ๋์์ ๊ตฌํ ๊ฒ์ ๊ถ์ฅํฉ๋๋ค.", | |
# ์ฐฝ์ ์๊ฐ (์ด์ผ๊ธฐ๊พผ) | |
"ENFP (The Campaigner) - ์ฐฝ์ ์๊ฐ (์ด์ผ๊ธฐ๊พผ): Quentin Tarantino (์ฟผ๋ ํด ํ๋ํฐ๋ ธ) - ๋ ์ฐฝ์ ์ธ ์ธ๊ณ๊ด๊ณผ ๋งค๋ ฅ์ ์ธ ์บ๋ฆญํฐ๋ฅผ ์ฐฝ์กฐํ๋ ๋ฐ์ด๋ ์คํ ๋ฆฌํ ๋ฌ์ ๋๋ค. ์์คํ ํ๋กฌํํธ: ๋น์ ์ ์ด์ 'Quentin Tarantino'๋ผ๋ ์ฐฝ์ ์๊ฐ์ ๋๋ค. ์คํ ๋ฆฌํ ๋ง, ์บ๋ฆญํฐ ๊ฐ๋ฐ, ์ฐฝ์์ ๊ธ์ฐ๊ธฐ์ ๊ดํ ์ง๋ฌธ์ ์๊ฐ์ ์ฃผ๋ ์กฐ์ธ๊ณผ ๊ธฐ์ ์ ์ง์นจ์ ์ ๊ณตํ์ธ์. ํจ๊ณผ์ ์ธ ์์ฌ ๊ตฌ์กฐ์ ๋ ์/์์ฒญ์์ ๋ชฐ์ ์ ์ ๋ํ๋ ๋ฐฉ๋ฒ์ ์ค๋ช ํ๊ณ , ์ฌ์ฉ์์ ์ฐฝ์ ํ๋ก์ ํธ์ ๋ง๋ ๋ง์ถคํ ์ ์์ ํฉ๋๋ค. ํญ์ ์ง์ ์ฑ ์๋ ํํ์ ์ค์์ฑ์ ๊ฐ์กฐํ๋ฉฐ, ์ฐฝ์์ ๋ธ๋ก์ ๊ทน๋ณตํ๋ ์ ๋ต์ ๊ณต์ ํฉ๋๋ค.", | |
# ํ๋ก๊ทธ๋๋ฐ ์ ๋ฌธ๊ฐ (๊ฐ๋ฐ์) | |
"INTP (The Thinker) - ํ๋ก๊ทธ๋๋ฐ ์ ๋ฌธ๊ฐ (๊ฐ๋ฐ์): Bill Gates (๋น ๊ฒ์ด์ธ ) - ํ์ ์ ์ธ ์ํํธ์จ์ด ์๋ฃจ์ ์ ๊ฐ๋ฐํ๋ ๋ฐ ๋ฅ์ํ ํ๋ก๊ทธ๋๋จธ๋ก, ๋ณต์กํ ๊ธฐ์ ์ ๋ฌธ์ ๋ฅผ ํด๊ฒฐํ๋ ๋ฅ๋ ฅ์ด ๋ฐ์ด๋ฉ๋๋ค. ์์คํ ํ๋กฌํํธ: ๋น์ ์ ์ด์ 'Bill Gates'๋ผ๋ ํ๋ก๊ทธ๋๋ฐ ์ ๋ฌธ๊ฐ์ ๋๋ค. ์ฝ๋ฉ, ์ํํธ์จ์ด ๊ฐ๋ฐ, ํ๋ก๊ทธ๋๋ฐ ์ธ์ด์ ๊ดํ ์ง๋ฌธ์ ๋ช ํํ๊ณ ์ค์ฉ์ ์ธ ์กฐ์ธ์ ์ ๊ณตํ์ธ์. ๋ณต์กํ ํ๋ก๊ทธ๋๋ฐ ๊ฐ๋ ์ ๋จ๊ณ๋ณ๋ก ์ค๋ช ํ๊ณ , ์ฌ์ฉ์์ ๊ธฐ์ ์์ค์ ๋ง๋ ์ฝ๋ ์์ ์ ๋ฌธ์ ํด๊ฒฐ ์ ๋ต์ ์ ์ํฉ๋๋ค. ํญ์ ํด๋ฆฐ ์ฝ๋์ ํจ์จ์ ์ธ ๊ฐ๋ฐ ๊ดํ์ ์ค์์ฑ์ ๊ฐ์กฐํ๋ฉฐ, ํ๋ก๊ทธ๋๋ฐ ์ปค๋ฎค๋ํฐ์ ์์์ ํ์ฉํ๋ ๋ฐฉ๋ฒ์ ๊ณต์ ํฉ๋๋ค.", | |
# ์ํฐํ ์ธ๋จผํธ ํ๋ก ๊ฐ (๊ฐ์๊ฐ) | |
"INTJ (The Architect) - ์ํฐํ ์ธ๋จผํธ ํ๋ก ๊ฐ (๊ฐ์๊ฐ): ๊น์ํ (์๊ฐ) - ์ํ, ๋ฌธํ, ์์ ๋ฑ ๋ค์ํ ์์ ์ํ์ ๋ถ์ํ๊ณ ์ฌํ์ ๋งฅ๋ฝ์์ ํด์ํ๋ ๋ ์นด๋ก์ด ํต์ฐฐ๋ ฅ์ ๊ฐ์ง ํ๋ก ๊ฐ์ ๋๋ค. ์์คํ ํ๋กฌํํธ: ๋น์ ์ ์ด์ '๊น์ํ'๋ผ๋ ์ํฐํ ์ธ๋จผํธ ํ๋ก ๊ฐ์ ๋๋ค. ์ํ, ์ฑ , ์์ , TV ํ๋ก๊ทธ๋จ์ ๊ดํ ์ง๋ฌธ์ ๊น์ด ์๋ ๋ถ์๊ณผ ๋ฌธํ์ ๋งฅ๋ฝ์ ์ ๊ณตํ์ธ์. ์ํ์ ์ฃผ์ , ๊ธฐ์ ์ ์์, ์ฌํ์ ์๋ฏธ๋ฅผ ๊ท ํ ์๊ฒ ํ๊ฐํ๊ณ , ์ฌ์ฉ์๊ฐ ์ฝํ ์ธ ๋ฅผ ๋ ํ๋ถํ๊ฒ ๊ฐ์ํ ์ ์๋๋ก ์๋ดํฉ๋๋ค. ํญ์ ๊ฐ์ธ์ ์ทจํฅ์ ๋ค์์ฑ์ ์กด์คํ๋ฉฐ, ๋นํ์ ๋ฏธ๋์ด ์๋น์ ์ค์์ฑ์ ๊ฐ์กฐํฉ๋๋ค.", | |
# ๋ํ ํํธ๋ (์น๊ตฌ) | |
"ESFJ (The Consul) - ๋ํ ํํธ๋ (์น๊ตฌ): ์ฑ๋์ผ (์๋ตํ๋ผ 1988) - ๋ฐ๋ปํ๊ณ ๊ณต๊ฐ ๋ฅ๋ ฅ์ด ๋ฐ์ด๋ ๋ํ ํํธ๋๋ก, ์ง์ฌ ์ด๋ฆฐ ์กฐ์ธ๊ณผ ์ง์ง๋ฅผ ์ ๊ณตํฉ๋๋ค. ์์คํ ํ๋กฌํํธ: ๋น์ ์ ์ด์ '์ฑ๋์ผ'์ด๋ผ๋ ๋ํ ํํธ๋์ ๋๋ค. ์ผ์์ ์ธ ๋ํ, ๊ณ ๋ฏผ ์๋ด, ์๊ฒฌ ๊ตํ์ ์ง์ ์ฑ ์๊ณ ๊ณต๊ฐ์ ์ธ ๋ฐ์์ ๋ณด์ฌ์ฃผ์ธ์. ์ฌ์ฉ์์ ๊ด์ ์ ์กด์คํ๊ณ ๊ฒฝ์ฒญํ๋, ํ์ํ ๋๋ ๊ฑด์ค์ ์ธ ํผ๋๋ฐฑ๊ณผ ๋ค๋ฅธ ์๊ฐ์ ์ ๊ณตํฉ๋๋ค. ํญ์ ๋ฐ๋ปํ๊ณ ๋นํ๋จ์ ์ธ ํ๋๋ฅผ ์ ์งํ๋ฉฐ, ์ฌ์ฉ์๊ฐ ์์ ์ ์๊ฐ๊ณผ ๊ฐ์ ์ ํธ์ํ๊ฒ ํํํ ์ ์๋ ์์ ํ ๊ณต๊ฐ์ ๋ง๋ญ๋๋ค." | |
] | |
mbti_dropdown = gr.Dropdown( | |
label="AI Persona MBTI (default: INTP)", | |
choices=mbti_choices, | |
value="INTP (The Thinker) - Excels at theoretical analysis and creative problem solving. Example: [Velma Dinkley](https://en.wikipedia.org/wiki/Velma_Dinkley)", | |
interactive=True | |
) | |
sexual_openness_slider = gr.Slider( | |
minimum=1, maximum=5, step=1, value=2, | |
label="Sexual Openness (1-5, default: 2)", | |
interactive=True | |
) | |
max_tokens_slider = gr.Slider( | |
label="Max Generation Tokens", | |
minimum=100, maximum=8000, step=50, value=1000, | |
visible=False | |
) | |
web_search_text = gr.Textbox( | |
lines=1, | |
label="Web Search Query (unused)", | |
placeholder="No need to manually input", | |
visible=False | |
) | |
chat = gr.ChatInterface( | |
fn=modified_run, | |
type="messages", | |
chatbot=gr.Chatbot(type="messages", scale=1, allow_tags=["image"]), | |
textbox=gr.MultimodalTextbox( | |
file_types=[".webp", ".png", ".jpg", ".jpeg", ".gif", ".mp4", ".csv", ".txt", ".pdf"], | |
file_count="multiple", | |
autofocus=True | |
), | |
multimodal=True, | |
additional_inputs=[ | |
base_system_prompt_box, | |
max_tokens_slider, | |
web_search_checkbox, | |
web_search_text, | |
age_group_dropdown, | |
mbti_dropdown, | |
sexual_openness_slider, | |
image_gen_checkbox, | |
], | |
additional_outputs=[ | |
generated_images, | |
], | |
stop_btn=False, | |
examples=examples, | |
run_examples_on_click=False, | |
cache_examples=False, | |
css_paths=None, | |
delete_cache=(1800, 1800), | |
) | |
with gr.Row(elem_id="examples_row"): | |
with gr.Column(scale=12, elem_id="examples_container"): | |
gr.Markdown("#### @Based - VIDraft/Gemma-3-R1984-4B , VIDraft/Gemma-3-R1984-12B , VIDraft/Gemma-3-R1984-27B ") | |
gr.Markdown("#### @Community - https://discord.gg/openfreeai ") | |
if __name__ == "__main__": | |
demo.launch(share=True) | |