Spaces:
Sleeping
Sleeping
File size: 30,165 Bytes
ce507ec 0751433 c984bb4 9826cfc fe7d37a ce507ec 70efebe d97ac9f ce507ec a1354f4 fe7d37a ce507ec fe7d37a 1c5c923 70efebe d97ac9f 70efebe fe7d37a ce507ec 1c5c923 fe7d37a 70efebe fe7d37a 70efebe fe7d37a 70efebe ce507ec 1c5c923 70efebe fe7d37a 70efebe fe7d37a f4d7d15 fe7d37a 70efebe fe7d37a 9826cfc 70efebe fe7d37a 70efebe fe7d37a 70efebe fe7d37a 70efebe fe7d37a 70efebe fe7d37a fa7abf4 d97ac9f fa7abf4 ce507ec c984bb4 3c54be4 db01582 d97ac9f fa7abf4 db01582 70efebe d7896b6 70efebe d7896b6 70efebe d7896b6 70efebe a1354f4 db01582 fa7abf4 fe7d37a fa7abf4 70efebe d7896b6 fa7abf4 fe7d37a 70efebe fe7d37a fa7abf4 d7896b6 a1354f4 d7896b6 fa7abf4 fe7d37a db01582 fa7abf4 fe7d37a fa7abf4 d7896b6 fa7abf4 d7896b6 70efebe fa7abf4 8b1668a 70efebe fa7abf4 db01582 fa7abf4 70efebe fa7abf4 70efebe a1354f4 70efebe 3c54be4 d7896b6 70efebe d7896b6 d97ac9f fe7d37a 70efebe d97ac9f 70efebe fe7d37a a1354f4 9826cfc fe7d37a a1354f4 9826cfc a1354f4 70efebe a1354f4 70efebe fe7d37a 70efebe fe7d37a 70efebe fe7d37a d97ac9f fa7abf4 d97ac9f d7896b6 8b1668a d97ac9f fe7d37a 70efebe fe7d37a 8b1668a d97ac9f 1c5c923 d97ac9f d7896b6 d97ac9f fa7abf4 d97ac9f fe7d37a fa7abf4 d97ac9f fa7abf4 d7896b6 fa7abf4 d97ac9f fe7d37a 70efebe a1354f4 70efebe a1354f4 70efebe ce507ec d97ac9f ce507ec d97ac9f fa7abf4 ce507ec 70efebe 8b1668a ce507ec 70efebe a1354f4 70efebe ce507ec a1354f4 ce507ec d97ac9f ce507ec fa7abf4 8b1668a fa7abf4 ce507ec 9826cfc d97ac9f d7896b6 70efebe d7896b6 70efebe fe7d37a d7896b6 d97ac9f ce507ec fe7d37a 6aa264c db01582 0751433 70efebe 7dbc041 fa7abf4 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 |
# storyverse_weaver/app.py
import gradio as gr
import os
import time
import json
from PIL import Image, ImageDraw, ImageFont
import random
import traceback
# --- Core Logic Imports ---
from core.llm_services import initialize_text_llms, is_gemini_text_ready, is_hf_text_ready, generate_text_gemini, generate_text_hf
from core.image_services import initialize_image_llms, STABILITY_API_CONFIGURED, OPENAI_DALLE_CONFIGURED, HF_IMAGE_API_CONFIGURED, generate_image_stabilityai, generate_image_dalle, generate_image_hf_model, ImageGenResponse
from core.story_engine import Story, Scene
from prompts.narrative_prompts import get_narrative_system_prompt, format_narrative_user_prompt
from prompts.image_style_prompts import STYLE_PRESETS, COMMON_NEGATIVE_PROMPTS, format_image_generation_prompt
from core.utils import basic_text_cleanup
# --- Initialize Services ---
initialize_text_llms()
initialize_image_llms()
# --- Get API Readiness Status ---
GEMINI_TEXT_IS_READY = is_gemini_text_ready()
HF_TEXT_IS_READY = is_hf_text_ready() # For text models
STABILITY_API_IS_READY = STABILITY_API_CONFIGURED
OPENAI_DALLE_IS_READY = OPENAI_DALLE_CONFIGURED
HF_IMAGE_IS_READY = HF_IMAGE_API_CONFIGURED # For image models via HF API
# --- Application Configuration (Models, Defaults) ---
TEXT_MODELS = {}
UI_DEFAULT_TEXT_MODEL_KEY = None
if GEMINI_TEXT_IS_READY:
TEXT_MODELS["β¨ Gemini 1.5 Flash (Narrate)"] = {"id": "gemini-1.5-flash-latest", "type": "gemini"}
TEXT_MODELS["Legacy Gemini 1.0 Pro (Narrate)"] = {"id": "gemini-1.0-pro-latest", "type": "gemini"}
if HF_TEXT_IS_READY:
TEXT_MODELS["Mistral 7B (Narrate via HF)"] = {"id": "mistralai/Mistral-7B-Instruct-v0.2", "type": "hf_text"}
TEXT_MODELS["Gemma 2B (Narrate via HF)"] = {"id": "google/gemma-2b-it", "type": "hf_text"}
if TEXT_MODELS:
if GEMINI_TEXT_IS_READY and "β¨ Gemini 1.5 Flash (Narrate)" in TEXT_MODELS: UI_DEFAULT_TEXT_MODEL_KEY = "β¨ Gemini 1.5 Flash (Narrate)"
elif HF_TEXT_IS_READY and "Mistral 7B (Narrate via HF)" in TEXT_MODELS: UI_DEFAULT_TEXT_MODEL_KEY = "Mistral 7B (Narrate via HF)"
elif TEXT_MODELS: UI_DEFAULT_TEXT_MODEL_KEY = list(TEXT_MODELS.keys())[0]
else:
TEXT_MODELS["No Text Models Configured"] = {"id": "dummy_text_error", "type": "none"}
UI_DEFAULT_TEXT_MODEL_KEY = "No Text Models Configured"
IMAGE_PROVIDERS = {}
UI_DEFAULT_IMAGE_PROVIDER_KEY = None
# Order of preference for default: Stability, then HF, then DALL-E (simulated)
if STABILITY_API_IS_READY: IMAGE_PROVIDERS["π¨ Stability AI (SDXL)"] = "stability_ai"
if HF_IMAGE_IS_READY:
IMAGE_PROVIDERS["π‘ HF Model (SDXL Base)"] = "hf_image_sdxl_base" # Key for SDXL via HF
IMAGE_PROVIDERS["π HF Model (OpenJourney)"] = "hf_image_openjourney" # Key for OpenJourney via HF
if OPENAI_DALLE_IS_READY: IMAGE_PROVIDERS["πΌοΈ DALL-E 3 (Sim.)"] = "dalle"
if IMAGE_PROVIDERS:
if STABILITY_API_IS_READY and "π¨ Stability AI (SDXL)" in IMAGE_PROVIDERS: UI_DEFAULT_IMAGE_PROVIDER_KEY = "π¨ Stability AI (SDXL)"
elif HF_IMAGE_IS_READY and "π‘ HF Model (SDXL Base)" in IMAGE_PROVIDERS: UI_DEFAULT_IMAGE_PROVIDER_KEY = "π‘ HF Model (SDXL Base)"
elif OPENAI_DALLE_IS_READY and "πΌοΈ DALL-E 3 (Sim.)" in IMAGE_PROVIDERS: UI_DEFAULT_IMAGE_PROVIDER_KEY = "πΌοΈ DALL-E 3 (Sim.)"
else: UI_DEFAULT_IMAGE_PROVIDER_KEY = list(IMAGE_PROVIDERS.keys())[0]
else:
IMAGE_PROVIDERS["No Image Providers Configured"] = "none"
UI_DEFAULT_IMAGE_PROVIDER_KEY = "No Image Providers Configured"
# --- Gradio UI Theme and CSS ---
omega_theme = gr.themes.Base(
font=[gr.themes.GoogleFont("Lexend Deca"), "ui-sans-serif", "system-ui", "sans-serif"],
primary_hue=gr.themes.colors.purple, secondary_hue=gr.themes.colors.pink, neutral_hue=gr.themes.colors.slate
).set(
body_background_fill="#0F0F1A", block_background_fill="#1A1A2E", block_border_width="1px",
block_border_color="#2A2A4A", block_label_background_fill="#2A2A4A", input_background_fill="#2A2A4A",
input_border_color="#4A4A6A", button_primary_background_fill="linear-gradient(135deg, #7F00FF 0%, #E100FF 100%)",
button_primary_text_color="white", button_secondary_background_fill="#4A4A6A",
button_secondary_text_color="#E0E0FF", slider_color="#A020F0"
)
omega_css = """
body, .gradio-container { background-color: #0F0F1A !important; color: #D0D0E0 !important; }
.gradio-container { max-width: 1400px !important; margin: auto !important; border-radius: 20px; box-shadow: 0 10px 30px rgba(0,0,0,0.2); padding: 25px !important; border: 1px solid #2A2A4A;}
.gr-panel, .gr-box, .gr-accordion { background-color: #1A1A2E !important; border: 1px solid #2A2A4A !important; border-radius: 12px !important; box-shadow: 0 4px 15px rgba(0,0,0,0.1);}
.gr-markdown h1 { font-size: 2.8em !important; text-align: center; color: transparent; background: linear-gradient(135deg, #A020F0 0%, #E040FB 100%); -webkit-background-clip: text; background-clip: text; margin-bottom: 5px !important; letter-spacing: -1px;}
.gr-markdown h3 { color: #C080F0 !important; text-align: center; font-weight: 400; margin-bottom: 25px !important;}
.input-section-header { font-size: 1.6em; font-weight: 600; color: #D0D0FF; margin-top: 15px; margin-bottom: 8px; border-bottom: 2px solid #7F00FF; padding-bottom: 5px;}
.output-section-header { font-size: 1.8em; font-weight: 600; color: #D0D0FF; margin-top: 15px; margin-bottom: 12px;}
.gr-input input, .gr-input textarea, .gr-dropdown select, .gr-textbox textarea { background-color: #2A2A4A !important; color: #E0E0FF !important; border: 1px solid #4A4A6A !important; border-radius: 8px !important; padding: 10px !important;}
.gr-button { border-radius: 8px !important; font-weight: 500 !important; transition: all 0.2s ease-in-out !important;}
.gr-button-primary { padding-top: 10px !important; padding-bottom: 10px !important; } /* Ensure enough padding for icons/text */
.gr-button-primary:hover { transform: scale(1.03) translateY(-1px) !important; box-shadow: 0 8px 16px rgba(127,0,255,0.3) !important; }
.panel_image { border-radius: 12px !important; overflow: hidden; box-shadow: 0 6px 15px rgba(0,0,0,0.25) !important; background-color: #23233A;}
.panel_image img { max-height: 600px !important; }
.gallery_output { background-color: transparent !important; border: none !important; }
.gallery_output .thumbnail-item { border-radius: 8px !important; box-shadow: 0 3px 8px rgba(0,0,0,0.2) !important; margin: 6px !important; transition: transform 0.2s ease; height: 180px !important; width: 180px !important;}
.gallery_output .thumbnail-item:hover { transform: scale(1.05); }
.status_text { font-weight: 500; padding: 12px 18px; text-align: center; border-radius: 8px; margin-top:12px; border: 1px solid transparent; font-size: 1.05em;}
.error_text { background-color: #401010 !important; color: #FFB0B0 !important; border-color: #802020 !important; }
.success_text { background-color: #104010 !important; color: #B0FFB0 !important; border-color: #208020 !important;}
.processing_text { background-color: #102040 !important; color: #B0D0FF !important; border-color: #204080 !important;}
.important-note { background-color: rgba(127,0,255,0.1); border-left: 5px solid #7F00FF; padding: 15px; margin-bottom:20px; color: #E0E0FF; border-radius: 6px;}
.gr-tabitem { background-color: #1A1A2E !important; border-radius: 0 0 12px 12px !important; padding: 15px !important;}
.gr-tab-button.selected { background-color: #2A2A4A !important; color: white !important; border-bottom: 3px solid #A020F0 !important; border-radius: 8px 8px 0 0 !important; font-weight: 600 !important;}
.gr-tab-button { color: #A0A0C0 !important; border-radius: 8px 8px 0 0 !important;}
.gr-accordion > .gr-block { border-top: 1px solid #2A2A4A !important; }
.gr-markdown code { background-color: #2A2A4A !important; color: #C0C0E0 !important; padding: 0.2em 0.5em; border-radius: 4px; }
.gr-markdown pre { background-color: #23233A !important; padding: 1em !important; border-radius: 6px !important; border: 1px solid #2A2A4A !important;}
.gr-markdown pre > code { padding: 0 !important; background-color: transparent !important; }
#surprise_button { background: linear-gradient(135deg, #ff7e5f 0%, #feb47b 100%) !important; font-weight:600 !important;}
#surprise_button:hover { transform: scale(1.03) translateY(-1px) !important; box-shadow: 0 8px 16px rgba(255,126,95,0.3) !important; }
"""
# --- Helper: Placeholder Image Creation ---
def create_placeholder_image(text="Processing...", size=(512, 512), color="#23233A", text_color="#E0E0FF"):
# ... (Full implementation as before) ...
img = Image.new('RGB', size, color=color); draw = ImageDraw.Draw(img)
try: font_path = "arial.ttf" if os.path.exists("arial.ttf") else None
except: font_path = None
try: font = ImageFont.truetype(font_path, 40) if font_path else ImageFont.load_default()
except IOError: font = ImageFont.load_default()
if hasattr(draw, 'textbbox'): bbox = draw.textbbox((0,0), text, font=font); tw, th = bbox[2]-bbox[0], bbox[3]-bbox[1]
else: tw, th = draw.textsize(text, font=font)
draw.text(((size[0]-tw)/2, (size[1]-th)/2), text, font=font, fill=text_color); return img
# --- StoryVerse Weaver Orchestrator ---
def add_scene_to_story_orchestrator(
current_story_obj: Story, scene_prompt_text: str, image_style_dropdown: str, artist_style_text: str,
negative_prompt_text: str, text_model_key: str, image_provider_key: str,
narrative_length: str, image_quality: str,
progress=gr.Progress(track_tqdm=True)
):
start_time = time.time()
if not current_story_obj: current_story_obj = Story()
log_accumulator = [f"**π Scene {current_story_obj.current_scene_number + 1} - {time.strftime('%H:%M:%S')}**"]
# Placeholders for the final return tuple
ret_story_state = current_story_obj
ret_gallery = current_story_obj.get_all_scenes_for_gallery_display()
ret_latest_image = create_placeholder_image("π¨ Conjuring visuals...") # Initial placeholder for image component
ret_latest_narrative_md_obj = gr.Markdown(value=" Musing narrative...")
ret_status_bar_html_obj = gr.HTML(value=f"<p class='processing_text status_text'>π Weaving Scene {current_story_obj.current_scene_number + 1}...</p>")
# Log is built up
# Initial yield for UI updates handled by the .then() chain for buttons
# This first yield is for immediate feedback on other components
yield {
output_status_bar: ret_status_bar_html_obj,
output_latest_scene_image: gr.Image(value=ret_latest_image), # Show placeholder
output_latest_scene_narrative: ret_latest_narrative_md_obj,
output_interaction_log_markdown: gr.Markdown(value="\n".join(log_accumulator))
}
try:
if not scene_prompt_text.strip():
raise ValueError("Scene prompt cannot be empty!")
# --- 1. Generate Narrative Text ---
progress(0.1, desc="βοΈ Crafting narrative...")
narrative_text_generated = f"Narrative Error: Init failed."
text_model_info = TEXT_MODELS.get(text_model_key)
if text_model_info and text_model_info["type"] != "none":
system_p = get_narrative_system_prompt("default")
prev_narrative = current_story_obj.get_last_scene_narrative()
user_p = format_narrative_user_prompt(scene_prompt_text, prev_narrative)
log_accumulator.append(f" Narrative: Using {text_model_key} ({text_model_info['id']}). Length: {narrative_length}")
text_response = None
if text_model_info["type"] == "gemini": text_response = generate_text_gemini(user_p, model_id=text_model_info["id"], system_prompt=system_p, max_tokens=768 if narrative_length.startswith("Detailed") else 400)
elif text_model_info["type"] == "hf_text": text_response = generate_text_hf(user_p, model_id=text_model_info["id"], system_prompt=system_p, max_tokens=768 if narrative_length.startswith("Detailed") else 400)
if text_response and text_response.success:
narrative_text_generated = basic_text_cleanup(text_response.text)
log_accumulator.append(f" Narrative: Success. (Snippet: {narrative_text_generated[:50]}...)")
elif text_response:
narrative_text_generated = f"**Narrative Error ({text_model_key}):** {text_response.error}"
log_accumulator.append(f" Narrative: FAILED - {text_response.error}")
else:
log_accumulator.append(f" Narrative: FAILED - No response object from {text_model_key}.")
else:
narrative_text_generated = "**Narrative Error:** Selected text model not available or misconfigured."
log_accumulator.append(f" Narrative: FAILED - Model '{text_model_key}' not available.")
ret_latest_narrative_str_content = f"## Scene Idea: {scene_prompt_text}\n\n{narrative_text_generated}"
ret_latest_narrative_md_obj = gr.Markdown(value=ret_latest_narrative_str_content)
yield { output_latest_scene_narrative: ret_latest_narrative_md_obj,
output_interaction_log_markdown: gr.Markdown(value="\n".join(log_accumulator)) }
# --- 2. Generate Image ---
progress(0.5, desc="π¨ Conjuring visuals...")
image_generated_pil = None
image_generation_error_message = None
selected_image_provider_type = IMAGE_PROVIDERS.get(image_provider_key)
image_content_prompt_for_gen = narrative_text_generated if narrative_text_generated and "Error" not in narrative_text_generated else scene_prompt_text
quality_keyword = "ultra detailed, intricate, masterpiece, " if image_quality == "High Detail" else ("concept sketch, line art, " if image_quality == "Sketch Concept" else "")
full_image_prompt = format_image_generation_prompt(quality_keyword + image_content_prompt_for_gen[:350], image_style_dropdown, artist_style_text)
log_accumulator.append(f" Image: Using {image_provider_key}. Style: {image_style_dropdown}. Artist: {artist_style_text or 'N/A'}.")
if selected_image_provider_type and selected_image_provider_type != "none":
image_response = None
if selected_image_provider_type == "stability_ai": image_response = generate_image_stabilityai(full_image_prompt, negative_prompt=negative_prompt_text or COMMON_NEGATIVE_PROMPTS, steps=40 if image_quality=="High Detail" else 25)
elif selected_image_provider_type == "dalle": image_response = generate_image_dalle(full_image_prompt, quality="hd" if image_quality=="High Detail" else "standard")
elif selected_image_provider_type == "hf_image_sdxl_base": image_response = generate_image_hf_model(full_image_prompt, model_id="stabilityai/stable-diffusion-xl-base-1.0", negative_prompt=negative_prompt_text or COMMON_NEGATIVE_PROMPTS)
elif selected_image_provider_type == "hf_image_openjourney": image_response = generate_image_hf_model(full_image_prompt, model_id="prompthero/openjourney", negative_prompt=negative_prompt_text or COMMON_NEGATIVE_PROMPTS, width=512, height=512)
if image_response and image_response.success:
image_generated_pil = image_response.image
log_accumulator.append(f" Image: Success from {image_response.provider}.")
elif image_response:
image_generation_error_message = f"**Image Error ({image_response.provider}):** {image_response.error}"
log_accumulator.append(f" Image: FAILED - {image_response.error}")
else:
image_generation_error_message = f"**Image Error:** No response object from {image_provider_key} service."
log_accumulator.append(f" Image: FAILED - No response object from {image_provider_key}.")
else:
image_generation_error_message = "**Image Error:** Selected image provider not available or misconfigured."
log_accumulator.append(f" Image: FAILED - Provider '{image_provider_key}' unavailable.")
ret_latest_image = image_generated_pil if image_generated_pil else create_placeholder_image("Image Gen Failed", color="#401010")
yield { output_latest_scene_image: gr.Image(value=ret_latest_image),
output_interaction_log_markdown: gr.Markdown(value="\n".join(log_accumulator)) }
# --- 3. Add Scene to Story Object ---
final_scene_error = None
if image_generation_error_message and "**Narrative Error**" in narrative_text_generated : final_scene_error = f"{narrative_text_generated}\n{image_generation_error_message}"
elif "**Narrative Error**" in narrative_text_generated: final_scene_error = narrative_text_generated
elif image_generation_error_message: final_scene_error = image_generation_error_message
current_story_obj.add_scene_from_elements(
user_prompt=scene_prompt_text,
narrative_text=narrative_text_generated if "**Narrative Error**" not in narrative_text_generated else "(Narrative gen failed)",
image=image_generated_pil,
image_style_prompt=f"{image_style_dropdown}{f', by {artist_style_text}' if artist_style_text and artist_style_text.strip() else ''}",
image_provider=image_provider_key if selected_image_provider_type != "none" else "N/A",
error_message=final_scene_error
)
ret_story_state = current_story_obj
log_accumulator.append(f" Scene {current_story_obj.current_scene_number} processed and added.")
# --- 4. Prepare Final Values for Return Tuple ---
ret_gallery = current_story_obj.get_all_scenes_for_gallery_display()
_ , latest_narr_for_display_final_str_temp = current_story_obj.get_latest_scene_details_for_display()
ret_latest_narrative_md_obj = gr.Markdown(value=latest_narr_for_display_final_str_temp)
status_html_str_temp = f"<p class='error_text status_text'>Scene {current_story_obj.current_scene_number} added with errors.</p>" if final_scene_error else f"<p class='success_text status_text'>π Scene {current_story_obj.current_scene_number} woven!</p>"
ret_status_bar_html_obj = gr.HTML(value=status_html_str_temp)
progress(1.0, desc="Scene Complete!")
except ValueError as ve:
log_accumulator.append(f"\n**INPUT/CONFIG ERROR:** {ve}")
ret_status_bar_html_obj = gr.HTML(value=f"<p class='error_text status_text'>β CONFIGURATION ERROR: {ve}</p>")
ret_latest_narrative_md_obj = gr.Markdown(value=f"## Error\n{ve}")
except Exception as e:
log_accumulator.append(f"\n**UNEXPECTED RUNTIME ERROR:** {type(e).__name__} - {e}\n{traceback.format_exc()}")
ret_status_bar_html_obj = gr.HTML(value=f"<p class='error_text status_text'>β UNEXPECTED ERROR: {type(e).__name__}. Check logs.</p>")
ret_latest_narrative_md_obj = gr.Markdown(value=f"## Unexpected Error\n{type(e).__name__}: {e}\nSee log for details.")
# `finally` block is removed from here; button re-enabling is handled by .then()
current_total_time = time.time() - start_time
log_accumulator.append(f" Cycle ended at {time.strftime('%H:%M:%S')}. Total time: {current_total_time:.2f}s")
ret_log_md = gr.Markdown(value="\n".join(log_accumulator))
return (
ret_story_state, ret_gallery, ret_latest_image,
ret_latest_narrative_md_obj, ret_status_bar_html_obj, ret_log_md
)
def clear_story_state_ui_wrapper():
new_story = Story()
placeholder_img = create_placeholder_image("Your StoryVerse is a blank canvas...", color="#1A1A2E", text_color="#A0A0C0")
cleared_gallery = [(placeholder_img, "Your StoryVerse is new and untold...")]
initial_narrative = "## β¨ A New Story Begins β¨\nDescribe your first scene idea..."
status_msg = "<p class='processing_text status_text'>π Story Cleared.</p>"
return (new_story, cleared_gallery, None, gr.Markdown(initial_narrative), gr.HTML(status_msg), "Log Cleared.", "")
def surprise_me_func():
themes = ["Cosmic Horror", "Solarpunk Utopia", "Mythic Fantasy", "Noir Detective"]; actions = ["unearths an artifact", "negotiates"]; settings = ["on a rogue planet", "in a city in a tree"]; prompt = f"A protagonist {random.choice(actions)} {random.choice(settings)}. Theme: {random.choice(themes)}."; style = random.choice(list(STYLE_PRESETS.keys())); artist = random.choice(["H.R. Giger", "Moebius", ""]*2); return prompt, style, artist
def disable_buttons_for_processing():
return gr.Button(interactive=False), gr.Button(interactive=False)
def enable_buttons_after_processing():
return gr.Button(interactive=True), gr.Button(interactive=True)
# --- Gradio UI Definition ---
with gr.Blocks(theme=omega_theme, css=omega_css, title="β¨ StoryVerse Omega β¨") as story_weaver_demo:
# Define Python variables for UI components that will be inputs or outputs
story_state_output = gr.State(Story())
# Input Column Components
scene_prompt_input = gr.Textbox()
image_style_input = gr.Dropdown()
artist_style_input = gr.Textbox()
negative_prompt_input = gr.Textbox()
text_model_dropdown = gr.Dropdown()
image_provider_dropdown = gr.Dropdown()
narrative_length_dropdown = gr.Dropdown()
image_quality_dropdown = gr.Dropdown()
# Output Column Components
output_gallery = gr.Gallery()
output_latest_scene_image = gr.Image()
output_latest_scene_narrative = gr.Markdown()
output_status_bar = gr.HTML()
output_interaction_log_markdown = gr.Markdown()
# Button components (updated via chained .then() events)
engage_button = gr.Button()
surprise_button = gr.Button()
clear_story_button = gr.Button()
# --- Layout the UI ---
gr.Markdown("<div align='center'><h1>β¨ StoryVerse Omega β¨</h1>\n<h3>Craft Immersive Multimodal Worlds with AI</h3></div>")
gr.HTML("<div class='important-note'><strong>Welcome, Worldsmith!</strong> Describe your vision, choose your style, and let Omega help you weave captivating scenes with narrative and imagery. Ensure API keys (<code>STORYVERSE_...</code>) are correctly set in Space Secrets!</div>")
with gr.Accordion("π§ AI Services Status & Info", open=False):
status_text_list = []; text_llm_ok, image_gen_ok = (GEMINI_TEXT_IS_READY or HF_TEXT_IS_READY), (STABILITY_API_IS_READY or OPENAI_DALLE_IS_READY or HF_IMAGE_IS_READY) # Added HF_IMAGE_IS_READY
if not text_llm_ok and not image_gen_ok: status_text_list.append("<p style='color:#FCA5A5;font-weight:bold;'>β οΈ CRITICAL: NO AI SERVICES CONFIGURED.</p>")
else:
if text_llm_ok: status_text_list.append("<p style='color:#A7F3D0;'>β
Text Generation Service(s) Ready.</p>")
else: status_text_list.append("<p style='color:#FCD34D;'>β οΈ Text Generation Service(s) NOT Ready.</p>")
if image_gen_ok: status_text_list.append("<p style='color:#A7F3D0;'>β
Image Generation Service(s) Ready.</p>")
else: status_text_list.append("<p style='color:#FCD34D;'>β οΈ Image Generation Service(s) NOT Ready.</p>")
gr.HTML("".join(status_text_list))
with gr.Row(equal_height=False, variant="panel"):
with gr.Column(scale=7, min_width=450):
gr.Markdown("### π‘ **Craft Your Scene**", elem_classes="input-section-header")
with gr.Group():
scene_prompt_input = gr.Textbox(lines=7, label="Scene Vision (Description, Dialogue, Action):", placeholder="e.g., Amidst swirling cosmic dust...")
with gr.Row(elem_classes=["compact-row"]):
with gr.Column(scale=2):
image_style_input = gr.Dropdown(choices=["Default (Cinematic Realism)"] + sorted(list(STYLE_PRESETS.keys())), value="Default (Cinematic Realism)", label="Visual Style Preset", allow_custom_value=True) # allow_custom_value for examples
with gr.Column(scale=2):
artist_style_input = gr.Textbox(label="Artistic Inspiration (Optional):", placeholder="e.g., Moebius...")
negative_prompt_input = gr.Textbox(lines=2, label="Exclude from Image (Negative Prompt):", value=COMMON_NEGATIVE_PROMPTS)
with gr.Accordion("βοΈ Advanced AI Configuration", open=False):
with gr.Group():
text_model_dropdown = gr.Dropdown(choices=list(TEXT_MODELS.keys()), value=UI_DEFAULT_TEXT_MODEL_KEY, label="Narrative AI Engine")
image_provider_dropdown = gr.Dropdown(choices=list(IMAGE_PROVIDERS.keys()), value=UI_DEFAULT_IMAGE_PROVIDER_KEY, label="Visual AI Engine")
with gr.Row():
narrative_length_dropdown = gr.Dropdown(["Short (1 paragraph)", "Medium (2-3 paragraphs)", "Detailed (4+ paragraphs)"], value="Medium (2-3 paragraphs)", label="Narrative Detail")
image_quality_dropdown = gr.Dropdown(["Standard", "High Detail", "Sketch Concept"], value="Standard", label="Image Detail/Style")
with gr.Row(elem_classes=["compact-row"], equal_height=True):
engage_button = gr.Button("π Weave This Scene!", variant="primary", scale=3, icon="β¨")
surprise_button = gr.Button("π² Surprise Me!", variant="secondary", scale=1, icon="π")
clear_story_button = gr.Button("ποΈ New Story", variant="stop", scale=1, icon="β»οΈ")
output_status_bar = gr.HTML(value="<p class='processing_text status_text'>Ready to weave your first masterpiece!</p>")
with gr.Column(scale=10, min_width=700):
gr.Markdown("### πΌοΈ **Your Evolving StoryVerse**", elem_classes="output-section-header")
with gr.Tabs():
with gr.TabItem("π Latest Scene", id="latest_scene_tab"):
output_latest_scene_image = gr.Image(label="Latest Scene Image", type="pil", interactive=False, show_download_button=True, height=512, show_label=False, elem_classes=["panel_image"])
output_latest_scene_narrative = gr.Markdown()
with gr.TabItem("π Story Scroll", id="story_scroll_tab"):
output_gallery = gr.Gallery(label="Story Scroll", show_label=False, columns=4, object_fit="cover", height=700, preview=True, allow_preview=True, elem_classes=["gallery_output"])
with gr.TabItem("βοΈ Interaction Log", id="log_tab"):
with gr.Accordion(label="Developer Interaction Log", open=False):
output_interaction_log_markdown = gr.Markdown("Log will appear here...")
# Chained event handling for engage_button
engage_button.click(
fn=disable_buttons_for_processing,
inputs=None,
outputs=[engage_button, surprise_button],
queue=False
).then(
fn=add_scene_to_story_orchestrator,
inputs=[
story_state_output, scene_prompt_input,
image_style_input, artist_style_input, negative_prompt_input,
text_model_dropdown, image_provider_dropdown,
narrative_length_dropdown, image_quality_dropdown
],
outputs=[
story_state_output, output_gallery, output_latest_scene_image,
output_latest_scene_narrative, output_status_bar, output_interaction_log_markdown
]
).then(
fn=enable_buttons_after_processing,
inputs=None,
outputs=[engage_button, surprise_button],
queue=False
)
clear_story_button.click(
fn=clear_story_state_ui_wrapper,
inputs=[],
outputs=[
story_state_output, output_gallery, output_latest_scene_image,
output_latest_scene_narrative, output_status_bar, output_interaction_log_markdown,
scene_prompt_input
]
)
surprise_button.click(
fn=surprise_me_func,
inputs=[],
outputs=[scene_prompt_input, image_style_input, artist_style_input]
)
gr.Examples(
examples=[ # Ensure these styles are in your STYLE_PRESETS in image_style_prompts.py
["A lone, weary traveler on a mechanical steed crosses a vast, crimson desert under twin suns. Dust devils dance in the distance.", "Sci-Fi Western", "Moebius", "greenery, water, modern city"],
["Deep within an ancient, bioluminescent forest, a hidden civilization of sentient fungi perform a mystical ritual around a pulsating crystal.", "Psychedelic Fantasy", "Alex Grey", "technology, buildings, roads"],
["A child sits on a crescent moon, fishing for stars in a swirling nebula. A friendly space whale swims nearby.", "Whimsical Cosmic", "James Jean", "realistic, dark, scary"],
["A grand, baroque library where the books fly freely and whisper forgotten lore to those who listen closely.", "Magical Realism", "Remedios Varo", "minimalist, simple, technology"]
],
inputs=[scene_prompt_input, image_style_input, artist_style_input, negative_prompt_input],
label="π Example Universes to Weave π",
)
gr.HTML("<div style='text-align:center; margin-top:30px; padding-bottom:20px;'><p style='font-size:0.9em; color:#8080A0;'>β¨ StoryVerse Omegaβ’ - Weaving Worlds with Words and Pixels β¨</p></div>")
# --- Entry Point ---
if __name__ == "__main__":
print("="*80); print("β¨ StoryVerse Omegaβ’ Launching... β¨")
print(f" Gemini Text Ready: {GEMINI_TEXT_IS_READY}"); print(f" HF Text Ready: {HF_TEXT_IS_READY}")
print(f" Stability AI Ready: {STABILITY_API_IS_READY}"); print(f" DALL-E Ready: {OPENAI_DALLE_IS_READY}"); print(f" HF Image Ready: {HF_IMAGE_IS_READY}")
if not (GEMINI_TEXT_IS_READY or HF_TEXT_IS_READY) or not (STABILITY_API_IS_READY or OPENAI_DALLE_IS_READY or HF_IMAGE_IS_READY): print(" π΄ WARNING: Not all services configured.")
print(f" Default Text Model: {UI_DEFAULT_TEXT_MODEL_KEY}"); print(f" Default Image Provider: {UI_DEFAULT_IMAGE_PROVIDER_KEY}")
print("="*80)
story_weaver_demo.launch(debug=True, server_name="0.0.0.0", share=False) |