Spaces:
Sleeping
Sleeping
File size: 19,955 Bytes
82f04d1 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 |
# # # story_script.py
# # import asyncio
# # import json
# # import logging
# # import random
# # logging.basicConfig(level=logging.INFO, format="%(asctime)s [%(levelname)s] %(message)s")
# # async def generate_story(script: str) -> dict:
# # """
# # Convert the ad script into a structured storyboard JSON format:
# # - Characters
# # - Scenes with keyframes, camera instructions, and music
# # """
# # # Sample character extraction (for simplicity, can be improved)
# # characters = [{"name": name, "seed": idx+1} for idx, name in enumerate(["MainGuy", "Friend1", "Friend2"])]
# # # Split script into lines and create scenes (basic heuristic, can be improved)
# # lines = [line.strip() for line in script.split("\n") if line.strip()]
# # scenes = {}
# # for idx, line in enumerate(lines, start=1):
# # char = characters[idx % len(characters)]["name"]
# # seed = characters[idx % len(characters)]["seed"]
# # scenes[f"scene{idx}"] = {
# # "character": char,
# # "scene": line,
# # "keyframes": [
# # {
# # "seed": seed,
# # "keyframe1": f"{char} in action based on script line: '{line[:40]}...'",
# # "keyframe2": f"{char} expressive close-up reacting to: '{line[:40]}...'"
# # }
# # ],
# # "camera": "Medium shot with dynamic zoom-ins",
# # "music": "Appropriate upbeat or dramatic tune based on action"
# # }
# # storyboard = {
# # "characters": characters,
# # **scenes
# # }
# # logging.info("Story script generated successfully")
# # return storyboard
# # # best
# # # story_script.py
# # import os
# # import asyncio
# # import httpx
# # import logging
# # from dotenv import load_dotenv
# # from dotenv import load_dotenv
# # dotenv_path = os.path.join(os.path.dirname(os.path.dirname(__file__)), '.env')
# # load_dotenv(dotenv_path)
# # load_dotenv()
# # logging.basicConfig(level=logging.INFO, format="%(asctime)s [%(levelname)s] %(message)s")
# # OPENROUTER_API_KEY = os.getenv("OPENROUTER_API_KEY")
# # MODEL_NAME = "deepseek/deepseek-r1-distill-llama-70b:free"
# # OPENROUTER_URL = "https://openrouter.ai/api/v1/chat/completions"
# # async def generate_story(script: str) -> dict:
# # """
# # Convert the ad script into a structured storyboard JSON using AI.
# # The JSON format includes:
# # - characters (with seeds)
# # - scenes (character, scene description, keyframes, camera, music)
# # """
# # prompt = f"""
# # You are a professional ad storyboard generator.
# # Take this ad script and convert it into a **storyboard JSON**.
# # Follow this format exactly:
# # ADD {{
# # "characters": [
# # {{"name": "MainGuy", "seed": 1}},
# # {{"name": "Friend1", "seed": 2}},
# # {{"name": "Friend2", "seed": 3}}
# # ],
# # "scene1": {{
# # "character": "MainGuy",
# # "scene": "Description of the scene",
# # "keyframes": [
# # {{
# # "seed": 1,
# # "keyframe1": "First keyframe description",
# # "keyframe2": "Second keyframe description"
# # }}
# # ],
# # "camera": "Camera instructions",
# # "music": "Music instructions"
# # }}
# # ...
# # }}
# # Ensure:
# # - Use the **script lines as scenes**.
# # - Assign characters logically to actions.
# # - Provide **keyframes, camera, and music**.
# # - Return **valid JSON only**, no extra text.
# # Script:
# # \"\"\"{script}\"\"\"
# # """
# # headers = {"Authorization": f"Bearer {OPENROUTER_API_KEY}"}
# # payload = {
# # "model": MODEL_NAME,
# # "messages": [{"role": "user", "content": prompt}],
# # "temperature": 0.8,
# # "max_tokens": 1200
# # }
# # async with httpx.AsyncClient(timeout=120) as client:
# # response = await client.post(OPENROUTER_URL, json=payload, headers=headers)
# # response.raise_for_status()
# # data = response.json()
# # # AI returns the JSON as a string
# # story_json_str = data["choices"][0]["message"]["content"]
# # # Remove possible extra text before/after JSON (some AI outputs might wrap with "ADD {...}")
# # if story_json_str.startswith("ADD"):
# # story_json_str = story_json_str[story_json_str.find("{"):]
# # # Convert string to dict
# # try:
# # story_dict = eval(story_json_str) # safe because AI returns JSON-like dict
# # except Exception as e:
# # logging.error(f"Failed to parse story JSON: {e}")
# # story_dict = {}
# # logging.info("Story script generated successfully using AI")
# # return story_dict
# # import os
# # import asyncio
# # import httpx
# # import logging
# # import json
# # from dotenv import load_dotenv
# # dotenv_path = os.path.join(os.path.dirname(os.path.dirname(__file__)), '.env')
# # load_dotenv(dotenv_path)
# # logging.basicConfig(level=logging.INFO, format="%(asctime)s [%(levelname)s] %(message)s")
# # OPENROUTER_API_KEY = os.getenv("OPENROUTER_API_KEY")
# # MODEL_NAME = "deepseek/deepseek-r1-distill-llama-70b:free"
# # OPENROUTER_URL = "https://openrouter.ai/api/v1/chat/completions"
# # async def generate_story(script: str) -> dict:
# # """
# # Convert the ad script into a structured storyboard JSON using AI.
# # The JSON format includes:
# # - characters (with seeds)
# # - scenes (character, scene description, keyframes, camera, music)
# # """
# # prompt = f"""
# # You are a professional ad storyboard generator.
# # Take this ad script and convert it into a **storyboard JSON**.
# # Follow this format exactly:
# # ADD {{
# # "characters": [
# # {{"name": "MainGuy", "seed": 1}},
# # {{"name": "Friend1", "seed": 2}},
# # {{"name": "Friend2", "seed": 3}}
# # ],
# # "scene1": {{
# # "character": "MainGuy",
# # "scene": "Description of the scene",
# # "keyframes": [
# # {{
# # "seed": 1,
# # "keyframe1": "First keyframe description",
# # "keyframe2": "Second keyframe description"
# # }}
# # ],
# # "camera": "Camera instructions",
# # "music": "Music instructions"
# # }}
# # ...
# # }}
# # Ensure:
# # - Use the **script lines as scenes**.
# # - Assign characters logically to actions.
# # - Provide **keyframes, camera, and music**.
# # - Return **valid JSON only**, no extra text, no markdown, no ``` fences.
# # Script:
# # \"\"\"{script}\"\"\"
# # """
# # headers = {"Authorization": f"Bearer {OPENROUTER_API_KEY}"}
# # payload = {
# # "model": MODEL_NAME,
# # "messages": [{"role": "user", "content": prompt}],
# # "temperature": 0.8,
# # "max_tokens": 1500
# # }
# # async with httpx.AsyncClient(timeout=120) as client:
# # response = await client.post(OPENROUTER_URL, json=payload, headers=headers)
# # response.raise_for_status()
# # data = response.json()
# # story_json_str = data["choices"][0]["message"]["content"]
# # # Clean unwanted wrappers
# # story_json_str = story_json_str.strip()
# # if story_json_str.startswith("ADD"):
# # story_json_str = story_json_str[story_json_str.find("{"):]
# # # Remove markdown fences
# # story_json_str = story_json_str.replace("```json", "").replace("```", "").strip()
# # # Parse safely
# # try:
# # story_dict = json.loads(story_json_str)
# # except json.JSONDecodeError as e:
# # logging.error(f"β Failed to parse story JSON properly: {e}")
# # story_dict = {"raw_output": story_json_str}
# # logging.info("Story script generated successfully using AI")
# # return story_dict
# import os
# import asyncio
# import httpx
# import logging
# import json
# import re
# from dotenv import load_dotenv
# dotenv_path = os.path.join(os.path.dirname(os.path.dirname(__file__)), '.env')
# load_dotenv(dotenv_path)
# logging.basicConfig(level=logging.INFO, format="%(asctime)s [%(levelname)s] %(message)s")
# OPENROUTER_API_KEY = os.getenv("OPENROUTER_API_KEY")
# MODEL_NAME = "deepseek/deepseek-r1-distill-llama-70b:free"
# OPENROUTER_URL = "https://openrouter.ai/api/v1/chat/completions"
# async def generate_story(script: str) -> dict:
# """
# Convert the ad script into a structured storyboard JSON using AI.
# Returns:
# {
# "raw_output": "<original AI string>",
# "parsed_output": <dict | None>
# }
# """
# prompt = f"""
# You are a professional ad storyboard generator.
# Take this ad script and convert it into a **storyboard JSON**.
# Follow this format exactly:
# ADD {{
# "characters": [
# {{"name": "MainGuy-", "seed": 1}},
# {{"name": "Friend1", "seed": 2}},
# {{"name": "Friend2", "seed": 3}}
# ],
# "scene1": {{
# "character": "MainGuy",
# "scene": "Description of the scene",
# "keyframes": [
# {{
# "seed": 1,
# "keyframe1": "First keyframe description",
# "keyframe2": "Second keyframe description"
# }}
# ],
# "camera": "Camera instructions",
# "music": "Music instructions"
# }}
# }}
# Ensure:
# - Each scene corresponds to a line in the script.
# - Assign logical characters.
# - Return only valid JSON (no markdown or explanations).
# Script:
# \"\"\"{script}\"\"\"
# """
# headers = {"Authorization": f"Bearer {OPENROUTER_API_KEY}"}
# payload = {
# "model": MODEL_NAME,
# "messages": [{"role": "user", "content": prompt}],
# "temperature": 0.8,
# "max_tokens": 1500
# }
# async with httpx.AsyncClient(timeout=120) as client:
# response = await client.post(OPENROUTER_URL, json=payload, headers=headers)
# response.raise_for_status()
# data = response.json()
# story_json_str = data["choices"][0]["message"]["content"].strip()
# raw_output = story_json_str
# # --- Cleaning Stage ---
# story_json_str = story_json_str.strip()
# if story_json_str.startswith("ADD"):
# story_json_str = story_json_str[story_json_str.find("{"):]
# # Remove markdown code fences and artifacts
# story_json_str = story_json_str.replace("```json", "").replace("```", "").strip()
# # Remove unwanted triple quotes, trailing commas, and unescaped slashes
# story_json_str = re.sub(r',\s*}', '}', story_json_str)
# story_json_str = re.sub(r',\s*\]', ']', story_json_str)
# story_json_str = story_json_str.replace('\\"', '"').replace("\\'", "'")
# parsed_story = None
# # --- Parsing Stage ---
# try:
# parsed_story = json.loads(story_json_str)
# except json.JSONDecodeError:
# try:
# # Handle double-encoded or escaped JSON
# cleaned_str = bytes(story_json_str, "utf-8").decode("unicode_escape")
# parsed_story = json.loads(cleaned_str)
# except Exception as e:
# logging.error(f"β JSON parse failed after cleaning: {e}")
# parsed_story = None
# logging.info("β
Storyboard generation completed")
# return parsed_story
##best best
# import os
# import asyncio
# import httpx
# import logging
# import json
# import re
# from dotenv import load_dotenv
# dotenv_path = os.path.join(os.path.dirname(os.path.dirname(__file__)), '.env')
# load_dotenv(dotenv_path)
# logging.basicConfig(level=logging.INFO, format="%(asctime)s [%(levelname)s] %(message)s")
# OPENROUTER_API_KEY = os.getenv("OPENROUTER_API_KEY")
# MODEL_NAME = "deepseek/deepseek-r1-distill-llama-70b:free"
# OPENROUTER_URL = "https://openrouter.ai/api/v1/chat/completions"
# async def generate_story(script: str) -> dict:
# """
# Convert the ad script into a structured storyboard JSON using AI.
# Always returns:
# {
# "raw_output": "<original text from AI>",
# "parsed_output": <dict or None>
# }
# """
# prompt = f"""
# You are a professional ad storyboard generator.
# Convert this ad script into a **storyboard JSON** only β no extra text.
# Format example:
# {{
# "characters": [
# {{"name": "MainGuy","Description":"complete decsripiton of the character", "seed": 1}},
# {{"name": "Dog","Description":"complete decsripiton of the character", "seed": 2}}
# ],
# "scene1": {{
# "character": "MainGuy",
# "scene": "Man wakes up late and rushes outside",
# "keyframes": [
# {{"seed": 1, "keyframe1": "Man stepping in puddle", "keyframe2": "Reaction close-up"}}
# ],
# "camera": "Medium shot with soft lighting",
# "music": "Playful upbeat tune"
# }}
# }}
# Script:
# \"\"\"{script}\"\"\"
# Return **only valid JSON**, no markdown or commentary.
# """
# headers = {"Authorization": f"Bearer {OPENROUTER_API_KEY}"}
# payload = {
# "model": MODEL_NAME,
# "messages": [{"role": "user", "content": prompt}],
# "temperature": 0.7,
# "max_tokens": 1500
# }
# async with httpx.AsyncClient(timeout=120) as client:
# response = await client.post(OPENROUTER_URL, json=payload, headers=headers)
# response.raise_for_status()
# data = response.json()
# story_json_str = data["choices"][0]["message"]["content"].strip()
# raw_output = story_json_str
# # --- Clean the output ---
# story_json_str = re.sub(r"^ADD\s*", "", story_json_str)
# story_json_str = story_json_str.replace("```json", "").replace("```", "")
# story_json_str = story_json_str.replace("β", "\"").replace("β", "\"").replace("β", "'")
# story_json_str = re.sub(r",\s*([}\]])", r"\1", story_json_str) # remove trailing commas
# story_json_str = story_json_str.strip()
# # --- Parse the JSON safely ---
# parsed_story = None
# try:
# parsed_story = json.loads(story_json_str)
# except json.JSONDecodeError as e:
# logging.warning(f"JSON parse failed: {e}")
# try:
# # try to find JSON substring in case AI wrapped it with text
# match = re.search(r"\{.*\}", story_json_str, re.DOTALL)
# if match:
# parsed_story = json.loads(match.group(0))
# except Exception as e2:
# logging.error(f"Final parsing failed: {e2}")
# parsed_story = None
# logging.info("β
Storyboard generation completed")
# return parsed_story
import os
import asyncio
import httpx
import logging
import json
import re
from dotenv import load_dotenv
# --- Load environment variables ---
dotenv_path = os.path.join(os.path.dirname(os.path.dirname(__file__)), '.env')
load_dotenv(dotenv_path)
# --- Configure logging ---
logging.basicConfig(level=logging.INFO, format="%(asctime)s [%(levelname)s] %(message)s")
# --- API Constants ---
OPENROUTER_API_KEY = os.getenv("OPENROUTER_API_KEY")
MODEL_NAME = "deepseek/deepseek-r1-distill-llama-70b:free"
OPENROUTER_URL = "https://openrouter.ai/api/v1/chat/completions"
async def generate_story(script: str) -> dict:
"""
Converts ad script into a structured storyboard JSON.
Each scene = only one action.
Reuses character seeds for consistency.
Always returns a non-null dictionary.
"""
prompt = f"""
You are a professional storyboard generator for AI video production.
Convert the given ad script into a JSON storyboard format.
Each scene must represent only ONE clear action or emotional beat.
Do not include any explanations, markdown, or text outside JSON.
### STRICT JSON FORMAT:
{{
"characters": [
{{"name": "Man","description":"average build, brown hair, casual outfit","seed":1}},
{{"name": "Librarian","description":"stern woman, cat-eye glasses, neat bun","seed":2}}
],
"scene1": {{
"character": "Man",
"scene": "Man enters the library holding a bag of potato chips",
"keyframes": [
{{
"seed": 1,
"keyframe1": "Man walking into a quiet library, holding a bag of potato chips, warm sunlight from windows, calm mood",
"keyframe2": "Side shot of man sitting down at a wooden table, casual expression, sunlight glows behind him"
}}
],
"camera": "Medium wide shot with soft natural lighting",
"music": "Gentle ambient tune"
}}
}}
Rules:
- Each scene = ONE clear action only.
- Use each characterβs 'seed' consistently across scenes.
- Each keyframe describes two cinematic angles of that action.
- Keep descriptions detailed but realistic.
- Return valid JSON only β no extra text, comments, or markdown.
Script:
\"\"\"{script}\"\"\"
"""
headers = {"Authorization": f"Bearer {OPENROUTER_API_KEY}"}
payload = {
"model": MODEL_NAME,
"messages": [{"role": "user", "content": prompt}],
"temperature": 0.7,
"max_tokens": 1800
}
async with httpx.AsyncClient(timeout=180) as client:
try:
response = await client.post(OPENROUTER_URL, json=payload, headers=headers)
response.raise_for_status()
data = response.json()
except Exception as e:
logging.error(f"API request failed: {e}")
return {"error": "API request failed", "details": str(e)}
story_json_str = data["choices"][0]["message"]["content"].strip()
# --- Clean the model output ---
story_json_str = re.sub(r"```(?:json)?", "", story_json_str)
story_json_str = story_json_str.replace("β", "\"").replace("β", "\"").replace("β", "'")
story_json_str = re.sub(r",\s*([}\]])", r"\1", story_json_str).strip()
parsed_story = None
try:
parsed_story = json.loads(story_json_str)
except json.JSONDecodeError as e:
logging.warning(f"Initial JSON parse failed: {e}")
match = re.search(r"\{.*\}", story_json_str, re.DOTALL)
if match:
try:
parsed_story = json.loads(match.group(0))
except Exception as e2:
logging.error(f"Fallback parse failed: {e2}")
# --- Final fallback: minimal structure ---
if not parsed_story:
logging.warning("Model output invalid, generating fallback JSON.")
parsed_story = {
"characters": [],
"scene1": {
"character": "Unknown",
"scene": "Failed to parse script properly",
"keyframes": [
{
"seed": 0,
"keyframe1": "Generic placeholder image",
"keyframe2": "Generic placeholder image"
}
],
"camera": "Static fallback frame",
"music": "None"
}
}
# --- Ensure consistent seeds & single-action scenes ---
characters = {c.get("name"): c for c in parsed_story.get("characters", [])}
for key, scene in parsed_story.items():
if not key.startswith("scene"):
continue
char_name = scene.get("character")
seed = characters.get(char_name, {}).get("seed", 0)
desc = characters.get(char_name, {}).get("description", "")
# Limit to one clear action
scene_text = scene.get("scene", "")
scene["scene"] = re.split(r"[,.] and |, then |;| but | while ", scene_text)[0].strip()
# Keyframe corrections
for kf in scene.get("keyframes", []):
kf["seed"] = seed
for i, k in enumerate(["keyframe1", "keyframe2"]):
if not kf.get(k):
angle = "wide shot" if i == 0 else "close-up"
kf[k] = (
f"{char_name} ({desc}) performing '{scene['scene']}', "
f"{angle}, cinematic tone, photorealistic lighting"
)
logging.info("β
Storyboard generated successfully with consistent single-action scenes.")
print(parsed_story)
return parsed_story
|