Spaces:
Running
Running
File size: 10,580 Bytes
6070720 a0199e2 6070720 a0199e2 6070720 a0199e2 6070720 a0199e2 6070720 a0199e2 6070720 a0199e2 6070720 a0199e2 6070720 a0199e2 6070720 a0199e2 6070720 a0199e2 6070720 a0199e2 6070720 a0199e2 6070720 a0199e2 6070720 | 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 | import os
import base64
import time
import requests
from typing import List, Optional, Tuple
from runwayml import RunwayML
import mimetypes
from urllib.parse import urlparse
def encode_image_to_data_uri(image_path: str) -> str:
"""Convert a local image file to a data URI."""
mime_type, _ = mimetypes.guess_type(image_path)
if not mime_type or not mime_type.startswith('image/'):
raise ValueError(f"Unsupported image type for {image_path}")
with open(image_path, 'rb') as image_file:
encoded_string = base64.b64encode(image_file.read()).decode('utf-8')
return f"data:{mime_type};base64,{encoded_string}"
def save_generated_image(image_url: str, filename: str = None, batch_folder: str = None) -> str:
"""
Download and save the generated image to a timestamped batch folder.
Args:
image_url: URL of the generated image
filename: Optional filename (auto-generated if not provided)
batch_folder: Optional batch folder name (auto-generated with timestamp if not provided)
Returns:
Path to the saved image file
"""
if not batch_folder:
timestamp = time.strftime("%Y%m%d_%H%M%S")
batch_folder = f"batch_{timestamp}"
output_dir = os.path.join("output", batch_folder)
os.makedirs(output_dir, exist_ok=True)
if not filename:
timestamp = int(time.time())
filename = f"generated_{timestamp}.jpg"
if not os.path.splitext(filename)[1]:
filename += ".jpg"
output_path = os.path.join(output_dir, filename)
response = requests.get(image_url)
response.raise_for_status()
with open(output_path, 'wb') as f:
f.write(response.content)
return output_path
def generate_image_with_references(
prompt_text: str,
reference_image_paths: List[str],
ratio: str = "1920:1080",
model: str = "gen4_image",
seed: Optional[int] = None,
api_key: Optional[str] = None,
auto_tag_prompt: bool = True
) -> str:
"""
Generate an image using RunwayML API with reference images.
Args:
prompt_text: Description of the image to generate (max 1000 characters)
reference_image_paths: List of local image file paths to use as references
ratio: Output image resolution (default: "1920:1080")
model: Model to use (default: "gen4_image")
seed: Optional seed for reproducible results
api_key: Optional API key (uses RUNWAYML_API_SECRET env var if not provided)
auto_tag_prompt: Whether to automatically append tags to prompt (default: True)
Returns:
Task ID for the generation request
"""
client = RunwayML(api_key=api_key or os.environ.get("RUNWAYML_API_SECRET"))
if len(reference_image_paths) > 3:
raise ValueError("Maximum 3 reference images allowed")
if len(prompt_text) > 1000:
raise ValueError("Prompt text must be 1000 characters or less")
reference_images = []
tags = []
used_standard_tags = set()
for i, image_path in enumerate(reference_image_paths):
if not os.path.exists(image_path):
raise FileNotFoundError(f"Image file not found: {image_path}")
filename = os.path.splitext(os.path.basename(image_path))[0]
path_parts = image_path.split(os.sep)
tag = None
for part in path_parts:
if part == 'characters' and 'character' not in used_standard_tags:
tag = 'character'
used_standard_tags.add('character')
break
elif part == 'scenes' and 'scene' not in used_standard_tags:
tag = 'scene'
used_standard_tags.add('scene')
break
elif part == 'styles' and 'style' not in used_standard_tags:
tag = 'style'
used_standard_tags.add('style')
break
if not tag:
tag = f"ref_{filename}".replace('-', '_').replace(' ', '_')[:16]
tag = ''.join(c for c in tag if c.isalnum() or c == '_')
if not tag[0].isalpha():
tag = f"img_{tag}"
tag = tag[:16]
tags.append(tag)
data_uri = encode_image_to_data_uri(image_path)
reference_images.append({"uri": data_uri, "tag": tag})
final_prompt = prompt_text
if auto_tag_prompt and tags:
tag_mentions = " ".join([f"@{tag}" for tag in tags])
final_prompt = f"{prompt_text} using references: {tag_mentions}"
if len(final_prompt) > 1000:
tag_mentions = " ".join([f"@{tag}" for tag in tags])
final_prompt = f"{prompt_text} {tag_mentions}"
if len(final_prompt) > 1000:
available_chars = 1000 - len(tag_mentions) - 1
final_prompt = f"{prompt_text[:available_chars]} {tag_mentions}"
print(f"Using tags: {tags}")
if auto_tag_prompt:
print(f"Auto-tagged prompt: {final_prompt}")
else:
print(f"Manual tagging mode - use @{', @'.join(tags)} in your prompt")
print(f"Original prompt: {final_prompt}")
create_params = {
"model": model,
"prompt_text": final_prompt,
"ratio": ratio,
"reference_images": reference_images
}
if seed is not None:
create_params["seed"] = seed
task = client.text_to_image.create(**create_params)
return task.id
def check_task_status(task_id: str, api_key: Optional[str] = None):
"""
Check the status of a generation task.
Args:
task_id: The task ID returned from generate_image_with_references
api_key: Optional API key (uses RUNWAYML_API_SECRET env var if not provided)
Returns:
Task details including status and output URLs if completed
"""
client = RunwayML(api_key=api_key or os.environ.get("RUNWAYML_API_SECRET"))
return client.tasks.retrieve(id=task_id)
def generate_and_wait_for_result(
prompt_text: str,
reference_image_paths: List[str],
ratio: str = "1920:1080",
model: str = "gen4_image",
seed: Optional[int] = None,
api_key: Optional[str] = None,
filename: str = None,
batch_folder: str = None,
max_retries: int = 8,
wait_interval: int = 15,
auto_tag_prompt: bool = True
) -> Tuple[str, str]:
"""
Generate an image and wait for completion with automatic retries.
Args:
prompt_text: Description of the image to generate
reference_image_paths: List of local image file paths to use as references
ratio: Output image resolution
model: Model to use
seed: Optional seed for reproducible results
api_key: Optional API key
filename: Optional filename for saved image
max_retries: Maximum number of status checks (default: 8)
wait_interval: Seconds to wait between checks (default: 15)
auto_tag_prompt: Whether to automatically append tags to prompt
Returns:
Tuple of (task_id, saved_image_path)
"""
task_id = generate_image_with_references(
prompt_text=prompt_text,
reference_image_paths=reference_image_paths,
ratio=ratio,
model=model,
seed=seed,
api_key=api_key,
auto_tag_prompt=auto_tag_prompt
)
print(f"Image generation started. Task ID: {task_id}")
print(f"Checking status every {wait_interval} seconds (max {max_retries} attempts)...")
for attempt in range(max_retries):
print(f"Attempt {attempt + 1}/{max_retries} - Waiting {wait_interval} seconds...")
time.sleep(wait_interval)
try:
status = check_task_status(task_id, api_key)
print(f"Status: {status.status}")
if status.status == "SUCCEEDED":
if hasattr(status, 'output') and status.output:
image_url = status.output[0]
print(f"Generation completed! Image URL: {image_url}")
saved_path = save_generated_image(image_url, filename, batch_folder)
print(f"Image saved to: {saved_path}")
return task_id, saved_path
else:
print("Task succeeded but no output found")
return task_id, None
elif status.status == "FAILED":
print("Task failed")
return task_id, None
elif status.status in ["PENDING", "RUNNING"]:
print("Task still in progress...")
continue
except Exception as e:
print(f"Error checking status: {e}")
if attempt == max_retries - 1:
print("Max retries reached. Task may still be processing.")
return task_id, None
print(f"Timeout after {max_retries} attempts. Task may still be processing.")
print(f"You can manually check status later using task ID: {task_id}")
return task_id, None
def main():
print("\n=== Testing RunwayML with Reference Images ===")
reference_images = [
"assets/characters/japanese_guy.jpg",
"assets/scenes/f1-fields.jpg",
"assets/styles/f1-cockpit.jpg"
]
print("=== Manual Tagging Mode ===")
manual_prompt = "@character in a @scene with @style composition, cinematic lighting, high detail"
try:
task_id, saved_path = generate_and_wait_for_result(
prompt_text=manual_prompt,
reference_image_paths=reference_images,
ratio="1920:1080",
filename="f1_driver_manual_tags.jpg",
auto_tag_prompt=False
)
if saved_path:
print(f"Manual tagging success! Image saved to: {saved_path}")
else:
print(f"Manual tagging incomplete. Task ID: {task_id}")
except Exception as e:
print(f"Manual tagging error: {e}")
print("\n" + "="*50)
print("=== Auto Tagging Mode Example ===")
auto_prompt = "A Japanese F1 driver in a cockpit style setting on a racing field, cinematic lighting, high detail"
try:
task_id, saved_path = generate_and_wait_for_result(
prompt_text=auto_prompt,
reference_image_paths=reference_images,
ratio="1920:1080",
filename="f1_driver_auto_tags.jpg",
auto_tag_prompt=True
)
if saved_path:
print(f"Auto tagging success! Image saved to: {saved_path}")
else:
print(f"Auto tagging incomplete. Task ID: {task_id}")
except Exception as e:
print(f"Auto tagging error: {e}")
if __name__ == "__main__":
main() |