File size: 7,120 Bytes
9abace2 | 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 | """
Image Generation
Part of SOVEREIGN PYTHON LLM ENGINE
Generate images using various providers (DALL-E, Stable Diffusion, etc.).
"""
from typing import Any
from pathlib import Path
from dataclasses import dataclass
import base64
from io import BytesIO
from ...core.evidence import WORMLedger
@dataclass
class GeneratedImage:
"""Generated image result"""
image_bytes: bytes
format: str # "png", "jpeg", etc.
size: tuple[int, int]
prompt: str
model: str
metadata: dict[str, Any]
class ImageGenerator:
"""
Image generation using multiple providers.
Supports:
- OpenAI DALL-E 3
- Stable Diffusion (via API or local)
- Amazon Bedrock (Stable Diffusion)
"""
def __init__(
self,
provider: str = "openai",
model: str | None = None,
worm_ledger: WORMLedger | None = None
):
"""
Initialize image generator.
Args:
provider: Provider name ("openai", "bedrock", "stability")
model: Optional model override
worm_ledger: Optional WORM ledger
"""
self.provider = provider
self.model = model
self.worm_ledger = worm_ledger
async def generate(
self,
prompt: str,
size: str = "1024x1024",
quality: str = "standard",
n: int = 1
) -> list[GeneratedImage]:
"""
Generate images from prompt.
Args:
prompt: Text prompt
size: Image size ("1024x1024", "1792x1024", etc.)
quality: Quality level ("standard", "hd")
n: Number of images
Returns:
List of GeneratedImage
"""
if self.provider == "openai":
return await self._generate_openai(prompt, size, quality, n)
elif self.provider == "bedrock":
return await self._generate_bedrock(prompt, size)
elif self.provider == "stability":
return await self._generate_stability(prompt, size)
else:
raise ValueError(f"Unknown provider: {self.provider}")
async def _generate_openai(
self,
prompt: str,
size: str,
quality: str,
n: int
) -> list[GeneratedImage]:
"""Generate using OpenAI DALL-E"""
try:
from openai import AsyncOpenAI
except ImportError:
raise ImportError("openai is required. Install: pip install openai")
client = AsyncOpenAI()
response = await client.images.generate(
model=self.model or "dall-e-3",
prompt=prompt,
size=size,
quality=quality,
n=n,
response_format="b64_json"
)
results = []
for image_data in response.data:
# Decode base64
image_bytes = base64.b64decode(image_data.b64_json)
# Parse size
width, height = map(int, size.split("x"))
results.append(GeneratedImage(
image_bytes=image_bytes,
format="png",
size=(width, height),
prompt=prompt,
model=self.model or "dall-e-3",
metadata={
"revised_prompt": image_data.revised_prompt
}
))
# Log to WORM
if self.worm_ledger:
await self.worm_ledger.append({
"event": "image_generate",
"provider": "openai",
"prompt": prompt,
"images_count": len(results)
})
return results
async def _generate_bedrock(
self,
prompt: str,
size: str
) -> list[GeneratedImage]:
"""Generate using AWS Bedrock (Stable Diffusion)"""
try:
import boto3
import json
except ImportError:
raise ImportError("boto3 is required")
client = boto3.client("bedrock-runtime", region_name="us-east-1")
# Parse size
width, height = map(int, size.split("x"))
# Build request
request_body = {
"text_prompts": [{"text": prompt}],
"cfg_scale": 7,
"steps": 30,
"width": width,
"height": height
}
# Invoke model
response = client.invoke_model(
modelId=self.model or "stability.stable-diffusion-xl-v1",
body=json.dumps(request_body)
)
# Parse response
response_body = json.loads(response["body"].read())
results = []
for artifact in response_body.get("artifacts", []):
image_bytes = base64.b64decode(artifact["base64"])
results.append(GeneratedImage(
image_bytes=image_bytes,
format="png",
size=(width, height),
prompt=prompt,
model=self.model or "stability.stable-diffusion-xl-v1",
metadata={}
))
# Log to WORM
if self.worm_ledger:
await self.worm_ledger.append({
"event": "image_generate",
"provider": "bedrock",
"prompt": prompt,
"images_count": len(results)
})
return results
async def _generate_stability(
self,
prompt: str,
size: str
) -> list[GeneratedImage]:
"""Generate using Stability AI API"""
try:
import httpx
except ImportError:
raise ImportError("httpx is required")
# This is a stub - would need Stability API key
raise NotImplementedError("Stability AI provider not fully implemented")
async def save_image(image: GeneratedImage, output_path: Path) -> None:
"""
Save generated image to file.
Args:
image: Generated image
output_path: Where to save
"""
output_path.write_bytes(image.image_bytes)
# Tool registration helper
async def generate_image_tool(
prompt: str,
provider: str = "bedrock",
size: str = "1024x1024",
output_path: str | None = None
) -> dict:
"""
Tool wrapper for image generation.
Args:
prompt: Image prompt
provider: Provider ("openai", "bedrock")
size: Image size
output_path: Optional path to save
Returns:
Dictionary with result
"""
generator = ImageGenerator(provider=provider)
images = await generator.generate(prompt, size=size)
result = {
"images_count": len(images),
"prompt": prompt,
"size": size,
"model": images[0].model if images else None
}
# Save if requested
if output_path and images:
await save_image(images[0], Path(output_path))
result["saved_to"] = output_path
return result
|