File size: 18,100 Bytes
2ed8996 | 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 | """
Evaluation tasks for AegisLM SaaS Backend.
Production-ready Celery tasks for AI evaluation processing,
background job execution, and result storage.
"""
import sys
import os
from datetime import datetime
from typing import Dict, Any, Optional
from celery import current_task
from sqlalchemy.ext.asyncio import AsyncSession, create_async_engine
from sqlalchemy.orm import sessionmaker
# Add the backend directory to Python path
sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
from workers.celery_worker import celery_app
from core.config import settings
from core.database import get_redis
from db_models.evaluation import Evaluation, EvaluationStatus
from schemas.evaluation_schema import EvaluationCreate
# Try to import AI pipeline components (now works with default path fix)
try:
from ai.pipelines.redteam_pipeline import run_redteam_pipeline, create_pipeline_config
from ai.models import PipelineConfig as AIPipelineConfig
AI_PIPELINE_AVAILABLE = True
except ImportError as e:
AI_PIPELINE_AVAILABLE = False
print(f"Warning: AI pipeline not available: {str(e)}")
# Database session for tasks
AsyncSessionLocal = sessionmaker(
bind=create_async_engine(settings.DATABASE_URL),
class_=AsyncSession,
expire_on_commit=False
)
class ModelInterface:
"""
Model interface for AI evaluation.
Implements the interface expected by the red team pipeline.
"""
def __init__(self, model_config: Dict[str, Any]):
"""
Initialize model interface.
Args:
model_config: Model configuration
"""
self.model_name = model_config.get("model_name", "unknown")
self.model_type = model_config.get("model_type", "api")
self.api_endpoint = model_config.get("api_endpoint")
self.api_key = model_config.get("api_key")
self.temperature = model_config.get("temperature", 0.7)
self.max_tokens = model_config.get("max_tokens", 1000)
self.custom_parameters = model_config.get("custom_parameters", {})
async def generate_response(self, prompt: str) -> str:
"""
Produce a real-time response using the AegisLM LLM Router.
"""
from ai.llm_router.llm_router import get_llm_router
router = get_llm_router()
try:
# Route to the actual model provider (Groq/OpenAI/Mistral)
response = await router.route_and_generate(
model=self.model_name,
prompt=prompt,
temperature=self.temperature,
max_tokens=self.max_tokens
)
return response.text
except Exception as e:
logger.error(f"Router failure: {str(e)}")
return f"Error: Model {self.model_name} failed to respond: {str(e)}"
@celery_app.task(bind=True, name="tasks.evaluation_task.run_evaluation_task")
def run_evaluation_task(self, job_id: str):
"""
Run evaluation task for given job ID.
Args:
job_id: Evaluation job ID
Returns:
dict: Task result
"""
async def _run_evaluation():
async with AsyncSessionLocal() as db:
try:
# Get evaluation from database
result = await db.execute(
select(Evaluation).where(Evaluation.job_id == job_id)
)
evaluation = result.scalar_one_or_none()
if not evaluation:
raise ValueError(f"Evaluation not found: {job_id}")
# Update status to running
evaluation.update_status(EvaluationStatus.RUNNING)
await db.commit()
# Create model interface
model = ModelInterface(evaluation.model_config)
# Create pipeline configuration
pipeline_config_dict = evaluation.pipeline_config
if AI_PIPELINE_AVAILABLE:
pipeline_config = create_pipeline_config(
max_iterations=pipeline_config_dict.get("max_iterations", 5),
attack_types=pipeline_config_dict.get("attack_types"),
num_prompts=pipeline_config_dict.get("num_prompts", 10),
mutation_enabled=pipeline_config_dict.get("mutation_enabled", True),
weights=pipeline_config_dict.get("weights")
)
else:
# Use REAL pipeline config with actual AI models if available
try:
from ai.models import PipelineConfig as AIPipelineConfig
pipeline_config = AIPipelineConfig(
max_iterations=pipeline_config_dict.get("max_iterations", 5),
attack_types=pipeline_config_dict.get("attack_types"),
num_prompts=pipeline_config_dict.get("num_prompts", 10),
mutation_enabled=pipeline_config_dict.get("mutation_enabled", True),
weights=pipeline_config_dict.get("weights"),
use_real_models=True, # Enable real model usage
fallback_to_simulation=True # Allow fallback if needed
)
except ImportError:
# Final fallback to basic config
pipeline_config = type('PipelineConfig', (), {
'max_iterations': pipeline_config_dict.get("max_iterations", 5),
'attack_types': pipeline_config_dict.get("attack_types", []),
'num_prompts': pipeline_config_dict.get("num_prompts", 10),
'mutation_enabled': pipeline_config_dict.get("mutation_enabled", True),
'weights': pipeline_config_dict.get("weights", {})
})()
# Run red team pipeline
start_time = datetime.utcnow()
if AI_PIPELINE_AVAILABLE:
# INTEGRATION: Awaiting the production async pipeline
pipeline_result = await run_redteam_pipeline(model, pipeline_config, job_id)
else:
# Use simplified result if AI pipeline not available
class SimplifiedPipelineResult:
def __init__(self):
self.pipeline_id = job_id
self.total_attacks = 0
self.successful_attacks = []
self.failed_attacks = []
self.success_rate = 0.0
self.execution_time_ms = 1000
def dict(self):
return {
"pipeline_id": self.pipeline_id,
"total_attacks": self.total_attacks,
"successful_attacks": self.successful_attacks,
"failed_attacks": self.failed_attacks,
"success_rate": self.success_rate,
"execution_time_ms": self.execution_time_ms
}
pipeline_result = SimplifiedPipelineResult()
end_time = datetime.utcnow()
# Calculate metrics
execution_time_ms = int((end_time - start_time).total_seconds() * 1000)
success_rate = pipeline_result.success_rate * 100 if pipeline_result.success_rate else 0.0
# Update evaluation with results
evaluation.status = EvaluationStatus.COMPLETED
evaluation.result_json = pipeline_result.dict()
evaluation.total_attacks = pipeline_result.total_attacks
evaluation.successful_attacks = len(pipeline_result.successful_attacks)
evaluation.success_rate = pipeline_result.success_rate # Store as float (0.0 to 1.0)
evaluation.execution_time_ms = execution_time_ms
evaluation.completed_at = end_time
await db.commit()
# Store result in Redis for quick access
redis_client = await get_redis()
result_key = f"evaluation_result:{job_id}"
await redis_client.setex(
result_key,
3600, # 1 hour expiry
str(pipeline_result.dict())
)
return {
"job_id": job_id,
"status": "completed",
"execution_time_ms": execution_time_ms,
"total_attacks": pipeline_result.total_attacks,
"successful_attacks": len(pipeline_result.successful_attacks),
"success_rate": success_rate,
"completed_at": end_time.isoformat()
}
except Exception as e:
# Update evaluation status to failed
if evaluation:
evaluation.update_status(EvaluationStatus.FAILED, str(e))
await db.commit()
# Log error
print(f"Evaluation task failed for job {job_id}: {str(e)}")
return {
"job_id": job_id,
"status": "failed",
"error": str(e),
"failed_at": datetime.utcnow().isoformat()
}
# Run the async function
import asyncio
return asyncio.run(_run_evaluation())
@celery_app.task(bind=True, name="tasks.evaluation_task.run_benchmark_task")
def run_benchmark_task(self, benchmark_id: str, job_ids: list):
"""
Run benchmark task for multiple evaluation jobs.
Args:
benchmark_id: Benchmark ID
job_ids: List of evaluation job IDs
Returns:
dict: Benchmark task result
"""
async def _run_benchmark():
try:
# Trigger individual evaluation tasks
task_results = []
for job_id in job_ids:
# Chain the evaluation task
task = run_evaluation_task.delay(job_id)
task_results.append({
"job_id": job_id,
"task_id": task.id,
"status": "queued"
})
# Store benchmark progress in Redis
redis_client = await get_redis()
benchmark_key = f"benchmark:{benchmark_id}"
benchmark_data = {
"benchmark_id": benchmark_id,
"job_ids": job_ids,
"task_ids": [result["task_id"] for result in task_results],
"status": "running",
"started_at": datetime.utcnow().isoformat(),
"total_jobs": len(job_ids),
"completed_jobs": 0,
"failed_jobs": 0
}
await redis_client.setex(
benchmark_key,
24 * 3600, # 24 hours expiry
str(benchmark_data)
)
return {
"benchmark_id": benchmark_id,
"status": "started",
"total_jobs": len(job_ids),
"task_results": task_results
}
except Exception as e:
print(f"Benchmark task failed for {benchmark_id}: {str(e)}")
return {
"benchmark_id": benchmark_id,
"status": "failed",
"error": str(e),
"failed_at": datetime.utcnow().isoformat()
}
# Run the async function
import asyncio
return asyncio.run(_run_benchmark())
@celery_app.task(name="tasks.evaluation_task.cleanup_expired_results")
def cleanup_expired_results():
"""
Cleanup expired evaluation results from Redis.
Returns:
dict: Cleanup results
"""
async def _cleanup():
try:
redis_client = await get_redis()
# Get all evaluation result keys
result_keys = await redis_client.keys("evaluation_result:*")
cleaned_count = 0
for key in result_keys:
# Check if key exists and is expired
ttl = await redis_client.ttl(key)
if ttl == -1: # No expiry set, set one
await redis_client.expire(key, 3600) # 1 hour
elif ttl == -2: # Key expired
cleaned_count += 1
# Also cleanup benchmark data
benchmark_keys = await redis_client.keys("benchmark:*")
for key in benchmark_keys:
ttl = await redis_client.ttl(key)
if ttl == -1: # No expiry set, set one
await redis_client.expire(key, 24 * 3600) # 24 hours
elif ttl == -2: # Key expired
cleaned_count += 1
return {
"status": "completed",
"timestamp": datetime.utcnow().isoformat(),
"cleaned_keys": cleaned_count,
"total_result_keys": len(result_keys),
"total_benchmark_keys": len(benchmark_keys)
}
except Exception as e:
print(f"Cleanup task failed: {str(e)}")
return {
"status": "failed",
"error": str(e),
"timestamp": datetime.utcnow().isoformat()
}
# Run the async function
import asyncio
return asyncio.run(_cleanup())
@celery_app.task(name="tasks.evaluation_task.update_evaluation_progress")
def update_evaluation_progress(job_id: str, progress: float):
"""
Update evaluation progress.
Args:
job_id: Evaluation job ID
progress: Progress percentage (0-100)
Returns:
dict: Update result
"""
async def _update_progress():
try:
redis_client = await get_redis()
progress_key = f"evaluation_progress:{job_id}"
progress_data = {
"job_id": job_id,
"progress": progress,
"updated_at": datetime.utcnow().isoformat()
}
await redis_client.setex(
progress_key,
3600, # 1 hour expiry
str(progress_data)
)
return {
"job_id": job_id,
"progress": progress,
"updated_at": progress_data["updated_at"]
}
except Exception as e:
print(f"Progress update failed for {job_id}: {str(e)}")
return {
"job_id": job_id,
"error": str(e)
}
# Run the async function
import asyncio
return asyncio.run(_update_progress())
@celery_app.task(name="tasks.evaluation_task.cancel_evaluation_task")
def cancel_evaluation_task(job_id: str):
"""
Cancel evaluation task.
Args:
job_id: Evaluation job ID
Returns:
dict: Cancellation result
"""
async def _cancel_evaluation():
async with AsyncSessionLocal() as db:
try:
# Get evaluation from database
result = await db.execute(
select(Evaluation).where(Evaluation.job_id == job_id)
)
evaluation = result.scalar_one_or_none()
if not evaluation:
raise ValueError(f"Evaluation not found: {job_id}")
# Update status to cancelled
evaluation.update_status(EvaluationStatus.CANCELLED)
await db.commit()
# Remove from Redis if exists
redis_client = await get_redis()
result_key = f"evaluation_result:{job_id}"
await redis_client.delete(result_key)
progress_key = f"evaluation_progress:{job_id}"
await redis_client.delete(progress_key)
return {
"job_id": job_id,
"status": "cancelled",
"cancelled_at": datetime.utcnow().isoformat()
}
except Exception as e:
print(f"Cancellation failed for {job_id}: {str(e)}")
return {
"job_id": job_id,
"error": str(e)
}
# Run the async function
import asyncio
return asyncio.run(_cancel_evaluation())
# Helper function to check task status
@celery_app.task(name="tasks.evaluation_task.check_task_status")
def check_task_status(task_id: str):
"""
Check status of a Celery task.
Args:
task_id: Celery task ID
Returns:
dict: Task status
"""
try:
task = celery_app.AsyncResult(task_id)
return {
"task_id": task_id,
"status": task.status,
"result": task.result if task.ready() else None,
"traceback": task.traceback if task.failed() else None,
"date_done": task.date_done.isoformat() if task.date_done else None
}
except Exception as e:
return {
"task_id": task_id,
"error": str(e)
}
|