Spaces:
Running
Running
File size: 20,227 Bytes
c262d1a |
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 |
#!/usr/bin/env python3
"""
Comprehensive Async Batch Logging System for GAIA Questions
Provides detailed per-question logs, batch summary, and classification analysis
"""
import os
import json
import asyncio
import logging
from datetime import datetime
from pathlib import Path
from typing import Dict, List, Optional, Any
from collections import defaultdict
from dataclasses import dataclass, asdict
@dataclass
class QuestionResult:
"""Data class for storing question processing results"""
task_id: str
question_text: str
classification: str
complexity: int
confidence: float
expected_answer: str
our_answer: str
status: str # CORRECT, INCORRECT, PARTIAL, ERROR
accuracy_score: float
total_duration: float
classification_time: float
solving_time: float
validation_time: float
error_type: Optional[str] = None
error_details: Optional[str] = None
tools_used: List[str] = None
anti_hallucination_applied: bool = False
override_reason: Optional[str] = None
def __post_init__(self):
if self.tools_used is None:
self.tools_used = []
class AsyncBatchLogger:
"""Comprehensive logging system for async batch processing"""
def __init__(self, base_log_dir: str = "logs"):
self.base_log_dir = Path(base_log_dir)
self.base_log_dir.mkdir(exist_ok=True)
# Initialize timestamps
self.batch_start_time = datetime.now()
self.timestamp = self.batch_start_time.strftime("%Y%m%d_%H%M%S")
# Create log files
self.summary_log_path = self.base_log_dir / f"async_batch_summary_{self.timestamp}.log"
self.batch_analysis_path = self.base_log_dir / f"async_batch_analysis_{self.timestamp}.json"
# Initialize data structures
self.question_results: Dict[str, QuestionResult] = {}
self.classification_results = defaultdict(list)
self.batch_metrics = {
"total_questions": 0,
"completed_questions": 0,
"correct_answers": 0,
"accuracy_rate": 0.0,
"total_duration": 0.0,
"start_time": self.batch_start_time.isoformat(),
"end_time": None
}
# Initialize summary logger
self.summary_logger = self._setup_summary_logger()
# Active question loggers for concurrent access
self.question_loggers: Dict[str, logging.Logger] = {}
def _setup_summary_logger(self) -> logging.Logger:
"""Set up the batch summary logger"""
logger = logging.getLogger(f"batch_summary_{self.timestamp}")
logger.setLevel(logging.INFO)
# Create file handler
handler = logging.FileHandler(self.summary_log_path)
formatter = logging.Formatter('[%(asctime)s] %(message)s', datefmt='%H:%M:%S')
handler.setFormatter(formatter)
logger.addHandler(handler)
# Also log to console
console_handler = logging.StreamHandler()
console_handler.setFormatter(formatter)
logger.addHandler(console_handler)
return logger
def _setup_question_logger(self, task_id: str) -> logging.Logger:
"""Set up detailed logger for a specific question"""
question_log_path = self.base_log_dir / f"async_batch_question_{task_id}_{self.timestamp}.log"
logger = logging.getLogger(f"question_{task_id}_{self.timestamp}")
logger.setLevel(logging.INFO)
# Create file handler
handler = logging.FileHandler(question_log_path)
formatter = logging.Formatter('%(message)s')
handler.setFormatter(formatter)
logger.addHandler(handler)
return logger
async def log_batch_start(self, total_questions: int, concurrency: int):
"""Log the start of batch processing"""
self.batch_metrics["total_questions"] = total_questions
self.summary_logger.info(f"BATCH_START | Total: {total_questions} questions | Concurrency: {concurrency}")
self.summary_logger.info(f"Timestamp: {self.batch_start_time.isoformat()}")
self.summary_logger.info(f"Log Directory: {self.base_log_dir}")
self.summary_logger.info("-" * 80)
async def log_question_start(self, task_id: str, question_data: Dict):
"""Log the start of processing a specific question"""
# Set up question-specific logger
question_logger = self._setup_question_logger(task_id)
self.question_loggers[task_id] = question_logger
# Log detailed question start
question_logger.info("=" * 80)
question_logger.info("ASYNC BATCH QUESTION PROCESSING")
question_logger.info("=" * 80)
question_logger.info(f"Question ID: {task_id}")
question_logger.info(f"Start Time: {datetime.now().isoformat()}")
question_logger.info(f"Question Text: {question_data.get('question', 'N/A')}")
question_logger.info(f"Level: {question_data.get('Level', 'Unknown')}")
question_logger.info(f"Has File: {'Yes' if question_data.get('file_name') else 'No'}")
if question_data.get('file_name'):
question_logger.info(f"File: {question_data.get('file_name')}")
question_logger.info("")
async def log_classification(self, task_id: str, classification: Dict):
"""Log question classification details"""
if task_id not in self.question_loggers:
return
logger = self.question_loggers[task_id]
logger.info("--- CLASSIFICATION PHASE ---")
logger.info(f"Primary Agent: {classification.get('primary_agent', 'unknown')}")
logger.info(f"Secondary Agents: {', '.join(classification.get('secondary_agents', []))}")
logger.info(f"Complexity: {classification.get('complexity', 0)}/5")
logger.info(f"Confidence: {classification.get('confidence', 0.0):.3f}")
logger.info(f"Tools Needed: {', '.join(classification.get('tools_needed', []))}")
logger.info(f"Reasoning: {classification.get('reasoning', 'N/A')}")
logger.info("")
async def log_solving_start(self, task_id: str, routing_plan: Dict):
"""Log the start of the solving phase"""
if task_id not in self.question_loggers:
return
logger = self.question_loggers[task_id]
logger.info("--- SOLVING PHASE ---")
logger.info(f"Route to: {routing_plan.get('primary_route', 'unknown')} agent")
logger.info(f"Coordination: {'Yes' if routing_plan.get('requires_coordination') else 'No'}")
logger.info(f"Estimated Duration: {routing_plan.get('estimated_duration', 'unknown')}")
logger.info("")
logger.info("Tool Executions:")
async def log_tool_execution(self, task_id: str, tool_name: str, duration: float, result_summary: str):
"""Log individual tool execution"""
if task_id not in self.question_loggers:
return
logger = self.question_loggers[task_id]
logger.info(f" - {tool_name}: {duration:.1f}s โ {result_summary[:100]}...")
async def log_answer_processing(self, task_id: str, raw_response: str, processed_answer: str,
anti_hallucination_applied: bool = False, override_reason: str = None):
"""Log answer processing and anti-hallucination details"""
if task_id not in self.question_loggers:
return
logger = self.question_loggers[task_id]
logger.info("")
logger.info("Agent Response (first 500 chars):")
logger.info(raw_response[:500] + ("..." if len(raw_response) > 500 else ""))
logger.info("")
logger.info(f"Processed Answer: {processed_answer}")
if anti_hallucination_applied:
logger.info(f"๐จ ANTI-HALLUCINATION OVERRIDE APPLIED")
logger.info(f"Reason: {override_reason}")
logger.info("")
async def log_question_complete(self, task_id: str, result: QuestionResult):
"""Log the completion of a question with full results"""
if task_id not in self.question_loggers:
return
logger = self.question_loggers[task_id]
# Store result
self.question_results[task_id] = result
self.classification_results[result.classification].append(result)
# Update batch metrics
self.batch_metrics["completed_questions"] += 1
if result.status == "CORRECT":
self.batch_metrics["correct_answers"] += 1
# Log validation phase
logger.info("--- VALIDATION PHASE ---")
logger.info(f"Expected Answer: {result.expected_answer}")
logger.info(f"Our Answer: {result.our_answer}")
logger.info(f"Status: {result.status}")
logger.info(f"Accuracy Score: {result.accuracy_score:.1%}")
logger.info("")
# Log performance metrics
logger.info("--- PERFORMANCE METRICS ---")
logger.info(f"Total Duration: {result.total_duration:.1f}s")
logger.info(f"Classification Time: {result.classification_time:.1f}s")
logger.info(f"Solving Time: {result.solving_time:.1f}s")
logger.info(f"Validation Time: {result.validation_time:.1f}s")
if result.error_type:
logger.info(f"Error Type: {result.error_type}")
logger.info(f"Error Details: {result.error_details}")
logger.info("")
logger.info("=" * 80)
logger.info("END QUESTION LOG")
logger.info("=" * 80)
# Log to summary
status_emoji = "โ
" if result.status == "CORRECT" else "๐ก" if result.status == "PARTIAL" else "โ"
override_info = f" | {result.override_reason}" if result.anti_hallucination_applied else ""
self.summary_logger.info(
f"{status_emoji} {task_id[:8]}... | {result.classification} | {result.status} | "
f"{result.accuracy_score:.0%} | {result.total_duration:.1f}s{override_info}"
)
async def log_batch_progress(self):
"""Log current batch progress with ETA"""
completed = self.batch_metrics["completed_questions"]
total = self.batch_metrics["total_questions"]
if completed == 0:
return
# Calculate accuracy
accuracy = (self.batch_metrics["correct_answers"] / completed) * 100
# Calculate ETA
elapsed_time = (datetime.now() - self.batch_start_time).total_seconds()
avg_time_per_question = elapsed_time / completed
remaining_questions = total - completed
eta_seconds = remaining_questions * avg_time_per_question
eta_minutes = int(eta_seconds // 60)
eta_seconds = int(eta_seconds % 60)
self.summary_logger.info(
f"๐ PROGRESS | {completed}/{total} completed | {accuracy:.1f}% accuracy | "
f"ETA: {eta_minutes}m {eta_seconds}s"
)
async def log_batch_complete(self):
"""Log batch completion with final summary"""
end_time = datetime.now()
total_duration = (end_time - self.batch_start_time).total_seconds()
# Update batch metrics
self.batch_metrics["end_time"] = end_time.isoformat()
self.batch_metrics["total_duration"] = total_duration
completed = self.batch_metrics["completed_questions"]
total = self.batch_metrics["total_questions"]
accuracy = (self.batch_metrics["correct_answers"] / completed * 100) if completed > 0 else 0
self.batch_metrics["accuracy_rate"] = accuracy / 100
self.summary_logger.info("-" * 80)
self.summary_logger.info(
f"๐ BATCH_COMPLETE | {completed}/{total} | {accuracy:.1f}% accuracy | "
f"Total: {int(total_duration//60)}m {int(total_duration%60)}s"
)
# Generate classification analysis
await self.generate_classification_analysis()
# Export final results
await self.export_results()
self.summary_logger.info(f"๐ Analysis exported: {self.batch_analysis_path}")
self.summary_logger.info(f"๐ Summary log: {self.summary_log_path}")
async def generate_classification_analysis(self):
"""Generate detailed analysis by classification"""
analysis = {
"batch_metadata": self.batch_metrics,
"classification_breakdown": {},
"overall_recommendations": []
}
for classification, results in self.classification_results.items():
if not results:
continue
# Calculate metrics
total = len(results)
correct = len([r for r in results if r.status == "CORRECT"])
partial = len([r for r in results if r.status == "PARTIAL"])
errors = len([r for r in results if r.status == "ERROR"])
accuracy_rate = correct / total if total > 0 else 0
avg_duration = sum(r.total_duration for r in results) / total if total > 0 else 0
# Error analysis
error_types = defaultdict(int)
failed_questions = []
for result in results:
if result.status in ["INCORRECT", "ERROR"]:
error_types[result.error_type or "unknown"] += 1
failed_questions.append({
"task_id": result.task_id,
"error_type": result.error_type,
"error_details": result.error_details
})
# Generate recommendations
recommendations = self._generate_recommendations(classification, results, error_types)
classification_analysis = {
"classification": classification,
"total_questions": total,
"accuracy_rate": accuracy_rate,
"successful": correct,
"partial": partial,
"failed": total - correct - partial,
"errors": errors,
"performance_metrics": {
"avg_duration": avg_duration,
"min_duration": min(r.total_duration for r in results) if results else 0,
"max_duration": max(r.total_duration for r in results) if results else 0
},
"error_breakdown": dict(error_types),
"failed_questions": failed_questions,
"improvement_recommendations": recommendations
}
analysis["classification_breakdown"][classification] = classification_analysis
# Generate overall recommendations
analysis["overall_recommendations"] = self._generate_overall_recommendations()
# Save classification analysis
with open(self.batch_analysis_path, 'w') as f:
json.dump(analysis, f, indent=2, ensure_ascii=False)
def _generate_recommendations(self, classification: str, results: List[QuestionResult],
error_types: Dict[str, int]) -> List[str]:
"""Generate specific recommendations for a classification"""
recommendations = []
accuracy_rate = len([r for r in results if r.status == "CORRECT"]) / len(results)
if accuracy_rate < 0.8:
recommendations.append(f"๐ง Low accuracy ({accuracy_rate:.1%}) - needs immediate attention")
# Classification-specific recommendations
if classification == "multimedia":
if "timeout" in error_types:
recommendations.append("โฑ๏ธ Optimize video processing timeout limits")
if "audio_processing" in error_types:
recommendations.append("๐ต Enhance audio transcription accuracy")
if accuracy_rate > 0.9:
recommendations.append("โ
Excellent multimedia processing - ready for production")
elif classification == "research":
if "hallucination" in error_types:
recommendations.append("๐จ Strengthen anti-hallucination safeguards")
if "wikipedia" in error_types:
recommendations.append("๐ Improve Wikipedia tool integration")
if accuracy_rate > 0.9:
recommendations.append("โ
Excellent research capabilities - ready for production")
elif classification == "logic_math":
if "chess" in error_types:
recommendations.append("โ๏ธ Enhance chess analysis algorithms")
if "calculation" in error_types:
recommendations.append("๐งฎ Improve mathematical calculation accuracy")
if accuracy_rate > 0.9:
recommendations.append("โ
Excellent logic/math processing - ready for production")
elif classification == "file_processing":
if "python_execution" in error_types:
recommendations.append("๐ Optimize Python code execution environment")
if "excel_processing" in error_types:
recommendations.append("๐ Enhance Excel file processing capabilities")
if accuracy_rate > 0.9:
recommendations.append("โ
Excellent file processing - ready for production")
# Performance recommendations
avg_duration = sum(r.total_duration for r in results) / len(results)
if avg_duration > 60:
recommendations.append(f"โก Optimize performance - avg duration {avg_duration:.1f}s")
return recommendations
def _generate_overall_recommendations(self) -> List[str]:
"""Generate overall system recommendations"""
recommendations = []
total_accuracy = self.batch_metrics["accuracy_rate"]
if total_accuracy >= 0.95:
recommendations.append("๐ EXCELLENT: 95%+ accuracy achieved - production ready!")
elif total_accuracy >= 0.90:
recommendations.append("โ
GREAT: 90%+ accuracy - minor optimizations needed")
elif total_accuracy >= 0.80:
recommendations.append("๐ง GOOD: 80%+ accuracy - moderate improvements needed")
elif total_accuracy >= 0.70:
recommendations.append("โ ๏ธ ACCEPTABLE: 70%+ accuracy - significant improvements needed")
else:
recommendations.append("๐จ CRITICAL: <70% accuracy - major system overhaul required")
# Add specific system recommendations
recommendations.extend([
"๐ Monitor performance metrics for production deployment",
"๐ Implement continuous improvement based on classification analysis",
"๐ Track accuracy trends over time",
"๐ ๏ธ Focus improvement efforts on lowest-performing classifications"
])
return recommendations
async def export_results(self):
"""Export comprehensive results for analysis"""
# Export individual question results
results_data = {
"batch_metadata": self.batch_metrics,
"question_results": [asdict(result) for result in self.question_results.values()],
"classification_summary": {
classification: {
"count": len(results),
"accuracy": len([r for r in results if r.status == "CORRECT"]) / len(results)
}
for classification, results in self.classification_results.items()
}
}
results_file = self.base_log_dir / f"async_batch_results_{self.timestamp}.json"
with open(results_file, 'w') as f:
json.dump(results_data, f, indent=2, ensure_ascii=False)
self.summary_logger.info(f"๐ Detailed results: {results_file}") |