File size: 13,720 Bytes
5e4510c |
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 |
#!/usr/bin/env python3
"""
AlgoTune Benchmark Script for OpenEvolve
This script runs all AlgoTune tasks in the examples/algotune folder sequentially,
measures speedup achieved by OpenEvolve, and outputs detailed JSON results.
Usage:
python examples/algotune/run_benchmark.py
python examples/algotune/run_benchmark.py --iterations 50 --timeout 3600
python examples/algotune/run_benchmark.py --tasks eigenvectors_complex,psd_cone_projection
"""
import argparse
import json
import os
import subprocess
import sys
import time
import traceback
from datetime import datetime
from pathlib import Path
from typing import Dict, List, Optional, Any
import statistics
def parse_args() -> argparse.Namespace:
"""Parse command-line arguments."""
parser = argparse.ArgumentParser(
description="Benchmark all AlgoTune tasks with OpenEvolve",
formatter_class=argparse.RawDescriptionHelpFormatter,
epilog="""
Examples:
python examples/algotune/run_benchmark.py
python examples/algotune/run_benchmark.py --iterations 50 --timeout 3600
python examples/algotune/run_benchmark.py --tasks eigenvectors_complex,psd_cone_projection
python examples/algotune/run_benchmark.py --output my_results.json
"""
)
parser.add_argument(
"--iterations", "-i",
type=int,
default=100,
help="Number of iterations to run for each task (default: 100)"
)
parser.add_argument(
"--timeout", "-t",
type=int,
default=7200, # 2 hours
help="Timeout in seconds for each task (default: 7200)"
)
parser.add_argument(
"--tasks",
type=str,
help="Comma-separated list of specific tasks to run (default: all tasks)"
)
parser.add_argument(
"--output", "-o",
type=str,
help="Output JSON file path (default: algotune_benchmark_TIMESTAMP.json)"
)
parser.add_argument(
"--config",
type=str,
help="Custom config file to use instead of task default configs"
)
parser.add_argument(
"--verbose", "-v",
action="store_true",
help="Enable verbose output"
)
return parser.parse_args()
def discover_algotune_tasks(algotune_dir: Path) -> List[str]:
"""
Discover all AlgoTune task directories.
Args:
algotune_dir: Path to examples/algotune directory
Returns:
List of task names (directory names)
"""
tasks = []
# Look for directories that contain the required files
for item in algotune_dir.iterdir():
if item.is_dir() and item.name not in ['__pycache__']:
# Check if it has the required files
required_files = ['initial_program.py', 'evaluator.py', 'config.yaml']
if all((item / file).exists() for file in required_files):
tasks.append(item.name)
return sorted(tasks)
def run_openevolve_task(
task_name: str,
task_dir: Path,
iterations: int,
timeout: int,
custom_config: Optional[str] = None,
verbose: bool = False
) -> Dict[str, Any]:
"""
Run OpenEvolve on a single AlgoTune task.
Args:
task_name: Name of the task
task_dir: Path to the task directory
iterations: Number of iterations to run
timeout: Timeout in seconds
custom_config: Optional custom config file path
verbose: Enable verbose output
Returns:
Dictionary containing task results
"""
print(f"Starting task: {task_name}")
start_time = time.time()
# Prepare command
initial_program = task_dir / "initial_program.py"
evaluator = task_dir / "evaluator.py"
config = Path(custom_config) if custom_config else (task_dir / "config.yaml")
# Get the project root (parent of examples/algotune)
project_root = Path(__file__).parent.parent.parent
openevolve_script = project_root / "openevolve-run.py"
cmd = [
sys.executable, str(openevolve_script),
str(initial_program),
str(evaluator),
"--config", str(config),
"--iterations", str(iterations)
]
if verbose:
print(f"Running command: {' '.join(cmd)}")
try:
# Run OpenEvolve
result = subprocess.run(
cmd,
cwd=task_dir, # Run from task directory
capture_output=True,
text=True,
timeout=timeout
)
end_time = time.time()
runtime_seconds = end_time - start_time
if verbose:
print(f"stdout: {result.stdout}")
if result.stderr:
print(f"stderr: {result.stderr}")
# Parse results from output directory
task_results = parse_task_results(task_dir, task_name, runtime_seconds)
task_results["command_output"] = {
"returncode": result.returncode,
"stdout": result.stdout,
"stderr": result.stderr
}
if result.returncode == 0:
print(f"β
{task_name} completed successfully in {runtime_seconds:.1f}s")
else:
print(f"β {task_name} failed with return code {result.returncode}")
task_results["status"] = "failed"
task_results["error"] = f"Process failed with return code {result.returncode}"
except subprocess.TimeoutExpired:
end_time = time.time()
runtime_seconds = end_time - start_time
print(f"β° {task_name} timed out after {timeout}s")
task_results = {
"status": "timeout",
"error": f"Task timed out after {timeout} seconds",
"runtime_seconds": runtime_seconds,
"speedup": None
}
except Exception as e:
end_time = time.time()
runtime_seconds = end_time - start_time
print(f"π₯ {task_name} crashed: {str(e)}")
task_results = {
"status": "error",
"error": str(e),
"traceback": traceback.format_exc(),
"runtime_seconds": runtime_seconds,
"speedup": None
}
return task_results
def parse_task_results(task_dir: Path, task_name: str, runtime_seconds: float) -> Dict[str, Any]:
"""
Parse OpenEvolve results from the output directory.
Args:
task_dir: Path to the task directory
task_name: Name of the task
runtime_seconds: Runtime in seconds
Returns:
Dictionary containing parsed results
"""
# Look for openevolve_output directory
output_dir = task_dir / "openevolve_output"
best_dir = output_dir / "best"
if not best_dir.exists():
return {
"status": "failed",
"error": "No openevolve_output/best directory found",
"runtime_seconds": runtime_seconds,
"speedup": None
}
# Try to read best_program_info.json
info_file = best_dir / "best_program_info.json"
if info_file.exists():
try:
with open(info_file, 'r') as f:
info = json.load(f)
# Extract key metrics
speedup = info.get("speedup_score", info.get("score"))
if speedup is None:
# Try to find speedup in evaluation_result
eval_result = info.get("evaluation_result", {})
speedup = eval_result.get("speedup_score", eval_result.get("score"))
return {
"status": "success",
"iterations_run": info.get("iteration", "unknown"),
"best_iteration": info.get("iteration", "unknown"),
"speedup": speedup,
"score": info.get("score"),
"evaluation_result": info.get("evaluation_result", {}),
"runtime_seconds": runtime_seconds,
"best_program_path": str(best_dir / "best_program.py"),
"best_program_info": info
}
except Exception as e:
return {
"status": "failed",
"error": f"Failed to parse best_program_info.json: {str(e)}",
"runtime_seconds": runtime_seconds,
"speedup": None
}
else:
return {
"status": "failed",
"error": "No best_program_info.json found",
"runtime_seconds": runtime_seconds,
"speedup": None
}
def calculate_algotune_score(speedups: List[float]) -> float:
"""
Calculate AlgoTune Score as harmonic mean of speedups.
Args:
speedups: List of speedup values
Returns:
Harmonic mean of speedups
"""
if not speedups:
return 0.0
# Filter out None and invalid values
valid_speedups = [s for s in speedups if s is not None and s > 0]
if not valid_speedups:
return 0.0
return statistics.harmonic_mean(valid_speedups)
def generate_summary(task_results: Dict[str, Dict[str, Any]]) -> Dict[str, Any]:
"""
Generate summary statistics from task results.
Args:
task_results: Dictionary of task results
Returns:
Summary statistics
"""
total_tasks = len(task_results)
successful_tasks = sum(1 for r in task_results.values() if r.get("status") == "success")
failed_tasks = total_tasks - successful_tasks
# Collect speedups for successful tasks
speedups = []
for result in task_results.values():
if result.get("status") == "success" and result.get("speedup") is not None:
speedups.append(result["speedup"])
# Calculate metrics
avg_speedup = statistics.mean(speedups) if speedups else 0.0
algotune_score = calculate_algotune_score(speedups)
total_runtime = sum(r.get("runtime_seconds", 0) for r in task_results.values())
return {
"total_tasks": total_tasks,
"successful_tasks": successful_tasks,
"failed_tasks": failed_tasks,
"average_speedup": round(avg_speedup, 3),
"algotune_score": round(algotune_score, 3),
"total_runtime_seconds": round(total_runtime, 1),
"total_runtime_minutes": round(total_runtime / 60, 1),
"speedups": speedups
}
def main():
"""Main function."""
args = parse_args()
# Setup paths
script_dir = Path(__file__).parent
algotune_dir = script_dir
print("π AlgoTune Benchmark for OpenEvolve")
print(f"π AlgoTune directory: {algotune_dir}")
print(f"π’ Iterations per task: {args.iterations}")
print(f"β±οΈ Timeout per task: {args.timeout}s")
# Discover tasks
all_tasks = discover_algotune_tasks(algotune_dir)
if args.tasks:
requested_tasks = [t.strip() for t in args.tasks.split(',')]
tasks_to_run = [t for t in requested_tasks if t in all_tasks]
missing_tasks = [t for t in requested_tasks if t not in all_tasks]
if missing_tasks:
print(f"β οΈ Warning: Tasks not found: {missing_tasks}")
else:
tasks_to_run = all_tasks
print(f"π Found {len(all_tasks)} total tasks, running {len(tasks_to_run)} tasks")
print(f"π Tasks to run: {', '.join(tasks_to_run)}")
print()
# Run tasks sequentially
task_results = {}
start_time = datetime.now()
for i, task_name in enumerate(tasks_to_run, 1):
print(f"[{i}/{len(tasks_to_run)}] Running {task_name}...")
task_dir = algotune_dir / task_name
result = run_openevolve_task(
task_name=task_name,
task_dir=task_dir,
iterations=args.iterations,
timeout=args.timeout,
custom_config=args.config,
verbose=args.verbose
)
task_results[task_name] = result
# Print speedup if available
if result.get("speedup") is not None:
print(f" Speedup: {result['speedup']:.3f}x")
print()
# Generate output
end_time = datetime.now()
summary = generate_summary(task_results)
output_data = {
"timestamp": start_time.isoformat(),
"duration": str(end_time - start_time),
"config": {
"iterations": args.iterations,
"timeout": args.timeout,
"custom_config": args.config
},
"tasks": task_results,
"summary": summary
}
# Determine output file
if args.output:
output_file = Path(args.output)
else:
timestamp = start_time.strftime("%Y%m%d_%H%M%S")
output_file = Path(f"algotune_benchmark_{timestamp}.json")
# Write results
with open(output_file, 'w') as f:
json.dump(output_data, f, indent=2)
# Print summary
print("=" * 60)
print("π BENCHMARK SUMMARY")
print("=" * 60)
print(f"Total tasks: {summary['total_tasks']}")
print(f"Successful: {summary['successful_tasks']}")
print(f"Failed: {summary['failed_tasks']}")
print(f"Average speedup: {summary['average_speedup']:.3f}x")
print(f"AlgoTune Score: {summary['algotune_score']:.3f}x")
print(f"Total runtime: {summary['total_runtime_minutes']:.1f} minutes")
print()
if summary['speedups']:
print("Individual speedups:")
for task_name, result in task_results.items():
if result.get("speedup") is not None:
print(f" {task_name:25}: {result['speedup']:6.3f}x")
print(f"π Full results saved to: {output_file}")
if __name__ == "__main__":
main() |