| import os | |
| import json | |
| import time | |
| from datetime import datetime | |
| import joblib | |
| import psutil | |
| import platform | |
| try: | |
| import cpuinfo | |
| CPU_NAME = cpuinfo.get_cpu_info().get('brand_raw', platform.processor()) | |
| except Exception: | |
| CPU_NAME = platform.processor() | |
| def log_experiment_results(logs, log_dir="experiments/logs/", log_file="experiment_log.jsonl"): | |
| os.makedirs(log_dir, exist_ok=True) | |
| log_path = os.path.join(log_dir, log_file) | |
| with open(log_path, "a") as f: | |
| for entry in logs: | |
| f.write(json.dumps(entry) + "\n") | |
| def create_log_entry(experiment_title, model_name, hyperparams, dataset_name, preprocessing, metrics, train_time, model_object): | |
| timestamp = datetime.now().isoformat() | |
| model_size = get_model_size(model_object) | |
| cpu_util = psutil.cpu_percent(interval=0.1) | |
| return { | |
| "experiment_title": experiment_title, | |
| "timestamp": timestamp, | |
| "model": model_name, | |
| "hyperparameters": hyperparams, | |
| "dataset": dataset_name, | |
| "preprocessing": preprocessing, | |
| "metrics": metrics, | |
| "training_time_sec": train_time, | |
| "model_size_bytes": model_size, | |
| "system_info": { | |
| "cpu": CPU_NAME, | |
| "cpu_utilization": cpu_util, | |
| "memory_used_mb": psutil.Process().memory_info().rss // 1024 ** 2 | |
| } | |
| } | |
| def get_model_size(model): | |
| temp_path = "experiments/logs/_temp_model.joblib" | |
| joblib.dump(model, temp_path) | |
| size = os.path.getsize(temp_path) | |
| os.remove(temp_path) | |
| return size | |