recycling-project-solutions/train.py

#!/usr/bin/env python3
"""
YOLOv8 Training Script for Memory Module Detection
This script trains a YOLOv8 nano model to detect memory modules in motherboard images.
"""

import os
import sys
from pathlib import Path
import yaml
from ultralytics import YOLO
import torch

def check_dataset_structure():
    """Verify that the dataset structure is correct."""
    required_paths = [
        'training/train/images',
        'training/train/labels', 
        'training/val/images',
        'training/val/labels',
        'dataset.yaml'
    ]
    
    for path in required_paths:
        if not os.path.exists(path):
            raise FileNotFoundError(f"Required path not found: {path}")
    
    # Check if we have images and labels
    train_images = len([f for f in os.listdir('training/train/images') if f.endswith('.png')])
    train_labels = len([f for f in os.listdir('training/train/labels') if f.endswith('.txt')])
    val_images = len([f for f in os.listdir('training/val/images') if f.endswith('.png')])
    val_labels = len([f for f in os.listdir('training/val/labels') if f.endswith('.txt')])
    
    print(f"Dataset structure verified:")
    print(f"  Training: {train_images} images, {train_labels} labels")
    print(f"  Validation: {val_images} images, {val_labels} labels")
    
    return True

def train_model(epochs=100, imgsz=640, batch_size=16, device='auto'):
    """
    Train YOLOv8 nano model on memory module dataset.
    
    Args:
        epochs (int): Number of training epochs
        imgsz (int): Image size for training
        batch_size (int): Batch size for training
        device (str): Device to use ('auto', 'cpu', 'cuda', or specific GPU id)
    """
    
    # Check dataset structure
    check_dataset_structure()
    
    # Initialize YOLOv8 nano model
    print("Initializing YOLOv8 nano model...")
    model = YOLO('yolov8n.pt')  # Load pretrained YOLOv8 nano model
    
    # Check available device
    if device == 'auto':
        device = 'cuda' if torch.cuda.is_available() else 'cpu'
    
    print(f"Using device: {device}")
    print(f"CUDA available: {torch.cuda.is_available()}")
    if torch.cuda.is_available():
        print(f"CUDA device: {torch.cuda.get_device_name()}")
    
    # Training configuration
    train_config = {
        'data': 'dataset.yaml',
        'epochs': epochs,
        'imgsz': imgsz,
        'batch': batch_size,
        'device': device,
        'project': 'runs/detect',
        'name': 'memory_module_detection',
        'save': True,
        'save_period': 10,  # Save checkpoint every 10 epochs
        'cache': False,  # Don't cache images (saves RAM)
        'workers': 4,
        'patience': 50,  # Early stopping patience
        'optimizer': 'AdamW',
        'lr0': 0.01,  # Initial learning rate
        'lrf': 0.01,  # Final learning rate factor
        'momentum': 0.937,
        'weight_decay': 0.0005,
        'warmup_epochs': 3,
        'warmup_momentum': 0.8,
        'warmup_bias_lr': 0.1,
        'box': 7.5,  # Box loss gain
        'cls': 0.5,  # Class loss gain
        'dfl': 1.5,  # DFL loss gain
        'pose': 12.0,  # Pose loss gain
        'kobj': 1.0,  # Keypoint obj loss gain
        'label_smoothing': 0.0,
        'nbs': 64,  # Nominal batch size
        'hsv_h': 0.015,  # Image HSV-Hue augmentation
        'hsv_s': 0.7,  # Image HSV-Saturation augmentation
        'hsv_v': 0.4,  # Image HSV-Value augmentation
        'degrees': 0.0,  # Image rotation (+/- deg)
        'translate': 0.1,  # Image translation (+/- fraction)
        'scale': 0.5,  # Image scale (+/- gain)
        'shear': 0.0,  # Image shear (+/- deg)
        'perspective': 0.0,  # Image perspective (+/- fraction)
        'flipud': 0.0,  # Image flip up-down (probability)
        'fliplr': 0.5,  # Image flip left-right (probability)
        'mosaic': 1.0,  # Image mosaic (probability)
        'mixup': 0.0,  # Image mixup (probability)
        'copy_paste': 0.0,  # Segment copy-paste (probability)
    }
    
    print("Starting training...")
    print(f"Configuration: {train_config}")
    
    # Train the model
    results = model.train(**train_config)
    
    # Print training results
    print("\nTraining completed!")
    print(f"Best model saved at: runs/detect/memory_module_detection/weights/best.pt")
    print(f"Last model saved at: runs/detect/memory_module_detection/weights/last.pt")
    
    return results

def validate_model(model_path='runs/detect/memory_module_detection/weights/best.pt'):
    """Validate the trained model."""
    if not os.path.exists(model_path):
        print(f"Model not found at {model_path}")
        return None
    
    print(f"Validating model: {model_path}")
    model = YOLO(model_path)
    
    # Run validation
    results = model.val(data='dataset.yaml')
    
    print("Validation completed!")
    return results

if __name__ == "__main__":
    import argparse
    
    parser = argparse.ArgumentParser(description='Train YOLOv8 for memory module detection')
    parser.add_argument('--epochs', type=int, default=100, help='Number of training epochs')
    parser.add_argument('--imgsz', type=int, default=640, help='Image size for training')
    parser.add_argument('--batch', type=int, default=16, help='Batch size')
    parser.add_argument('--device', type=str, default='auto', help='Device to use (auto, cpu, cuda)')
    parser.add_argument('--validate', action='store_true', help='Only run validation')
    parser.add_argument('--model', type=str, default='runs/detect/memory_module_detection/weights/best.pt', 
                       help='Model path for validation')
    
    args = parser.parse_args()
    
    if args.validate:
        validate_model(args.model)
    else:
        train_model(epochs=args.epochs, imgsz=args.imgsz, batch_size=args.batch, device=args.device)
        # Also run validation after training
        validate_model()