Blame for code/train_univfd_multiclass.py

#!/usr/bin/env python3
"""

Training script for UnivFD-style multi-class classification (per-model prediction).



Uses a frozen CLIP ViT-L/14 backbone with a trainable FC layer,

following the UniversalFakeDetect architecture but with multi-class

output for model identification instead of binary real/fake.

"""
import argparse
import json
import os
import random
import sys
from dataclasses import dataclass
from pathlib import Path
from typing import Dict, List, Tuple

import yaml
from PIL import Image

import torch
import torch.nn as nn
import torch.optim as optim
from torch.utils.data import Dataset, DataLoader
import torchvision.transforms as T
from accelerate import Accelerator
from tqdm import tqdm

try:
    import clip
except ImportError:
    raise ImportError("Please install clip: pip install git+https://github.com/openai/CLIP.git")


# CLIP normalization statistics
CLIP_MEAN = [0.48145466, 0.4578275, 0.40821073]
CLIP_STD = [0.26862954, 0.26130258, 0.27577711]

ALLOWED_EXTS = {".jpg", ".jpeg", ".png", ".webp"}


@dataclass
class Sample:
    image_path: Path
    class_index: int


class ImageClsDataset(Dataset):
    """Dataset for loading images organized by model ID folders."""
    REAL_MODEL_ID = 0  # Reserved model ID for real images

    @classmethod
    def from_scratch(cls, ai_root, model_mapping, transform: T.Compose, real_root: str = None):
        samples = dict()

        # Load real images first (model_id = 0)
        if real_root and os.path.isdir(real_root):
            real_files = []
            for file in os.listdir(real_root):
                ext = Path(file).suffix.lower()
                if ext in ALLOWED_EXTS:
                    real_files.append((os.path.join(real_root, file), cls.REAL_MODEL_ID))
            if real_files:
                samples[cls.REAL_MODEL_ID] = real_files
                print(f"Loaded {len(real_files)} real images")

        # Load AI-generated images
        subfolders = os.listdir(ai_root)
        for subfolder in subfolders:
            if '.' in subfolder:
                continue
            try:
                idx = int(subfolder)
            except ValueError:
                continue
            if idx not in model_mapping.values():
                print(subfolder, 'is not listed as a model.')
            else:
                samples[idx] = []
                files = os.listdir(os.path.join(ai_root, subfolder))
                for file in files:
                    ext = Path(file).suffix.lower()
                    if ext in ALLOWED_EXTS:
                        samples[idx].append((os.path.join(ai_root, subfolder, file), idx))
        return cls(samples, model_mapping, transform)

    def __init__(self, samples, model_mapping, transform):
        super().__init__()
        self.samples_dict = samples  # Keep dict for subset operations
        self.model_mapping = model_mapping
        self.transform = transform

        # Create mapping from model_id to class index (0, 1, 2, ...)
        self.model_id_to_class = {mid: i for i, mid in enumerate(sorted(samples.keys()))}
        self.class_to_model_id = {i: mid for mid, i in self.model_id_to_class.items()}

        # Track which class index corresponds to real images (None if no real images)
        self.real_class_idx = self.model_id_to_class.get(self.REAL_MODEL_ID, None)
        self.real_class_indices = [
            class_idx for model_id, class_idx in self.model_id_to_class.items()
            if model_id == self.REAL_MODEL_ID or model_id < 0
        ]

        # Flatten samples into a list for indexing: [(image_path, class_index), ...]
        self.samples = []
        for model_id, items in samples.items():
            class_idx = self.model_id_to_class[model_id]
            for img_path, _ in items:
                self.samples.append((img_path, class_idx))

    @property
    def num_classes(self):
        return len(self.samples_dict)

    def __len__(self) -> int:
        return len(self.samples)

    def __getitem__(self, idx: int):
        img_path, label = self.samples[idx]
        try:
            img = Image.open(img_path).convert("RGB")
        except Exception as exc:
            raise RuntimeError(f"Unreadable benchmark image: {img_path}") from exc
        img = self.transform(img)
        return img, label

    def crop_by_num_samples(self, limitation, min_threshold=None, real_limit=None):
        """Subsample to limitation samples per model."""
        if min_threshold is None:
            min_threshold = limitation
        subset = dict()
        for key, v in self.samples_dict.items():
            if key == self.REAL_MODEL_ID:
                limit = real_limit if real_limit is not None else limitation
                if len(v) >= min_threshold:
                    subset[key] = random.sample(v, k=min(limit, len(v)))
            else:
                if len(v) >= min_threshold:
                    subset[key] = random.sample(v, k=min(limitation, len(v)))
        return ImageClsDataset(subset, self.model_mapping, self.transform)

    def subsample_from(self, n_samples: int, real_ratio: float = None):
        """Create a smaller dataset by taking first n_samples from each AI model."""
        subset = dict()
        real_keys = sorted(k for k in self.samples_dict.keys() if k == self.REAL_MODEL_ID or k < 0)
        num_ai_models = len([k for k in self.samples_dict.keys() if k not in real_keys])
        real_counts = {}
        if real_ratio is not None and real_keys:
            total_real = int(real_ratio * n_samples * num_ai_models)
            base_count = total_real // len(real_keys)
            remainder = total_real % len(real_keys)
            real_counts = {
                key: base_count + (1 if i < remainder else 0)
                for i, key in enumerate(real_keys)
            }

        for key, v in self.samples_dict.items():
            if key in real_counts:
                subset[key] = v[:real_counts[key]]
            else:
                subset[key] = v[:n_samples]
        return ImageClsDataset(subset, self.model_mapping, self.transform)

    def safe_split(self, ratio=0.9, real_test_count: int = None):
        """Split dataset into train and test sets."""
        train, val = dict(), dict()
        for k, v in self.samples_dict.items():
            v_copy = list(v)
            random.shuffle(v_copy)
            if k == self.REAL_MODEL_ID and real_test_count is not None:
                val[k] = v_copy[:real_test_count]
                train[k] = v_copy[real_test_count:]
            else:
                train[k] = v_copy[:int(len(v_copy) * ratio)]
                val[k] = v_copy[int(len(v_copy) * ratio):]
        return ImageClsDataset(train, self.model_mapping, self.transform), ImageClsDataset(val, self.model_mapping, self.transform)

    def get_split_config(self) -> Dict:
        """Export the current dataset configuration for reproducibility."""
        config = {}
        for model_id, items in self.samples_dict.items():
            config[str(model_id)] = [img_path for img_path, _ in items]
        return config

    @classmethod
    def from_split_config(cls, config: Dict, model_mapping, transform: T.Compose):
        """Create dataset from a saved split configuration."""
        samples = {}
        for model_id_str, paths in config.items():
            model_id = int(model_id_str)
            samples[model_id] = [(path, model_id) for path in paths]
        return cls(samples, model_mapping, transform)


class UnivFDMulticlassModel(nn.Module):
    """

    UnivFD-style model for multi-class classification: Frozen CLIP backbone + trainable FC layer.



    Architecture:

        Input -> CLIP ViT-L/14 (frozen) -> FC(768, num_classes) -> Logits

    """

    def __init__(self, num_classes: int, init_gain: float = 0.02):
        super().__init__()

        # Load CLIP model (will be moved to device later by accelerate)
        self.clip_model, _ = clip.load("ViT-L/14", device="cpu")
        self.clip_model.eval()

        # Freeze CLIP backbone
        for param in self.clip_model.parameters():
            param.requires_grad = False

        # Trainable FC layer for multi-class classification
        self.fc = nn.Linear(768, num_classes)
        nn.init.normal_(self.fc.weight, 0.0, init_gain)
        nn.init.zeros_(self.fc.bias)

    def forward(self, x):
        # Extract features with frozen backbone
        with torch.no_grad():
            features = self.clip_model.encode_image(x)
        # Apply trainable FC layer
        return self.fc(features.float())


def save_split_config(train_config: Dict, test_config: Dict, output_path: Path):
    """Save train/test split configuration to JSON file."""
    split_data = {
        "train": train_config,
        "test": test_config,
    }
    with open(output_path, "w", encoding="utf-8") as f:
        json.dump(split_data, f, indent=2)
    print(f"Saved split config to: {output_path}")


def load_split_config(split_path: Path) -> Tuple[Dict, Dict]:
    """Load train/test split configuration from JSON file."""
    with open(split_path, "r", encoding="utf-8") as f:
        split_data = json.load(f)
    return split_data["train"], split_data["test"]


def make_clip_transforms(img_size: int = 224) -> Tuple[T.Compose, T.Compose]:
    """Create CLIP-normalized transforms for training and testing."""
    train_tf = T.Compose([
        T.RandomResizedCrop(img_size, scale=(0.6, 1.0)),
        T.RandomHorizontalFlip(),
        T.ToTensor(),
        T.Normalize(CLIP_MEAN, CLIP_STD),
    ])
    test_tf = T.Compose([
        T.Resize(int(img_size * 1.15)),
        T.CenterCrop(img_size),
        T.ToTensor(),
        T.Normalize(CLIP_MEAN, CLIP_STD),
    ])
    return train_tf, test_tf


def train_one_epoch(

    model: nn.Module,

    loader: DataLoader,

    criterion: nn.Module,

    optimizer: optim.Optimizer,

    accelerator: Accelerator,

    device: torch.device,

    epoch: int = 0,

    total_epochs: int = 0

) -> Tuple[float, float]:
    """Train for one epoch with cross-entropy loss."""
    model.train()
    if hasattr(model, "clip_model"):
        model.clip_model.eval()
    running_loss = 0.0
    correct = 0
    total = 0

    pbar = tqdm(loader, desc=f"Epoch {epoch}/{total_epochs}", disable=not accelerator.is_local_main_process)
    for images, labels in pbar:
        images = images.to(device, non_blocking=True)
        labels = labels.to(device, non_blocking=True)

        optimizer.zero_grad(set_to_none=True)
        outputs = model(images)
        loss = criterion(outputs, labels)
        accelerator.backward(loss)
        optimizer.step()

        running_loss += loss.item() * images.size(0)
        preds = outputs.argmax(dim=1)
        correct += (preds == labels).sum().item()
        total += labels.size(0)

        pbar.set_postfix(loss=f"{loss.item():.4f}", acc=f"{correct/total:.4f}")

    # Reduce metrics across processes
    loss_tensor = torch.tensor([running_loss], device=device)
    correct_tensor = torch.tensor([correct], device=device)
    total_tensor = torch.tensor([total], device=device)
    loss_sum = accelerator.reduce(loss_tensor, reduction="sum").item()
    correct_sum = accelerator.reduce(correct_tensor, reduction="sum").item()
    total_sum = accelerator.reduce(total_tensor, reduction="sum").item()

    avg_loss = loss_sum / max(1, total_sum)
    acc = correct_sum / max(1, total_sum)
    return avg_loss, acc


def evaluate(

    model: nn.Module,

    loader: DataLoader,

    criterion: nn.Module,

    accelerator: Accelerator,

    device: torch.device,

    real_class_indices: List[int] = None,

    num_classes: int = None

) -> Dict:
    """

    Evaluate model with multi-class classification metrics.



    Returns:

        Dict with keys: 'loss', 'per_model_acc', 'top5_acc', 'ai_vs_real_acc',

                       'real_recall', 'confusion_matrix'

    """
    model.eval()
    running_loss = 0.0
    correct_per_model = 0
    correct_top5 = 0
    correct_ai_vs_real = 0
    total = 0

    # Track real class recall
    real_correct = 0
    real_total = 0

    # Track confusion matrix: confusion_matrix[true_label, pred_label]
    if num_classes is not None:
        confusion_matrix = torch.zeros(num_classes, num_classes, device=device, dtype=torch.long)
    else:
        confusion_matrix = None
    real_class_set = set(real_class_indices or [])

    with torch.no_grad():
        for images, labels in loader:
            images = images.to(device, non_blocking=True)
            labels = labels.to(device, non_blocking=True)
            outputs = model(images)
            loss = criterion(outputs, labels)

            running_loss += loss.item() * images.size(0)
            preds = outputs.argmax(dim=1)

            # Per-model accuracy (top-1)
            correct_per_model += (preds == labels).sum().item()

            # Top-5 accuracy
            if num_classes is not None and num_classes >= 5:
                _, top5_preds = outputs.topk(5, dim=1)
                top5_correct = top5_preds.eq(labels.view(-1, 1).expand_as(top5_preds))
                correct_top5 += top5_correct.any(dim=1).sum().item()
            else:
                # If less than 5 classes, top-5 is same as checking if prediction exists
                correct_top5 += (preds == labels).sum().item()

            # Update confusion matrix
            if num_classes is not None:
                for t, p in zip(labels, preds):
                    confusion_matrix[t, p] += 1

            # AI vs Real accuracy (binary classification)
            if real_class_set:
                labels_real_mask = torch.zeros_like(labels, dtype=torch.bool)
                preds_real_mask = torch.zeros_like(preds, dtype=torch.bool)
                for class_idx in real_class_set:
                    labels_real_mask |= labels == class_idx
                    preds_real_mask |= preds == class_idx
                labels_binary = (~labels_real_mask).long()
                preds_binary = (~preds_real_mask).long()
                correct_ai_vs_real += (preds_binary == labels_binary).sum().item()

                # Track real class recall
                real_total += labels_real_mask.sum().item()
                real_correct += ((preds == labels) & labels_real_mask).sum().item()

            total += labels.size(0)

    # Reduce metrics across processes
    loss_tensor = torch.tensor([running_loss], device=device)
    correct_per_model_tensor = torch.tensor([correct_per_model], device=device)
    correct_top5_tensor = torch.tensor([correct_top5], device=device)
    correct_ai_vs_real_tensor = torch.tensor([correct_ai_vs_real], device=device)
    total_tensor = torch.tensor([total], device=device)
    real_correct_tensor = torch.tensor([real_correct], device=device)
    real_total_tensor = torch.tensor([real_total], device=device)

    loss_sum = accelerator.reduce(loss_tensor, reduction="sum").item()
    correct_per_model_sum = accelerator.reduce(correct_per_model_tensor, reduction="sum").item()
    correct_top5_sum = accelerator.reduce(correct_top5_tensor, reduction="sum").item()
    correct_ai_vs_real_sum = accelerator.reduce(correct_ai_vs_real_tensor, reduction="sum").item()
    total_sum = accelerator.reduce(total_tensor, reduction="sum").item()
    real_correct_sum = accelerator.reduce(real_correct_tensor, reduction="sum").item()
    real_total_sum = accelerator.reduce(real_total_tensor, reduction="sum").item()

    # Reduce confusion matrix across processes
    if num_classes is not None:
        confusion_matrix = accelerator.reduce(confusion_matrix, reduction="sum")
        confusion_matrix_list = confusion_matrix.cpu().tolist()
    else:
        confusion_matrix_list = None

    avg_loss = loss_sum / max(1, total_sum)
    per_model_acc = correct_per_model_sum / max(1, total_sum)
    top5_acc = correct_top5_sum / max(1, total_sum)
    ai_vs_real_acc = correct_ai_vs_real_sum / max(1, total_sum) if real_class_set else None
    real_recall = real_correct_sum / max(1, real_total_sum) if real_class_set and real_total_sum > 0 else None

    return {
        'loss': avg_loss,
        'per_model_acc': per_model_acc,
        'top5_acc': top5_acc,
        'ai_vs_real_acc': ai_vs_real_acc,
        'real_recall': real_recall,
        'confusion_matrix': confusion_matrix_list,
    }


def save_outputs(output_dir: Path, model: nn.Module, is_best: bool = False, is_final: bool = False):
    """Save model checkpoints."""
    output_dir.mkdir(parents=True, exist_ok=True)

    if is_best:
        # Save only the FC layer weights (like UnivFD)
        fc_path = output_dir / "fc_weights_best.pth"
        torch.save(model.fc.state_dict(), fc_path)
        print(f"Saved best FC weights to: {fc_path}")

    if is_final:
        fc_path = output_dir / "fc_weights_final.pth"
        torch.save(model.fc.state_dict(), fc_path)
        print(f"Saved final FC weights to: {fc_path}")


def validation_epoch_interval(num_epochs: int, num_validations: int | None) -> int:
    """Return the epoch interval that keeps validation passes bounded."""
    if num_validations is None or num_validations <= 0:
        return 1
    return max(1, num_epochs // num_validations)


def should_validate_epoch(

    epoch: int,

    num_epochs: int,

    num_validations: int | None,

    no_val_between_epochs: bool,

) -> bool:
    if no_val_between_epochs:
        return epoch == num_epochs
    interval = validation_epoch_interval(num_epochs, num_validations)
    return epoch == num_epochs or epoch % interval == 0


def train_with_sample_size(

    train_ds: ImageClsDataset,

    test_ds: ImageClsDataset,

    sample_size: int,

    num_epochs: int,

    args,

    accelerator: Accelerator,

    output_dir: Path,

):
    """Train a model with a specific sample size setting."""
    accelerator.print(f"\n{'='*60}")
    accelerator.print(f"Training with sample_size={sample_size}, epochs={num_epochs}")
    accelerator.print(f"{'='*60}")

    # Create subsampled training set
    train_tf, test_tf = make_clip_transforms(args.img_size)
    n_samples_per_ai = int(sample_size * args.train_ratio)
    sub_train_ds = train_ds.subsample_from(n_samples_per_ai, real_ratio=args.real_ratio)

    sub_train_ds.transform = train_tf
    test_ds.transform = test_tf

    # Count samples
    real_model_ids = {k for k in sub_train_ds.samples_dict if k == ImageClsDataset.REAL_MODEL_ID or k < 0}
    real_train_count = sum(len(v) for k, v in sub_train_ds.samples_dict.items() if k in real_model_ids)
    ai_train_count = sum(len(v) for k, v in sub_train_ds.samples_dict.items() if k not in real_model_ids)
    accelerator.print(f"Train samples: {len(sub_train_ds)} (Real: {real_train_count}, AI: {ai_train_count})")
    accelerator.print(f"Test samples: {len(test_ds)}")
    accelerator.print(f"Number of classes: {sub_train_ds.num_classes}")

    pin_memory = torch.cuda.is_available()
    train_loader = DataLoader(
        sub_train_ds, batch_size=args.batch_size, shuffle=True, num_workers=args.num_workers, pin_memory=pin_memory
    )
    test_loader = DataLoader(
        test_ds, batch_size=args.batch_size, shuffle=False, num_workers=args.num_workers, pin_memory=pin_memory
    )

    device = accelerator.device
    model = UnivFDMulticlassModel(num_classes=sub_train_ds.num_classes, init_gain=args.init_gain)
    criterion = nn.CrossEntropyLoss()

    # Only optimize FC layer parameters
    optimizer = optim.AdamW(model.fc.parameters(), lr=args.lr, weight_decay=args.weight_decay)

    model, optimizer, train_loader, test_loader = accelerator.prepare(model, optimizer, train_loader, test_loader)

    best_acc = -1.0
    results = []

    # Get class to model_id mapping for per-model analysis
    class_to_model_id = sub_train_ds.class_to_model_id
    num_classes = sub_train_ds.num_classes
    real_class_indices = sub_train_ds.real_class_indices

    best_confusion_matrix = None
    best_top5_acc = -1.0
    best_real_recall = -1.0

    for epoch in range(1, num_epochs + 1):
        train_loss, train_acc = train_one_epoch(
            model, train_loader, criterion, optimizer, accelerator, device, epoch, num_epochs
        )

        validate_now = should_validate_epoch(
            epoch, num_epochs, args.num_validations, args.no_val_between_epochs
        )
        if not validate_now:
            accelerator.print(f"Epoch {epoch:02d}/{num_epochs} | Train loss {train_loss:.4f} acc {train_acc:.4f}")
            results.append({
                'epoch': epoch,
                'validated': False,
                'train_loss': train_loss,
                'train_acc': train_acc,
                'test_loss': None,
                'per_model_acc': None,
                'top5_acc': None,
                'ai_vs_real_acc': None,
                'real_recall': None,
                'confusion_matrix': None,
            })
            continue

        metrics = evaluate(
            model, test_loader, criterion, accelerator, device,
            real_class_indices=real_class_indices, num_classes=num_classes
        )

        test_loss = metrics['loss']
        per_model_acc = metrics['per_model_acc']
        top5_acc = metrics['top5_acc']
        ai_vs_real_acc = metrics['ai_vs_real_acc']
        real_recall = metrics['real_recall']
        confusion_matrix = metrics['confusion_matrix']

        # Build output string
        output_str = (
            f"Epoch {epoch:02d}/{num_epochs} | Train loss {train_loss:.4f} acc {train_acc:.4f} | "
            f"Test loss {test_loss:.4f} top1 {per_model_acc:.4f} top5 {top5_acc:.4f}"
        )
        if real_recall is not None:
            output_str += f" real_recall {real_recall:.4f}"
        if ai_vs_real_acc is not None:
            output_str += f" ai-vs-real {ai_vs_real_acc:.4f}"
        accelerator.print(output_str)

        is_best = per_model_acc > best_acc
        if is_best:
            best_acc = per_model_acc
            best_top5_acc = top5_acc
            best_real_recall = real_recall if real_recall is not None else -1.0
            best_confusion_matrix = confusion_matrix

        # Save best model (only when validation is performed)

        results.append({
            'epoch': epoch,
            'validated': True,
            'train_loss': train_loss,
            'train_acc': train_acc,
            'test_loss': test_loss,
            'per_model_acc': per_model_acc,
            'top5_acc': top5_acc,
            'ai_vs_real_acc': ai_vs_real_acc,
            'real_recall': real_recall,
            'confusion_matrix': confusion_matrix,
        })

    # Save final model
    accelerator.wait_for_everyone()
    if accelerator.is_main_process:
        unwrapped = accelerator.unwrap_model(model)
        save_outputs(output_dir, unwrapped, is_final=True)

    # Save results summary
    # Get final epoch metrics for summary
    final_metrics = results[-1] if results else {}
    if accelerator.is_main_process:
        results_path = output_dir / "results.json"
        with open(results_path, "w") as f:
            json.dump({
                'sample_size': sample_size,
                'num_epochs': num_epochs,
                'num_validations': args.num_validations,
                'validation_interval': validation_epoch_interval(num_epochs, args.num_validations),
                'num_classes': num_classes,
                'best_per_model_acc': best_acc,
                'best_top5_acc': best_top5_acc,
                'best_real_recall': best_real_recall if best_real_recall >= 0 else None,
                'final_per_model_acc': final_metrics.get('per_model_acc'),
                'final_top5_acc': final_metrics.get('top5_acc'),
                'final_real_recall': final_metrics.get('real_recall'),
                'best_confusion_matrix': best_confusion_matrix,
                'class_to_model_id': {str(k): v for k, v in class_to_model_id.items()},
                'epochs': results,
            }, f, indent=2)
        accelerator.print(f"Saved results to: {results_path}")

    return final_metrics['per_model_acc'], final_metrics['top5_acc'], final_metrics['real_recall']


def parse_args():
    parser = argparse.ArgumentParser(description="Train UnivFD-style multi-class classifier (per-model prediction).")
    parser.add_argument("--mapping", type=str, default="metadata/model_mapping.yaml",
                       help="Path to model_mapping.yaml")
    parser.add_argument("--images-root", type=str, default="data_local/images/ai",
                       help="Root directory containing {id} subfolders with images")
    parser.add_argument("--real-root", type=str, default="data_local/images/real",
                       help="Root directory containing real images")
    parser.add_argument("--output-dir", type=str, default="./outputs_univfd_multiclass",
                       help="Directory to save model and results")
    parser.add_argument("--batch-size", type=int, default=64)
    parser.add_argument("--lr", type=float, default=1e-4, help="Learning rate (UnivFD default: 1e-4)")
    parser.add_argument("--weight-decay", type=float, default=0.0, help="Weight decay for AdamW")
    parser.add_argument("--init-gain", type=float, default=0.02, help="FC layer weight initialization std")
    parser.add_argument("--num-workers", type=int, default=4)
    parser.add_argument("--img-size", type=int, default=224)
    parser.add_argument("--train-ratio", type=float, default=1.0, help="Train split ratio per class")
    parser.add_argument("--sample-sizes", type=str, default="900,90,45,18",
                       help="Comma-separated sample sizes to train on (descending order)")
    parser.add_argument("--total-samples-per-model", type=int, default=9000,
                       help="Total samples per model across all epochs")
    parser.add_argument("--real-ratio", type=float, default=0.1,
                       help="Ratio of real images to total AI images in training")
    parser.add_argument("--real-test-count", type=int, default=100,
                       help="Fixed number of real images in test set")
    parser.add_argument("--split-file", type=str, required=True,
                       help="Path to split config JSON. If provided, reuse this split")
    parser.add_argument("--no-val-between-epochs", action="store_true",
                       help="Skip validation between epochs, only evaluate at the end")
    parser.add_argument("--num-validations", type=int, default=None,
                       help="Target validation passes per sample size. Default validates every epoch.")
    parser.add_argument("--seed", type=int, default=42)
    parser.add_argument("--mixed-precision", type=str, default="no", choices=["no", "fp16", "bf16"],
                       help="Mixed precision mode for accelerate")
    return parser.parse_args()


def main():
    args = parse_args()
    accelerator = Accelerator(mixed_precision=args.mixed_precision)

    random.seed(args.seed)
    torch.manual_seed(args.seed)

    images_root = Path(args.images_root)
    output_dir = Path(args.output_dir)

    # Parse sample sizes
    sample_sizes = [int(s.strip()) for s in args.sample_sizes.split(",")]
    sample_sizes = sorted(sample_sizes, reverse=True)
    max_sample_size = sample_sizes[0]

    accelerator.print("=" * 60)
    accelerator.print("UnivFD-style Multi-class Classification Training")
    accelerator.print("=" * 60)
    accelerator.print(f"Model: CLIP ViT-L/14 (frozen) + FC layer (trainable)")
    accelerator.print(f"Task: Multi-class classification (per-model prediction)")
    accelerator.print(f"Sample sizes to train: {sample_sizes}")
    accelerator.print(f"Total samples per model target: {args.total_samples_per_model}")
    accelerator.print(f"Real image ratio: {args.real_ratio}")
    accelerator.print(f"Real test count: {args.real_test_count}")

    accelerator.print("\nLoading mapping...")
    with open(args.mapping, "r", encoding="utf-8") as f:
        model_mapping = yaml.safe_load(f)

    accelerator.print("Preparing base dataset...")
    train_tf, test_tf = make_clip_transforms(args.img_size)

    if not Path(args.split_file).is_file():
        raise FileNotFoundError(args.split_file)
    with open(args.split_file, encoding="utf-8") as split_handle:
        release_split = json.load(split_handle)
    if release_split.get("protocol", {}).get("seed") != args.seed:
        raise ValueError("Build a release split for this seed before training")
    if args.split_file:
        accelerator.print(f"Loading split config from: {args.split_file}")
        train_config, test_config = load_split_config(Path(args.split_file))
        base_train_ds = ImageClsDataset.from_split_config(train_config, model_mapping, train_tf)
        base_test_ds = ImageClsDataset.from_split_config(test_config, model_mapping, test_tf)
        accelerator.print(f"Loaded {len(base_train_ds)} train samples, {len(base_test_ds)} test samples")
    else:
        dataset = ImageClsDataset.from_scratch(images_root, model_mapping, train_tf, real_root=args.real_root)

        num_ai_models = len([k for k in dataset.samples_dict.keys() if k != ImageClsDataset.REAL_MODEL_ID])
        max_real_train = int(args.real_ratio * (max_sample_size * args.train_ratio) * num_ai_models)
        real_limit = max_real_train + args.real_test_count
        accelerator.print(f"Real images needed: {real_limit} (train: {max_real_train}, test: {args.real_test_count})")

        subset = dataset.crop_by_num_samples(max_sample_size, min_threshold=1000, real_limit=real_limit)
        base_train_ds, base_test_ds = subset.safe_split(args.train_ratio, real_test_count=args.real_test_count)
        base_test_ds.transform = test_tf

        output_dir.mkdir(parents=True, exist_ok=True)
        split_path = output_dir / "split_config.json"
        if accelerator.is_main_process:
            save_split_config(base_train_ds.get_split_config(), base_test_ds.get_split_config(), split_path)

    accelerator.print(f"Base dataset: {base_train_ds.num_classes} classes, {len(base_train_ds)} train, {len(base_test_ds)} test")

    real_class_indices = base_train_ds.real_class_indices
    if real_class_indices:
        accelerator.print(f"Real images included as classes {real_class_indices}")
    else:
        accelerator.print("No real images - AI-vs-real accuracy will not be computed")

    # Train for each sample size
    all_results = {}
    for sample_size in sample_sizes:
        num_epochs = args.total_samples_per_model // sample_size
        accelerator.print(f"\nSample size {sample_size}: {num_epochs} epochs")

        exp_output_dir = output_dir / f"exp_{sample_size}"

        final_acc, final_top5, final_recall = train_with_sample_size(
            base_train_ds,
            base_test_ds,
            sample_size,
            num_epochs,
            args,
            accelerator,
            exp_output_dir,
        )

        all_results[sample_size] = {
            'num_epochs': num_epochs,
            'final_per_model_acc': final_acc,
            'final_top5_acc': final_top5,
            'final_real_recall': final_recall,
        }

    # Save summary
    if accelerator.is_main_process:
        summary_path = output_dir / "summary.json"
        with open(summary_path, "w") as f:
            json.dump(all_results, f, indent=2)
        accelerator.print(f"\nSaved summary to: {summary_path}")

    accelerator.print("\nAll training completed!")
    accelerator.print("\nResults summary:")
    for sample_size, result in all_results.items():
        output_str = f"  Sample size {sample_size}: {result['num_epochs']} epochs, top1 {result['final_per_model_acc']:.4f}, top5 {result['final_top5_acc']:.4f}"
        if result['final_real_recall'] is not None:
            output_str += f", real_recall {result['final_real_recall']:.4f}"
        accelerator.print(output_str)


if __name__ == "__main__":
    try:
        main()
    except Exception as e:
        print(f"Error: {e}", file=sys.stderr)
        sys.exit(1)