WZMIAOMIAO/deep-learning-for-image-processing · error · ValueError

dataset have {} classes, but input {}

Error message

dataset have {} classes, but input {}

What it means

The single-GPU training script compares args.num_classes (used to build the model's classifier) with the number of distinct labels in the training dataset; if they differ it raises ValueError showing both numbers. This guard prevents training a model whose output dimensionality doesn't match the label space.

Source

Thrown at pytorch_classification/mini_imagenet/train_single_gpu.py:45

                                     transforms.RandomHorizontalFlip(),
                                     transforms.ToTensor(),
                                     transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])]),
        "val": transforms.Compose([transforms.Resize(256),
                                   transforms.CenterCrop(224),
                                   transforms.ToTensor(),
                                   transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])])}

    data_root = args.data_path
    json_path = "./classes_name.json"
    # 实例化训练数据集
    train_dataset = MyDataSet(root_dir=data_root,
                              csv_name="new_train.csv",
                              json_path=json_path,
                              transform=data_transform["train"])

    # check num_classes
    if args.num_classes != len(train_dataset.labels):
        raise ValueError("dataset have {} classes, but input {}".format(len(train_dataset.labels),
                                                                        args.num_classes))

    # 实例化验证数据集
    val_dataset = MyDataSet(root_dir=data_root,
                            csv_name="new_val.csv",
                            json_path=json_path,
                            transform=data_transform["val"])

    batch_size = args.batch_size
    nw = min([os.cpu_count(), batch_size if batch_size > 1 else 0, 8])  # number of workers
    print('Using {} dataloader workers every process'.format(nw))
    train_loader = torch.utils.data.DataLoader(train_dataset,
                                               batch_size=batch_size,
                                               shuffle=True,
                                               pin_memory=True,
                                               num_workers=nw,
                                               collate_fn=train_dataset.collate_fn)

View on GitHub (pinned to 1ec3fe6f37)

Solutions

  1. Pass the correct --num_classes matching the dataset (e.g. 64 for Mini-ImageNet train split).
  2. Regenerate class_indices.json and the CSVs after any dataset change.
  3. Print len(train_dataset.labels) before launching to confirm the expected count.

Example fix

// before
python train_single_gpu.py --num_classes=5
// after
python train_single_gpu.py --num_classes=64
Defensive patterns

Strategy: validation

Validate before calling

import json
with open(json_path) as f:
    class_indices = json.load(f)
assert args.num_classes == len(class_indices), \
    f"--num_classes={args.num_classes} but dataset has {len(class_indices)} classes"

Type guard

def num_classes_matches(args, dataset_labels) -> bool:
    return getattr(args, "num_classes", None) == len(dataset_labels)

Try / catch

try:
    train_loop(args)
except ValueError as e:
    if "classes" in str(e):
        print(f"Fix: pass --num_classes={len(train_dataset.labels)}")
    raise

Prevention

When it happens

Trigger: Running train_single_gpu.py with --num_classes N where N != number of classes in class_indices.json built from new_train.csv.

Common situations: Switching datasets without updating the CLI argument, stale class_indices.json from an older dataset, forgetting that Mini-ImageNet has 64 train classes instead of 1000.

Related errors

AI-assisted analysis of WZMIAOMIAO/deep-learning-for-image-processing@1ec3fe6f37 (2026-08-30). Data as JSON: /api/errors/613209bb358947b1. Report an issue: GitHub.