Training configuration

Training configuration lives under the Trainer root object.

Trainer:
  Model:
    classpath: segmentation.UNet.UNet
    UNet:
      ...
  Dataset:
    ...
  train_name: SEG_BASELINE
  epochs: 100

Top-level fields

Field	Type	Default in code	Required	Effect
Model	mapping	ModelLoader()	Yes	Selects and configures the model graph.
Dataset	mapping	DataTrain()	Yes	Defines training data loading, transforms, augmentation, and patching.
train_name	string	TRAIN_01	No	Names the run and its output folders.
manual_seed	int or null	None	No	Sets the random seed when provided.
epochs	int	100	No	Number of training epochs.
it_validation	int or null	None	No	Validation and checkpoint interval in iterations.
autocast	bool	false	No	Enables AMP during training.
gradient_checkpoints	list or null	None	No	Activates gradient checkpointing on selected modules.
gpu_checkpoints	list or null	None	No	Pins selected modules to dedicated GPUs.
ema_decay	float	0	No	Enables exponential moving average tracking when greater than zero.
data_log	list or null	None	No	TensorBoard logging directives for dataset groups or model outputs.
EarlyStopping	mapping or null	None	No	Configures early stopping.
save_checkpoint_mode	string	BEST	No	BEST keeps the best checkpoint, ALL keeps every saved checkpoint.

Trainer.Model

Trainer.Model always starts with a classpath, then a section named after the selected class.

Model:
  classpath: segmentation.UNet.UNet
  UNet:
    optimizer:
      name: AdamW
      lr: 0.001

Common nested fields used by built-in and local models:

Field	Type	Required	Effect
classpath	string	Yes	Selects the model class to import.
<SelectedClass>	mapping	Yes	Constructor arguments for the chosen class.
optimizer	mapping	Usually	Optimizer configuration passed through OptimizerLoader.
schedulers	mapping	Optional	Learning-rate schedulers keyed by classpath.
outputs_criterions	mapping	Usually	Declares losses and metrics attached to specific model outputs.
Patch	mapping	Optional	Enables model-level patching via ModelPatch.
dim	int	Model-dependent	Declares whether the network operates in 2D or 3D.

outputs_criterions

This is the most important training structure after the dataset definition.

outputs_criterions:
  UNetBlock_0:Head:Conv:
    targets_criterions:
      SEG:
        criterions_loader:
          torch:nn:CrossEntropyLoss:
            is_loss: true
            schedulers:
              Constant:
                nb_step: 0
                value: 1

Structure:

• output key → model output or module path

• targets_criterions → one or more target groups

• criterions_loader → one or more criteria for that target

• each criterion can define is_loss, group, start, stop, accumulation, and scheduler weights

Trainer.Dataset

Training datasets are instantiated through DataTrain.

Common fields:

Field	Type	Default in code	Effect
dataset_filenames	list[str]	["./Dataset:mha"]	Dataset sources and selection mode.
groups_src	mapping	required in practice	Maps on-disk groups to loaded tensors.
augmentations	mapping or null	one default augmentation list	Data augmentations sampled during training.
inline_augmentations	bool	false	Keeps base samples cached and generates augmentation tensors only when an augmented sample is requested; augmentation states are re-sampled on each epoch.
Patch	mapping or null	DatasetPatch()	Dataset-level patch extraction.
use_cache	bool	true	Cache transformed data in memory.
subset	object	TrainSubset()	Restricts which cases are used.
batch_size	int	1	Batch size.
num_workers	int or null	None	Number of DataLoader workers. When use_cache: false, None auto-enables worker prefetching.
pin_memory	bool	false	Enables pinned host memory for DataLoader batches.
prefetch_factor	int or null	None	Number of prefetched batches per worker when workers are enabled.
persistent_workers	bool or null	None	Keep workers alive across epochs when workers are enabled.
validation	float / string / list / null	0.2	Validation split or explicit validation set.
shuffle	bool	true through subset	Shuffles the training sampler.

When use_cache: false, KonfAI now tries to stream patches directly from disk instead of materializing full volumes in RAM. This path is used automatically when the configured preprocessing chain is compatible with patch-wise loading; otherwise KonfAI falls back to the existing full-volume loading path.

groups_src

Each source group contains one or more destination groups:

groups_src:
  CT:
    groups_dest:
      CT:
        transforms:
          Standardize:
            lazy: false
            mean: None
            std: None
            mask: None
            inverse: false
        patch_transforms: None
        is_input: true

Use this section to define:

• what exists on disk

• preprocessing transforms

• patch-specific transforms

• whether the tensor is a model input

Examples

The most practical examples in the repository are:

• examples/Segmentation/Config.yml

• examples/Synthesis/Config.yml

• examples/Synthesis/Config_GAN.yml

See also

• Common configuration patterns

• Prediction configuration

• Datasets and groups

• Training workflows