Docs /Templates/ sinapsis-anomalib

GlassTrain

sinapsis-anomalibAnomalibanomalibanomaly_detectiondynamicmodels

Python template · sinapsis_anomalib.templates.anomalib_train

Overview

Training implementation for Anomalib models.

Glass — inherited documentation

PyTorch Lightning Implementation of the GLASS Model.

The model uses a pre-trained feature extractor to extract features and a feature adaptor to mitigate latent domain bias. Global anomaly features are synthesized from adapted normal features using gradient ascent. Local anomaly images are synthesized using texture overlay datasets like dtd which are then processed by feature extractor and feature adaptor. All three different features are passed to the discriminator trained using loss functions.

Args: input_shape (tuple[int, int]): Input image dimensions as a tuple of (height, width). Required for shaping the input pipeline. Defaults to `(288, 288)`. anomaly_source_path (str): Path to the dataset or source directory containing normal images and anomaly textures backbone (str, optional): Name of the CNN backbone used for feature extraction. Defaults to `"wide_resnet50_2"`. pretrain_embed_dim (int, optional): Dimensionality of features extracted by the pre-trained backbone before adaptation. Defaults to `1536`. target_embed_dim (int, optional): Dimensionality of the target adapted features after projection. Defaults to `1536`. patchsize (int, optional): Size of the local patch used in feature aggregation (e.g., for neighborhood pooling). Defaults to `3`. patchstride (int, optional): Stride used when extracting patches for local feature aggregation. Defaults to `1`. pre_trained (bool, optional): Whether to use ImageNet pre-trained weights for the backbone network. Defaults to `True`. layers (list[str], optional): List of backbone layers to extract features from. Defaults to `["layer2", "layer3"]`. pre_projection (int, optional): Number of projection layers used in the feature adaptor (e.g., MLP before discriminator). Defaults to `1`. discriminator_layers (int, optional): Number of layers in the discriminator network. Defaults to `2`. discriminator_hidden (int, optional): Number of hidden units in each discriminator layer. Defaults to `1024`. learning_rate (float, optional): Learning rate for training the feature adaptor and discriminator networks. Defaults to `0.0001`. step (int, optional): Number of gradient ascent steps for anomaly synthesis. Defaults to `20`. svd (int, optional): Flag to enable SVD-based feature projection. Defaults to `0`. gaussian_noise_std (float, optional): Standard deviation of Gaussian noise added to features for global anomaly synthesis. Defaults to ``0.015``. radius_quantile (float, optional): Quantile used to compute the truncated projection radius during gradient ascent. Defaults to ``0.75``. focal_loss_quantile_threshold (float, optional): Quantile threshold for hard example mining in focal loss computation. When ``0``, all samples are used. Defaults to ``0.5``. mining (bool): Whether to perform gradient ascent or skip it. Defaults to ``True``. pre_processor (PreProcessor | bool, optional): reprocessing module or flag to enable default preprocessing. Set to `True` to apply default normalization and resizing. Defaults to `True`. post_processor (PostProcessor | bool, optional): Postprocessing module or flag to enable default output smoothing or thresholding. Defaults to `True`. evaluator (Evaluator | bool, optional): Evaluation module for calculating metrics such as AUROC and PRO. Defaults to `True`. visualizer (Visualizer | bool, optional): Visualization module to generate heatmaps, segmentation overlays, and anomaly scores. Defaults to `True`.

Capabilities

Capabilities: not set

Installation

Install the published package using pip or uv.

Terminalbash
pip install sinapsis-anomalib==1.0.0
uv pip install sinapsis-anomalib==1.0.0
uv add sinapsis-anomalib==1.0.0

Configuration

Effective fields from the inspector schema, including nested Pydantic model fields.

logger
lightning.pytorch.loggers.logger.Logger | collections.abc.Iterable[lightning.pytorch.loggers.logger.Logger] | bool | None
Optional
Default: null
metadata
sinapsis_core.template_base.base_models.TemplateMetadata | None
Optional
Default: null
callbacks
list[lightning.pytorch.callbacks.callback.Callback] | None
Optional
Default: null
ckpt_path

Path to checkpoint for resuming training. If None, starts training from scratch.

str | pathlib._local.Path | None
Optional
Default: null
threshold
anomalib.metrics.threshold.base.Threshold | tuple[anomalib.metrics.threshold.base.Threshold, anomalib.metrics.threshold.base.Threshold] | omegaconf.dictconfig.DictConfig | omegaconf.listconfig.ListConfig | list[dict[str, str | float]] | str | None
Optional
Default: null
glass_initRequired
__init__Attributes
Required
Default: null
svd
int
Optional
Default: 0
step
int
Optional
Default: 20
layers
NoneType
Optional
Default: null
mining
bool
Optional
Default: true
backbone
str
Optional
Default: "wide_resnet50_2"
evaluator
typing.Union[anomalib.metrics.evaluator.Evaluator, bool]
Optional
Default: true
patchsize
int
Optional
Default: 3
visualizer
typing.Union[anomalib.visualization.base.Visualizer, bool]
Optional
Default: true
input_shape
tuple[int, int]
Optional
Default: [288,288]
patchstride
int
Optional
Default: 1
pre_trained
bool
Optional
Default: true
learning_rate
float
Optional
Default: 0.0001
pre_processor
typing.Union[anomalib.pre_processing.pre_processor.PreProcessor, bool]
Optional
Default: true
post_processor
typing.Union[anomalib.post_processing.post_processor.PostProcessor, bool]
Optional
Default: true
pre_projection
int
Optional
Default: 1
radius_quantile
float
Optional
Default: 0.75
target_embed_dim
int
Optional
Default: 1536
gaussian_noise_std
float
Optional
Default: 0.015
pretrain_embed_dim
int
Optional
Default: 1536
anomaly_source_path
typing.Optional[str]
Optional
Default: null
discriminator_hidden
int
Optional
Default: 1024
discriminator_layers
int
Optional
Default: 2
focal_loss_quantile_threshold
float
Optional
Default: 0.5
train_root
str | pathlib._local.Path | None
Optional
Default: null
trainer_args
(TrainerConfig): General trainer arguments. For more details see: https://lightning.ai/docs/pytorch/stable/common/trainer.html#trainer-flags glass_init input_shape tuple[int, int]: default: (288, 288)… Read full field description

(TrainerConfig): General trainer arguments. For more details see: https://lightning.ai/docs/pytorch/stable/common/trainer.html#trainer-flags glass_init input_shape tuple[int, int]: default: (288, 288) anomaly_source_path typing.Optional[str]: default: None backbone <class 'str'>: default: wide_resnet50_2 pretrain_embed_dim <class 'int'>: default: 1536 target_embed_dim <class 'int'>: default: 1536 patchsize <class 'int'>: default: 3 patchstride <class 'int'>: default: 1 pre_trained <class 'bool'>: default: True layers <class 'NoneType'>: default: None pre_projection <class 'int'>: default: 1 discriminator_layers <class 'int'>: default: 2 discriminator_hidden <class 'int'>: default: 1024 learning_rate <class 'float'>: default: 0.0001 step <class 'int'>: default: 20 svd <class 'int'>: default: 0 gaussian_noise_std <class 'float'>: default: 0.015 radius_quantile <class 'float'>: default: 0.75 focal_loss_quantile_threshold <class 'float'>: default: 0.5 mining <class 'bool'>: default: True pre_processor typing.Union[anomalib.pre_processing.pre_processor.PreProcessor, bool]: default: True post_processor typing.Union[anomalib.post_processing.post_processor.PostProcessor, bool]: default: True evaluator typing.Union[anomalib.metrics.evaluator.Evaluator, bool]: default: True visualizer typing.Union[anomalib.visualization.base.Visualizer, bool]: default: True

TrainerConfig
Optional
Default: null
devices

The devices to use for training (e.g., 1, [0, 1], "auto"). Defaults to "auto".

int | list[int] | str | None
Optional
Default: "auto"
max_epochs

The maximum number of epochs to train for. Defaults to 5.

int | None
Optional
Default: 5
min_epochs

The minimum number of epochs to train for. Defaults to 1.

int
Optional
Default: 1
accelerator

The hardware accelerator to use. Defaults to "auto".

typing.Literal['cpu', 'gpu', 'tpu', 'hpu', 'auto']
Optional
Default: "cpu"
image_metrics
list[str] | str | dict[str, dict[str, typing.Any]] | None
Optional
Default: null
normalization
anomalib.utils.normalization.NormalizationMethod | omegaconf.dictconfig.DictConfig | lightning.pytorch.callbacks.callback.Callback | str | None
Optional
Default: null
pixel_metrics
list[str] | str | dict[str, dict[str, typing.Any]] | None
Optional
Default: null
logger_configs
dict[str, dict[str, typing.Any]] | None
Optional
Default: null
callback_configs
dict[str, dict[str, typing.Any]] | None
Optional
Default: null
folder_attributes
sinapsis_anomalib.helpers.configs.FolderConfig | None
Optional
Default: null
nameRequired

A descriptive name for the dataset.

str
Required
Default: null
root

The root directory where the dataset is located.

str | pathlib._local.Path | None
Optional
Default: null
seed

A random seed for reproducibility in data splitting.

int | None
Optional
Default: null
mask_dir
str | pathlib._local.Path | collections.abc.Sequence[str | pathlib._local.Path] | None
Optional
Default: null
extensions
list[str] | None
Optional
Default: null
normal_dirRequired

Path to the directory containing normal images.

str | pathlib._local.Path | collections.abc.Sequence[str | pathlib._local.Path]
Required
Default: null
num_workers

The number of subprocesses to use for data loading. Defaults to 8.

int
Optional
Default: 8
abnormal_dir

Path to the directory for abnormal images.

str | pathlib._local.Path | collections.abc.Sequence[str | pathlib._local.Path] | None
Optional
Default: null
val_split_mode
anomalib.data.utils.split.ValSplitMode | str
Optional
Default: "from_test"
eval_batch_size

The number of samples per batch for evaluation. Defaults to 32.

int
Optional
Default: 32
normal_test_dir
str | pathlib._local.Path | collections.abc.Sequence[str | pathlib._local.Path] | None
Optional
Default: null
test_split_mode
anomalib.data.utils.split.TestSplitMode | str
Optional
Default: "from_dir"
val_split_ratio
float
Optional
Default: 0.5
test_split_ratio
float
Optional
Default: 0.2
train_batch_size

The number of samples per batch for training. Defaults to 32.

int
Optional
Default: 32
normal_split_ratio
float
Optional
Default: 0.2

Configuration example

Generated from documented defaults. Replace any placeholders and complete required nested settings before running.

yaml
templates:
  - template_name: "GlassTrain"
    class_name: "GlassTrain"
    attributes:
      logger: null
      callbacks: null
      ckpt_path: null
      threshold: null
      glass_init:
        svd: 0
        step: 20
        layers: null
        mining: true
        backbone: "wide_resnet50_2"
        evaluator: true
        patchsize: 3
        visualizer: true
        input_shape: [288, 288]
        patchstride: 1
        pre_trained: true
        learning_rate: 0.0001
        pre_processor: true
        post_processor: true
        pre_projection: 1
        radius_quantile: 0.75
        target_embed_dim: 1536
        gaussian_noise_std: 0.015
        pretrain_embed_dim: 1536
        anomaly_source_path: null
        discriminator_hidden: 1024
        discriminator_layers: 2
        focal_loss_quantile_threshold: 0.5
      train_root: null
      trainer_args:
        devices: "auto"
        max_epochs: 5
        min_epochs: 1
        accelerator: "cpu"
      image_metrics: null
      normalization: null
      pixel_metrics: null
      logger_configs: null
      callback_configs: null
      folder_attributes:
        name: "<required>"
        root: null
        seed: null
        mask_dir: null
        extensions: null
        normal_dir: "<required>"
        num_workers: 8
        abnormal_dir: null
        val_split_mode: "from_test"
        eval_batch_size: 32
        normal_test_dir: null
        test_split_mode: "from_dir"
        val_split_ratio: 0.5
        test_split_ratio: 0.2
        train_batch_size: 32
        normal_split_ratio: 0.2
View original docstring example

The source example references CfaTrain, not GlassTrain.

yaml
agent:
    name: my_test_agent
    templates:
    - template_name: InputTemplate
    class_name: InputTemplate
    attributes: {}
    - template_name: CfaTrain
    class_name: CfaTrain
    template_input: InputTemplate
    attributes:
        folder_attributes_config_path: 'path/to/config.yaml'
        generic_key: 'my_generic_key'
        callbacks: null
        normalization: null
        threshold: null
        image_metrics: null
        pixel_metrics: null
        logger: null
        callback_configs: null
        logger_configs: null
        ckpt_path: null
        trainer_args:
            devices: "0"
            accelerator: gpu
            max_epochs: null
        cfa_init:
            backbone: wide_resnet50_2
            gamma_c: 1
            gamma_d: 1
            num_nearest_neighbors: 3
            num_hard_negative_features: 3
            radius: 1.0e-05
For a full list of options use the sinapsis cli: sinapsis info --all-template-names.
If you want to see all available models, please visit:
https://anomalib.readthedocs.io/en/v1.2.0/markdown/guides/reference/models/image/index.html.

Inheritance

Immediate parent: sinapsis_anomalib.templates.anomalib_train.AnomalibTrain

Environment variables

Environment definitions belong to their canonical Python modules.

ANOMALIB_ROOT_DIR · defined elsewhereSINAPSIS_BUILD_DOCS · defined elsewhere