v0_5

Used by:

• API Reference  model
  •  ModelDescr
  •  ModelDescr_v0_5

Classes:

Name	Description
ArchitectureFromFileDescr
ArchitectureFromLibraryDescr
Author
AxisBase
AxisId
BadgeDescr	A custom badge
BatchAxis
BiasRisksLimitations	Known biases, risks, technical limitations, and recommendations for model use.
BinarizeAlongAxisKwargs	key word arguments for BinarizeDescr
BinarizeDescr	Binarize the tensor with a fixed threshold.
BinarizeKwargs	key word arguments for BinarizeDescr
BioimageioConfig
CallableFromDepencency
CellposeFlowDynamicsDescr	Cellpose flow dynamics postprocessing as described in:
CellposeFlowDynamicsKwargs	key word arguments for CellposeFlowDynamicsDescr
ChannelAxis
CiteEntry	A citation that should be referenced in work using this resource.
ClipDescr	Set tensor values below min to min and above max to max.
ClipKwargs	key word arguments for ClipDescr
Config
ConstantPadding
ContinuousReproducibilityTolerance
CustomProcessingDescr	Custom (post)processing op — source file shipped inline with the model.
DataDependentSize
DatasetDescr	A bioimage.io dataset resource description file (dataset RDF) describes a dataset relevant to bioimage
DatasetId
Datetime	Timestamp in ISO 8601 format
DeprecatedLicenseId
DiscreteReproducibilityTolerance
Doi	A digital object identifier, see https://www.doi.org/
EdgePadding
EnsureDtypeDescr	Cast the tensor data type to EnsureDtypeKwargs.dtype (if not matching).
EnsureDtypeKwargs	key word arguments for EnsureDtypeDescr
EnvironmentalImpact	Environmental considerations for model training and deployment.
Evaluation
FileDescr	A file description
FixedZeroMeanUnitVarianceAlongAxisKwargs	key word arguments for FixedZeroMeanUnitVarianceDescr
FixedZeroMeanUnitVarianceDescr	Subtract a given mean and divide by the standard deviation.
FixedZeroMeanUnitVarianceKwargs	key word arguments for FixedZeroMeanUnitVarianceDescr
HttpUrl	A URL with the HTTP or HTTPS scheme.
Identifier
IndexAxisBase
IndexInputAxis
IndexOutputAxis
InputTensorDescr
IntervalOrRatioDataDescr
KerasHdf5WeightsDescr
KerasV3WeightsDescr
LicenseId
LinkedDataset	Reference to a bioimage.io dataset.
LinkedModel	Reference to a bioimage.io model.
LinkedResource	Reference to a bioimage.io resource
Maintainer
ModelDescr	Specification of the fields used in a bioimage.io-compliant RDF to describe AI models with pretrained weights.
ModelId
NominalOrOrdinalDataDescr
OnnxWeightsDescr
OrcidId	An ORCID identifier, see https://orcid.org/
OutputTensorDescr
ParameterizedSize	Describes a range of valid tensor axis sizes as size = min + n*step.
PytorchStateDictWeightsDescr
ReflectPadding
RelativeFilePath	A path relative to the rdf.yaml file (also if the RDF source is a URL).
ReproducibilityTolerance
ResourceId
RunMode
ScaleLinearAlongAxisKwargs	Key word arguments for ScaleLinearDescr
ScaleLinearDescr	Fixed linear scaling.
ScaleLinearKwargs	Key word arguments for ScaleLinearDescr
ScaleMeanVarianceDescr	Scale a tensor's data distribution to match another tensor's mean/std.
ScaleMeanVarianceKwargs	key word arguments for ScaleMeanVarianceKwargs
ScaleRangeDescr	Scale with percentiles.
ScaleRangeKwargs	key word arguments for ScaleRangeDescr
Sha256	A SHA-256 hash value
SiUnit	An SI unit
SigmoidDescr	The logistic sigmoid function, a.k.a. expit function.
SizeReference	A tensor axis size (extent in pixels/frames) defined in relation to a reference axis.
SoftmaxDescr	The softmax function.
SoftmaxKwargs	key word arguments for SoftmaxDescr
SpaceAxisBase
SpaceInputAxis
SpaceOutputAxis
SpaceOutputAxisWithHalo
StardistPostprocessingDescr	Stardist postprocessing including non-maximum suppression and converting polygon representations to instance labels
StardistPostprocessingKwargs2D
StardistPostprocessingKwargs3D
SymmetricPadding
TensorDescrBase
TensorId
TensorflowJsWeightsDescr
TensorflowSavedModelBundleWeightsDescr
TimeAxisBase
TimeInputAxis
TimeOutputAxis
TimeOutputAxisWithHalo
TorchscriptWeightsDescr
TrainingDetails
Uploader
Version	wraps a packaging.version.Version instance for validation in pydantic models
WeightsDescr
WeightsEntryDescrBase
WithHalo
ZeroMeanUnitVarianceDescr	Subtract mean and divide by variance.
ZeroMeanUnitVarianceKwargs	key word arguments for ZeroMeanUnitVarianceDescr

Functions:

Name	Description
convert_axes
generate_covers
get_halos	Get all input and output halos from tensor descriptions.
validate_tensors	Validate all inputs (and optionally output tensors) against their tensor descriptions.

Attributes:

Name	Type	Description
ANY_AXIS_TYPES		intended for isinstance comparisons in py<3.10
AbsoluteTolerance
AnyAxis
AxisType
BATCH_AXIS_ID
BioimageioYamlContent
CHANNEL_AXIS_ID
DEFAULT_INDEX_AXIS_ID
DEFAULT_SPACE_AXIS_ID
DEFAULT_TIME_AXIS_ID
FileDescr_dependencies
FileDescr_external_data
INPUT_AXIS_TYPES		intended for isinstance comparisons in py<3.10
IO_AxisT
InputAxis
IntervalOrRatioDType
KnownRunMode
MismatchedElementsPerMillion
NominalOrOrdinalDType
NonBatchAxisId
NotEmpty
OUTPUT_AXIS_TYPES		intended for isinstance comparisons in py<3.10
OutputAxis
Padding
ParameterizedSize_N	TypeAlias	Annotates an integer to calculate a concrete axis size from a ParameterizedSize.
PostprocessingDescr
PostprocessingId
PreprocessingDescr
PreprocessingId
RelativeTolerance
SAME_AS_TYPE
SpaceUnit		Space unit compatible to the OME-Zarr axes specification 0.5
SpecificWeightsDescr
TVs
TensorDataDescr
TensorDescr
TimeUnit		Time unit compatible to the OME-Zarr axes specification 0.5
VALID_COVER_IMAGE_EXTENSIONS
WeightsFormat

ANY_AXIS_TYPES module-attribute

ANY_AXIS_TYPES = INPUT_AXIS_TYPES + OUTPUT_AXIS_TYPES

intended for isinstance comparisons in py<3.10

AbsoluteTolerance module-attribute

AbsoluteTolerance = Annotated[
    float, annotated_types.Interval(ge=0)
]

Subclassed by:

•  v0_5
  •  ContinuousReproducibilityTolerance
  •  ReproducibilityTolerance

AnyAxis module-attribute

AnyAxis = Union[InputAxis, OutputAxis]

AxisType module-attribute

AxisType = Literal[
    "batch", "channel", "index", "time", "space"
]

BATCH_AXIS_ID module-attribute

BATCH_AXIS_ID = AxisId('batch')

BioimageioYamlContent module-attribute

BioimageioYamlContent = Dict[str, YamlValue]

Returned by:

• API Reference  spec
  •  dump_description
  •  get_package_content
  •  get_resource_package_content

Used by:

• API Reference
  •  common  BioimageioYamlSource
  •  spec
    •  save_bioimageio_yaml_only
    •  update_format
    •  update_hashes
    •  upload
    •  validate_format

CHANNEL_AXIS_ID module-attribute

CHANNEL_AXIS_ID = AxisId('channel')

DEFAULT_INDEX_AXIS_ID module-attribute

DEFAULT_INDEX_AXIS_ID = AxisId('index')

DEFAULT_SPACE_AXIS_ID module-attribute

DEFAULT_SPACE_AXIS_ID = AxisId('x')

DEFAULT_TIME_AXIS_ID module-attribute

DEFAULT_TIME_AXIS_ID = AxisId('time')

FileDescr_dependencies module-attribute

FileDescr_dependencies = Annotated[
    FileDescr_package,
    WithSuffix((".yaml", ".yml"), case_sensitive=True),
    Field(examples=[{"source": "environment.yaml"}]),
]

FileDescr_external_data module-attribute

FileDescr_external_data = Annotated[
    FileDescr_package,
    WithSuffix(".data", case_sensitive=True),
    Field(examples=[{"source": "weights.onnx.data"}]),
]

INPUT_AXIS_TYPES module-attribute

INPUT_AXIS_TYPES = (
    BatchAxis,
    ChannelAxis,
    IndexInputAxis,
    TimeInputAxis,
    SpaceInputAxis,
)

intended for isinstance comparisons in py<3.10

Used by:

•  v0_5  ANY_AXIS_TYPES

IO_AxisT module-attribute

IO_AxisT = TypeVar('IO_AxisT', InputAxis, OutputAxis)

Subclassed by:

•  v0_5  TensorDescrBase

InputAxis module-attribute

InputAxis = Annotated[
    _InputAxisUnion, Discriminator("type")
]

Subclassed by:

•  v0_5  InputTensorDescr

Used by:

•  v0_5
  •  AnyAxis
  •  IO_AxisT

IntervalOrRatioDType module-attribute

IntervalOrRatioDType = Literal[
    "float32",
    "float64",
    "uint8",
    "int8",
    "uint16",
    "int16",
    "uint32",
    "int32",
    "uint64",
    "int64",
]

KnownRunMode module-attribute

KnownRunMode = Literal['deepimagej']

MismatchedElementsPerMillion module-attribute

MismatchedElementsPerMillion = Annotated[
    int, annotated_types.Interval(ge=0, le=30000)
]

NominalOrOrdinalDType module-attribute

NominalOrOrdinalDType = Literal[
    "float32",
    "float64",
    "uint8",
    "int8",
    "uint16",
    "int16",
    "uint32",
    "int32",
    "uint64",
    "int64",
    "bool",
]

NonBatchAxisId module-attribute

NonBatchAxisId = Annotated[AxisId, Predicate(_is_not_batch)]

NotEmpty module-attribute

NotEmpty = Annotated[S, annotated_types.MinLen(1)]

Used by:

•  v0_5  TVs

OUTPUT_AXIS_TYPES module-attribute

OUTPUT_AXIS_TYPES = (
    BatchAxis,
    ChannelAxis,
    IndexOutputAxis,
    TimeOutputAxis,
    TimeOutputAxisWithHalo,
    SpaceOutputAxis,
    SpaceOutputAxisWithHalo,
)

intended for isinstance comparisons in py<3.10

Used by:

•  v0_5  ANY_AXIS_TYPES

OutputAxis module-attribute

OutputAxis = Annotated[
    _OutputAxisUnion, Discriminator("type")
]

Subclassed by:

•  v0_5  OutputTensorDescr

Used by:

•  v0_5
  •  AnyAxis
  •  IO_AxisT

Padding module-attribute

Padding = Union[
    ConstantPadding,
    EdgePadding,
    ReflectPadding,
    SymmetricPadding,
]

ParameterizedSize_N module-attribute

ParameterizedSize_N: TypeAlias = int

Annotates an integer to calculate a concrete axis size from a ParameterizedSize.

PostprocessingDescr module-attribute

PostprocessingDescr = Annotated[
    Union[
        BinarizeDescr,
        CellposeFlowDynamicsDescr,
        ClipDescr,
        CustomProcessingDescr,
        EnsureDtypeDescr,
        FixedZeroMeanUnitVarianceDescr,
        ScaleLinearDescr,
        ScaleMeanVarianceDescr,
        ScaleRangeDescr,
        SigmoidDescr,
        SoftmaxDescr,
        StardistPostprocessingDescr,
        ZeroMeanUnitVarianceDescr,
    ],
    Discriminator("id"),
]

PostprocessingId module-attribute

PostprocessingId = Literal[
    "binarize",
    "clip",
    "custom",
    "ensure_dtype",
    "fixed_zero_mean_unit_variance",
    "scale_linear",
    "scale_mean_variance",
    "scale_range",
    "sigmoid",
    "softmax",
    "zero_mean_unit_variance",
]

PreprocessingDescr module-attribute

PreprocessingDescr = Annotated[
    Union[
        BinarizeDescr,
        ClipDescr,
        EnsureDtypeDescr,
        FixedZeroMeanUnitVarianceDescr,
        ScaleLinearDescr,
        ScaleRangeDescr,
        SigmoidDescr,
        SoftmaxDescr,
        ZeroMeanUnitVarianceDescr,
    ],
    Discriminator("id"),
]

PreprocessingId module-attribute

PreprocessingId = Literal[
    "binarize",
    "clip",
    "ensure_dtype",
    "fixed_zero_mean_unit_variance",
    "scale_linear",
    "scale_range",
    "sigmoid",
    "softmax",
]

RelativeTolerance module-attribute

RelativeTolerance = Annotated[
    float, annotated_types.Interval(ge=0, le=0.01)
]

Subclassed by:

•  v0_5
  •  ContinuousReproducibilityTolerance
  •  ReproducibilityTolerance

SAME_AS_TYPE module-attribute

SAME_AS_TYPE = '<same as type>'

SpaceUnit module-attribute

SpaceUnit = Literal[
    "attometer",
    "angstrom",
    "centimeter",
    "decimeter",
    "exameter",
    "femtometer",
    "foot",
    "gigameter",
    "hectometer",
    "inch",
    "kilometer",
    "megameter",
    "meter",
    "micrometer",
    "mile",
    "millimeter",
    "nanometer",
    "parsec",
    "petameter",
    "picometer",
    "terameter",
    "yard",
    "yoctometer",
    "yottameter",
    "zeptometer",
    "zettameter",
]

Space unit compatible to the OME-Zarr axes specification 0.5

SpecificWeightsDescr module-attribute

SpecificWeightsDescr = Union[
    KerasHdf5WeightsDescr,
    KerasV3WeightsDescr,
    OnnxWeightsDescr,
    PytorchStateDictWeightsDescr,
    TensorflowJsWeightsDescr,
    TensorflowSavedModelBundleWeightsDescr,
    TorchscriptWeightsDescr,
]

TVs module-attribute

TVs = Union[
    NotEmpty[List[int]],
    NotEmpty[List[float]],
    NotEmpty[List[bool]],
    NotEmpty[List[str]],
]

TensorDataDescr module-attribute

TensorDataDescr = Union[
    NominalOrOrdinalDataDescr, IntervalOrRatioDataDescr
]

TensorDescr module-attribute

TensorDescr = Union[InputTensorDescr, OutputTensorDescr]

Used by:

•  v0_5
  •  get_halos
  •  validate_tensors

TimeUnit module-attribute

TimeUnit = Literal[
    "attosecond",
    "centisecond",
    "day",
    "decisecond",
    "exasecond",
    "femtosecond",
    "gigasecond",
    "hectosecond",
    "hour",
    "kilosecond",
    "megasecond",
    "microsecond",
    "millisecond",
    "minute",
    "nanosecond",
    "petasecond",
    "picosecond",
    "second",
    "terasecond",
    "yoctosecond",
    "yottasecond",
    "zeptosecond",
    "zettasecond",
]

Time unit compatible to the OME-Zarr axes specification 0.5

VALID_COVER_IMAGE_EXTENSIONS module-attribute

VALID_COVER_IMAGE_EXTENSIONS = (
    ".gif",
    ".jpeg",
    ".jpg",
    ".png",
    ".svg",
)

WeightsFormat module-attribute

WeightsFormat = Literal[
    "keras_hdf5",
    "keras_v3",
    "onnx",
    "pytorch_state_dict",
    "tensorflow_js",
    "tensorflow_saved_model_bundle",
    "torchscript",
]

ArchitectureFromFileDescr pydantic-model

Bases: _ArchitectureCallableDescr, FileDescr

Show JSON schema:

{
  "$defs": {
    "RelativeFilePath": {
      "description": "A path relative to the `rdf.yaml` file (also if the RDF source is a URL).",
      "format": "path",
      "title": "RelativeFilePath",
      "type": "string"
    },
    "YamlValue": {
      "anyOf": [
        {
          "type": "boolean"
        },
        {
          "format": "date",
          "type": "string"
        },
        {
          "format": "date-time",
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        },
        {
          "type": "string"
        },
        {
          "items": {
            "$ref": "#/$defs/YamlValue"
          },
          "type": "array"
        },
        {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "type": "object"
        },
        {
          "type": "null"
        }
      ]
    }
  },
  "additionalProperties": false,
  "properties": {
    "source": {
      "anyOf": [
        {
          "description": "A URL with the HTTP or HTTPS scheme.",
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "title": "HttpUrl",
          "type": "string"
        },
        {
          "$ref": "#/$defs/RelativeFilePath"
        },
        {
          "format": "file-path",
          "title": "FilePath",
          "type": "string"
        }
      ],
      "description": "Architecture source file",
      "title": "FileSource"
    },
    "sha256": {
      "anyOf": [
        {
          "description": "A SHA-256 hash value",
          "maxLength": 64,
          "minLength": 64,
          "title": "Sha256",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "SHA256 hash value of the **source** file.",
      "title": "Sha256"
    },
    "callable": {
      "description": "Identifier of the callable that returns a torch.nn.Module instance.",
      "examples": [
        "MyNetworkClass",
        "get_my_model"
      ],
      "minLength": 1,
      "title": "Identifier",
      "type": "string"
    },
    "kwargs": {
      "additionalProperties": {
        "$ref": "#/$defs/YamlValue"
      },
      "description": "key word arguments for the `callable`",
      "title": "Kwargs",
      "type": "object"
    }
  },
  "required": [
    "source",
    "callable"
  ],
  "title": "model.v0_5.ArchitectureFromFileDescr",
  "type": "object"
}

Fields:

• sha256 (Sha256 | None)
• callable (Annotated[Identifier, Field(examples=['MyNetworkClass', 'get_my_model'])])
• kwargs (dict[str, YamlValue])
• source (Annotated[FileSource, AfterValidator(wo_special_file_name)])

callable pydantic-field

callable: Annotated[
    Identifier,
    Field(examples=["MyNetworkClass", "get_my_model"]),
]

Identifier of the callable that returns a torch.nn.Module instance.

kwargs pydantic-field

kwargs: dict[str, YamlValue]

key word arguments for the callable

sha256 pydantic-field

sha256: Sha256 | None = None

SHA256 hash value of the source file.

source pydantic-field

source: Annotated[
    FileSource, AfterValidator(wo_special_file_name)
]

Architecture source file

suffix property

suffix: str

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

download

download(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

alias for .get_reader

Source code in src/bioimageio/spec/_internal/io.py

def download(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """alias for `.get_reader`"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

get_reader

get_reader(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

open the file source (download if needed)

Source code in src/bioimageio/spec/_internal/io.py

def get_reader(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """open the file source (download if needed)"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

validate_sha256

validate_sha256(force_recompute: bool = False) -> None

validate the sha256 hash value of the source file

Source code in src/bioimageio/spec/_internal/io.py

def validate_sha256(self, force_recompute: bool = False) -> None:
    """validate the sha256 hash value of the **source** file"""
    context = get_validation_context()
    src_str = str(self.source)
    if force_recompute:
        actual_sha = None
    else:
        actual_sha = context.known_files.get(src_str)

    if actual_sha is None:
        if context.perform_io_checks or force_recompute:
            reader = get_reader(self.source, sha256=self.sha256)
            if force_recompute:
                actual_sha = get_sha256(reader)
            else:
                actual_sha = reader.sha256

            context.known_files[src_str] = actual_sha
        elif context.known_files and src_str not in context.known_files:
            # perform_io_checks is False, but known files were given,
            # so we expect all file references to be in there
            raise ValueError(f"File {src_str} not found in `known_files`.")

    if actual_sha is None or self.sha256 == actual_sha:
        return
    elif self.sha256 is None or context.update_hashes:
        self.sha256 = actual_sha
    elif self.sha256 != actual_sha:
        raise ValueError(
            f"Sha256 mismatch for {self.source}. Expected {self.sha256}, got "
            + f"{actual_sha}. Update expected `sha256` or point to the matching "
            + "file."
        )

ArchitectureFromLibraryDescr pydantic-model

Bases: _ArchitectureCallableDescr

Show JSON schema:

{
  "$defs": {
    "YamlValue": {
      "anyOf": [
        {
          "type": "boolean"
        },
        {
          "format": "date",
          "type": "string"
        },
        {
          "format": "date-time",
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        },
        {
          "type": "string"
        },
        {
          "items": {
            "$ref": "#/$defs/YamlValue"
          },
          "type": "array"
        },
        {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "type": "object"
        },
        {
          "type": "null"
        }
      ]
    }
  },
  "additionalProperties": false,
  "properties": {
    "callable": {
      "description": "Identifier of the callable that returns a torch.nn.Module instance.",
      "examples": [
        "MyNetworkClass",
        "get_my_model"
      ],
      "minLength": 1,
      "title": "Identifier",
      "type": "string"
    },
    "kwargs": {
      "additionalProperties": {
        "$ref": "#/$defs/YamlValue"
      },
      "description": "key word arguments for the `callable`",
      "title": "Kwargs",
      "type": "object"
    },
    "import_from": {
      "description": "Where to import the callable from, i.e. `from <import_from> import <callable>`",
      "title": "Import From",
      "type": "string"
    }
  },
  "required": [
    "callable",
    "import_from"
  ],
  "title": "model.v0_5.ArchitectureFromLibraryDescr",
  "type": "object"
}

Fields:

• callable (Annotated[Identifier, Field(examples=['MyNetworkClass', 'get_my_model'])])
• kwargs (dict[str, YamlValue])
• import_from (str)

callable pydantic-field

callable: Annotated[
    Identifier,
    Field(examples=["MyNetworkClass", "get_my_model"]),
]

Identifier of the callable that returns a torch.nn.Module instance.

import_from pydantic-field

import_from: str

Where to import the callable from, i.e. from <import_from> import <callable>

kwargs pydantic-field

kwargs: dict[str, YamlValue]

key word arguments for the callable

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

Author pydantic-model

Bases: _Author_v0_2

Show JSON schema:

{
  "additionalProperties": false,
  "properties": {
    "affiliation": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Affiliation",
      "title": "Affiliation"
    },
    "email": {
      "anyOf": [
        {
          "format": "email",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Email",
      "title": "Email"
    },
    "orcid": {
      "anyOf": [
        {
          "description": "An ORCID identifier, see https://orcid.org/",
          "title": "OrcidId",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "An [ORCID iD](https://support.orcid.org/hc/en-us/sections/360001495313-What-is-ORCID\n) in hyphenated groups of 4 digits, (and [valid](\nhttps://support.orcid.org/hc/en-us/articles/360006897674-Structure-of-the-ORCID-Identifier\n) as per ISO 7064 11,2.)",
      "examples": [
        "0000-0001-2345-6789"
      ],
      "title": "Orcid"
    },
    "name": {
      "title": "Name",
      "type": "string"
    },
    "github_user": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Github User"
    }
  },
  "required": [
    "name"
  ],
  "title": "generic.v0_3.Author",
  "type": "object"
}

Fields:

• affiliation (str | None)
• email (EmailStr | None)
• orcid (Annotated[OrcidId | None, Field(examples=['0000-0001-2345-6789'])])
• name (Annotated[str, Predicate(_has_no_slash)])
• github_user (str | None)

Validators:

• _validate_github_user → github_user

affiliation pydantic-field

affiliation: str | None = None

Affiliation

email pydantic-field

email: EmailStr | None = None

Email

github_user pydantic-field

github_user: str | None = None

name pydantic-field

name: Annotated[str, Predicate(_has_no_slash)]

orcid pydantic-field

orcid: Annotated[
    OrcidId | None, Field(examples=["0000-0001-2345-6789"])
] = None

An ORCID iD in hyphenated groups of 4 digits, (and valid as per ISO 7064 11,2.)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

AxisBase pydantic-model

Bases: NodeWithExplicitlySetFields

Show JSON schema:

{
  "additionalProperties": false,
  "properties": {
    "id": {
      "description": "An axis id unique across all axes of one tensor.",
      "maxLength": 16,
      "minLength": 1,
      "title": "AxisId",
      "type": "string"
    },
    "description": {
      "default": "",
      "description": "A short description of this axis beyond its type and id.",
      "maxLength": 128,
      "title": "Description",
      "type": "string"
    }
  },
  "required": [
    "id"
  ],
  "title": "model.v0_5.AxisBase",
  "type": "object"
}

Fields:

• id (AxisId)
• description (Annotated[str, MaxLen(128)])

description pydantic-field

description: Annotated[str, MaxLen(128)] = ''

A short description of this axis beyond its type and id.

id pydantic-field

id: AxisId

An axis id unique across all axes of one tensor.

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

AxisId

Bases: LowerCaseIdentifier

              flowchart TD
              bioimageio.spec.model.v0_5.AxisId[AxisId]
              bioimageio.spec._internal.types.LowerCaseIdentifier[LowerCaseIdentifier]
              bioimageio.spec._internal.validated_string.ValidatedString[ValidatedString]

                              bioimageio.spec._internal.types.LowerCaseIdentifier --> bioimageio.spec.model.v0_5.AxisId
                                bioimageio.spec._internal.validated_string.ValidatedString --> bioimageio.spec._internal.types.LowerCaseIdentifier
                



              click bioimageio.spec.model.v0_5.AxisId href "" "bioimageio.spec.model.v0_5.AxisId"
              click bioimageio.spec._internal.types.LowerCaseIdentifier href "" "bioimageio.spec._internal.types.LowerCaseIdentifier"
              click bioimageio.spec._internal.validated_string.ValidatedString href "" "bioimageio.spec._internal.validated_string.ValidatedString"
            

Returned by:

•  v0_5  get_halos

Used by:

•  v0_5
  •  BATCH_AXIS_ID
  •  CHANNEL_AXIS_ID
  •  DEFAULT_INDEX_AXIS_ID
  •  DEFAULT_SPACE_AXIS_ID
  •  DEFAULT_TIME_AXIS_ID
  •  NonBatchAxisId

Methods:

Name	Description
__get_pydantic_core_schema__
__get_pydantic_json_schema__
__new__

Attributes:

Name	Type	Description
root_model	type[RootModel[Any]]	the pydantic root model to validate the string

root_model class-attribute

root_model: type[RootModel[Any]] = RootModel[
    Annotated[
        NotEmpty[str],
        AfterValidator(_normalize_axis_id),
        MaxLen(16),
        AfterValidator(validate_identifier),
        AfterValidator(validate_is_not_keyword),
    ]
]

the pydantic root model to validate the string

__get_pydantic_core_schema__ classmethod

__get_pydantic_core_schema__(
    source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_core_schema__(
    cls, source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema:
    return no_info_after_validator_function(cls, handler(str))

__get_pydantic_json_schema__ classmethod

__get_pydantic_json_schema__(
    core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_json_schema__(
    cls, core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue:
    json_schema = cls.root_model.model_json_schema(mode=handler.mode)
    json_schema["title"] = cls.__name__.strip("_")
    if cls.__doc__:
        json_schema["description"] = cls.__doc__

    return json_schema

__new__

__new__(object: object)

Source code in src/bioimageio/spec/_internal/validated_string.py

def __new__(cls, object: object):
    _validated = cls.root_model.model_validate(object).root
    self = super().__new__(cls, _validated)
    self._validated = _validated
    return self._after_validator()

BadgeDescr pydantic-model

Bases: Node

A custom badge

Show JSON schema:

{
  "$defs": {
    "RelativeFilePath": {
      "description": "A path relative to the `rdf.yaml` file (also if the RDF source is a URL).",
      "format": "path",
      "title": "RelativeFilePath",
      "type": "string"
    }
  },
  "additionalProperties": false,
  "description": "A custom badge",
  "properties": {
    "label": {
      "description": "badge label to display on hover",
      "examples": [
        "Open in Colab"
      ],
      "title": "Label",
      "type": "string"
    },
    "icon": {
      "anyOf": [
        {
          "format": "file-path",
          "title": "FilePath",
          "type": "string"
        },
        {
          "$ref": "#/$defs/RelativeFilePath"
        },
        {
          "description": "A URL with the HTTP or HTTPS scheme.",
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "title": "HttpUrl",
          "type": "string"
        },
        {
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "badge icon (included in bioimage.io package if not a URL)",
      "examples": [
        "https://colab.research.google.com/assets/colab-badge.svg"
      ],
      "title": "Icon"
    },
    "url": {
      "description": "target URL",
      "examples": [
        "https://colab.research.google.com/github/HenriquesLab/ZeroCostDL4Mic/blob/master/Colab_notebooks/U-net_2D_ZeroCostDL4Mic.ipynb"
      ],
      "format": "uri",
      "maxLength": 2083,
      "minLength": 1,
      "title": "HttpUrl",
      "type": "string"
    }
  },
  "required": [
    "label",
    "url"
  ],
  "title": "generic.v0_2.BadgeDescr",
  "type": "object"
}

Fields:

• label (Annotated[str, Field(examples=['Open in Colab'])])
• icon (Annotated[Annotated[FilePath | RelativeFilePath, AfterValidator(wo_special_file_name), include_in_package] | HttpUrl | pydantic.HttpUrl | None, Field(examples=['https://colab.research.google.com/assets/colab-badge.svg'])])
• url (Annotated[HttpUrl, Field(examples=['https://colab.research.google.com/github/HenriquesLab/ZeroCostDL4Mic/blob/master/Colab_notebooks/U-net_2D_ZeroCostDL4Mic.ipynb'])])

icon pydantic-field

icon: Annotated[
    Annotated[
        FilePath | RelativeFilePath,
        AfterValidator(wo_special_file_name),
        include_in_package,
    ]
    | HttpUrl
    | pydantic.HttpUrl
    | None,
    Field(
        examples=[
            "https://colab.research.google.com/assets/colab-badge.svg"
        ]
    ),
] = None

badge icon (included in bioimage.io package if not a URL)

label pydantic-field

label: Annotated[str, Field(examples=['Open in Colab'])]

badge label to display on hover

url pydantic-field

url: Annotated[
    HttpUrl,
    Field(
        examples=[
            "https://colab.research.google.com/github/HenriquesLab/ZeroCostDL4Mic/blob/master/Colab_notebooks/U-net_2D_ZeroCostDL4Mic.ipynb"
        ]
    ),
]

target URL

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

BatchAxis pydantic-model

Bases: AxisBase

Show JSON schema:

{
  "additionalProperties": false,
  "properties": {
    "id": {
      "default": "batch",
      "maxLength": 16,
      "minLength": 1,
      "title": "AxisId",
      "type": "string"
    },
    "description": {
      "default": "",
      "description": "A short description of this axis beyond its type and id.",
      "maxLength": 128,
      "title": "Description",
      "type": "string"
    },
    "type": {
      "const": "batch",
      "title": "Type",
      "type": "string"
    },
    "size": {
      "anyOf": [
        {
          "const": 1,
          "type": "integer"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The batch size may be fixed to 1,\notherwise (the default) it may be chosen arbitrarily depending on available memory",
      "title": "Size"
    }
  },
  "required": [
    "type"
  ],
  "title": "model.v0_5.BatchAxis",
  "type": "object"
}

Fields:

• description (Annotated[str, MaxLen(128)])
• type (Literal['batch'])
• id (Annotated[AxisId, Predicate(_is_batch)])
• size (Literal[1] | None)

concatenable property

concatenable

description pydantic-field

description: Annotated[str, MaxLen(128)] = ''

A short description of this axis beyond its type and id.

id pydantic-field

id: Annotated[AxisId, Predicate(_is_batch)] = BATCH_AXIS_ID

An axis id unique across all axes of one tensor.

implemented_type class-attribute

implemented_type: Literal['batch'] = 'batch'

scale property

scale

size pydantic-field

size: Literal[1] | None = None

The batch size may be fixed to 1, otherwise (the default) it may be chosen arbitrarily depending on available memory

type pydantic-field

type: Literal['batch'] = 'batch'

unit property

unit

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

BiasRisksLimitations pydantic-model

Bases: Node

Known biases, risks, technical limitations, and recommendations for model use.

Show JSON schema:

{
  "additionalProperties": true,
  "description": "Known biases, risks, technical limitations, and recommendations for model use.",
  "properties": {
    "known_biases": {
      "default": "In general bioimage models may suffer from biases caused by:\n\n- Imaging protocol dependencies\n- Use of a specific cell type\n- Species-specific training data limitations\n\n",
      "description": "Biases in training data or model behavior.",
      "title": "Known Biases",
      "type": "string"
    },
    "risks": {
      "default": "Common risks in bioimage analysis include:\n\n- Erroneously assuming generalization to unseen experimental conditions\n- Trusting (overconfident) model outputs without validation\n- Misinterpretation of results\n\n",
      "description": "Potential risks in the context of bioimage analysis.",
      "title": "Risks",
      "type": "string"
    },
    "limitations": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Technical limitations and failure modes.",
      "title": "Limitations"
    },
    "recommendations": {
      "default": "Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model.",
      "description": "Mitigation strategies regarding `known_biases`, `risks`, and `limitations`, as well as applicable best practices.\n\nConsider:\n- How to use a validation dataset?\n- How to manually validate?\n- Feasibility of domain adaptation for different experimental setups?",
      "title": "Recommendations",
      "type": "string"
    }
  },
  "title": "model.v0_5.BiasRisksLimitations",
  "type": "object"
}

Fields:

• known_biases (str)
• risks (str)
• limitations (str | None)
• recommendations (str)

known_biases pydantic-field

known_biases: str

Biases in training data or model behavior.

limitations pydantic-field

limitations: str | None = None

Technical limitations and failure modes.

recommendations pydantic-field

recommendations: str = "Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model."

Mitigation strategies regarding known_biases, risks, and limitations, as well as applicable best practices.

Consider: - How to use a validation dataset? - How to manually validate? - Feasibility of domain adaptation for different experimental setups?

risks pydantic-field

risks: str

Potential risks in the context of bioimage analysis.

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

format_md

format_md() -> str

Source code in src/bioimageio/spec/model/v0_5.py

    def format_md(self) -> str:
        if self.limitations is None:
            limitations_header = ""
        else:
            limitations_header = "## Limitations\n\n"

        return f"""# Bias, Risks, and Limitations

{self.known_biases}

{self.risks}

{limitations_header}{self.limitations or ""}

## Recommendations

{self.recommendations}

"""

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

BinarizeAlongAxisKwargs pydantic-model

Bases: KwargsNode

key word arguments for BinarizeDescr

Show JSON schema:

{
  "additionalProperties": false,
  "description": "key word arguments for [BinarizeDescr][]",
  "properties": {
    "threshold": {
      "description": "The fixed threshold values along `axis`",
      "items": {
        "type": "number"
      },
      "minItems": 1,
      "title": "Threshold",
      "type": "array"
    },
    "axis": {
      "description": "The `threshold` axis",
      "examples": [
        "channel"
      ],
      "maxLength": 16,
      "minLength": 1,
      "title": "AxisId",
      "type": "string"
    }
  },
  "required": [
    "threshold",
    "axis"
  ],
  "title": "model.v0_5.BinarizeAlongAxisKwargs",
  "type": "object"
}

Fields:

• threshold (NotEmpty[list[float]])
• axis (Annotated[NonBatchAxisId, Field(examples=['channel'])])

axis pydantic-field

axis: Annotated[NonBatchAxisId, Field(examples=["channel"])]

The threshold axis

threshold pydantic-field

threshold: NotEmpty[list[float]]

The fixed threshold values along axis

Referenced by:

•  v0_5  BinarizeDescr

__contains__

__contains__(item: str) -> bool

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __contains__(self, item: str) -> bool:
    return item in self.__class__.model_fields

__getitem__

__getitem__(item: str) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __getitem__(self, item: str) -> Any:
    if item in self.__class__.model_fields:
        return getattr(self, item)
    else:
        raise KeyError(item)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get

get(item: str, default: Any = None) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def get(self, item: str, default: Any = None) -> Any:
    return self[item] if item in self else default  # ruff: ignore[SIM401]

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

BinarizeDescr pydantic-model

Bases: NodeWithExplicitlySetFields

Binarize the tensor with a fixed threshold.

Values above BinarizeKwargs.threshold/BinarizeAlongAxisKwargs.threshold will be set to one, values below the threshold to zero.

Examples:

• in YAML

  postprocessing:
    - id: binarize
      kwargs:
        axis: 'channel'
        threshold: [0.25, 0.5, 0.75]
  

• in Python:

  >>> postprocessing = [BinarizeDescr(
  ...   kwargs=BinarizeAlongAxisKwargs(
  ...       axis=AxisId('channel'),
  ...       threshold=[0.25, 0.5, 0.75],
  ...   )
  ... )]
  

Show JSON schema:

{
  "$defs": {
    "BinarizeAlongAxisKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [BinarizeDescr][]",
      "properties": {
        "threshold": {
          "description": "The fixed threshold values along `axis`",
          "items": {
            "type": "number"
          },
          "minItems": 1,
          "title": "Threshold",
          "type": "array"
        },
        "axis": {
          "description": "The `threshold` axis",
          "examples": [
            "channel"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        }
      },
      "required": [
        "threshold",
        "axis"
      ],
      "title": "model.v0_5.BinarizeAlongAxisKwargs",
      "type": "object"
    },
    "BinarizeKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [BinarizeDescr][]",
      "properties": {
        "threshold": {
          "description": "The fixed threshold",
          "title": "Threshold",
          "type": "number"
        }
      },
      "required": [
        "threshold"
      ],
      "title": "model.v0_5.BinarizeKwargs",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "description": "Binarize the tensor with a fixed threshold.\n\nValues above [BinarizeKwargs.threshold][]/[BinarizeAlongAxisKwargs.threshold][]\nwill be set to one, values below the threshold to zero.\n\nExamples:\n- in YAML\n    ```yaml\n    postprocessing:\n      - id: binarize\n        kwargs:\n          axis: 'channel'\n          threshold: [0.25, 0.5, 0.75]\n    ```\n- in Python:\n\n    >>> postprocessing = [BinarizeDescr(\n    ...   kwargs=BinarizeAlongAxisKwargs(\n    ...       axis=AxisId('channel'),\n    ...       threshold=[0.25, 0.5, 0.75],\n    ...   )\n    ... )]",
  "properties": {
    "id": {
      "const": "binarize",
      "title": "Id",
      "type": "string"
    },
    "kwargs": {
      "anyOf": [
        {
          "$ref": "#/$defs/BinarizeKwargs"
        },
        {
          "$ref": "#/$defs/BinarizeAlongAxisKwargs"
        }
      ],
      "title": "Kwargs"
    }
  },
  "required": [
    "id",
    "kwargs"
  ],
  "title": "model.v0_5.BinarizeDescr",
  "type": "object"
}

Fields:

• id (Literal['binarize'])
• kwargs (BinarizeKwargs | BinarizeAlongAxisKwargs)

id pydantic-field

id: Literal['binarize'] = 'binarize'

implemented_id class-attribute

implemented_id: Literal['binarize'] = 'binarize'

kwargs pydantic-field

kwargs: BinarizeKwargs | BinarizeAlongAxisKwargs

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

BinarizeKwargs pydantic-model

Bases: KwargsNode

key word arguments for BinarizeDescr

Show JSON schema:

{
  "additionalProperties": false,
  "description": "key word arguments for [BinarizeDescr][]",
  "properties": {
    "threshold": {
      "description": "The fixed threshold",
      "title": "Threshold",
      "type": "number"
    }
  },
  "required": [
    "threshold"
  ],
  "title": "model.v0_5.BinarizeKwargs",
  "type": "object"
}

Fields:

• threshold (float)

threshold pydantic-field

threshold: float

The fixed threshold

Referenced by:

•  v0_5  BinarizeDescr

__contains__

__contains__(item: str) -> bool

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __contains__(self, item: str) -> bool:
    return item in self.__class__.model_fields

__getitem__

__getitem__(item: str) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __getitem__(self, item: str) -> Any:
    if item in self.__class__.model_fields:
        return getattr(self, item)
    else:
        raise KeyError(item)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get

get(item: str, default: Any = None) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def get(self, item: str, default: Any = None) -> Any:
    return self[item] if item in self else default  # ruff: ignore[SIM401]

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

BioimageioConfig pydantic-model

Bases: Node

Show JSON schema:

{
  "$defs": {
    "BiasRisksLimitations": {
      "additionalProperties": true,
      "description": "Known biases, risks, technical limitations, and recommendations for model use.",
      "properties": {
        "known_biases": {
          "default": "In general bioimage models may suffer from biases caused by:\n\n- Imaging protocol dependencies\n- Use of a specific cell type\n- Species-specific training data limitations\n\n",
          "description": "Biases in training data or model behavior.",
          "title": "Known Biases",
          "type": "string"
        },
        "risks": {
          "default": "Common risks in bioimage analysis include:\n\n- Erroneously assuming generalization to unseen experimental conditions\n- Trusting (overconfident) model outputs without validation\n- Misinterpretation of results\n\n",
          "description": "Potential risks in the context of bioimage analysis.",
          "title": "Risks",
          "type": "string"
        },
        "limitations": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Technical limitations and failure modes.",
          "title": "Limitations"
        },
        "recommendations": {
          "default": "Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model.",
          "description": "Mitigation strategies regarding `known_biases`, `risks`, and `limitations`, as well as applicable best practices.\n\nConsider:\n- How to use a validation dataset?\n- How to manually validate?\n- Feasibility of domain adaptation for different experimental setups?",
          "title": "Recommendations",
          "type": "string"
        }
      },
      "title": "model.v0_5.BiasRisksLimitations",
      "type": "object"
    },
    "ContinuousReproducibilityTolerance": {
      "additionalProperties": true,
      "properties": {
        "mismatched_elements_per_million": {
          "default": 100,
          "description": "Maximum number of mismatched elements/pixels per million to tolerate.",
          "maximum": 30000,
          "minimum": 0,
          "title": "Mismatched Elements Per Million",
          "type": "integer"
        },
        "weights_formats": {
          "default": [],
          "description": "Limits the weights formats these details apply to.",
          "items": {
            "enum": [
              "keras_hdf5",
              "keras_v3",
              "onnx",
              "pytorch_state_dict",
              "tensorflow_js",
              "tensorflow_saved_model_bundle",
              "torchscript"
            ],
            "type": "string"
          },
          "title": "Weights Formats",
          "type": "array"
        },
        "relative_tolerance": {
          "default": 0.001,
          "description": "Maximum relative tolerance of reproduced test tensor.\nNeeds to be 0 for discrete (integer and boolean) outputs.",
          "maximum": 0.01,
          "minimum": 0,
          "title": "Relative Tolerance",
          "type": "number"
        },
        "absolute_tolerance": {
          "default": 0.001,
          "description": "Maximum absolute tolerance of reproduced test tensor.\n\nThe `absolute_tolerance` may not be greater than 1% of the maximum absolute value of the test tensor\n(this serves to prevent false positives in the presence of very small numbers).\nNeeds to be 0 for discrete (integer and boolean) outputs.",
          "minimum": 0,
          "title": "Absolute Tolerance",
          "type": "number"
        },
        "output_ids": {
          "const": "any_continuous_output",
          "default": "any_continuous_output",
          "description": "Limits the output tensor IDs to all outputs of float data type.",
          "title": "Output Ids",
          "type": "string"
        }
      },
      "title": "model.v0_5.ContinuousReproducibilityTolerance",
      "type": "object"
    },
    "DiscreteReproducibilityTolerance": {
      "additionalProperties": true,
      "properties": {
        "mismatched_elements_per_million": {
          "default": 1000,
          "description": "Maximum number of mismatched elements/pixels per million to tolerate.\n\nNote: Increased default mismatched elements per million compared to\n[`ReproducibilityTolerance`][ReproducibilityTolerance]/[`ContinuousReproducibilityTolerance`][ContinuousReproducibilityTolerance],\nsince a single pixel mismatch can be more significant than for continuous outputs with numeric tolerance.",
          "maximum": 30000,
          "minimum": 0,
          "title": "Mismatched Elements Per Million",
          "type": "integer"
        },
        "weights_formats": {
          "default": [],
          "description": "Limits the weights formats these details apply to.",
          "items": {
            "enum": [
              "keras_hdf5",
              "keras_v3",
              "onnx",
              "pytorch_state_dict",
              "tensorflow_js",
              "tensorflow_saved_model_bundle",
              "torchscript"
            ],
            "type": "string"
          },
          "title": "Weights Formats",
          "type": "array"
        },
        "relative_tolerance": {
          "const": 0,
          "default": 0,
          "description": "Relative tolerance does not apply to discrete outputs",
          "title": "Relative Tolerance",
          "type": "integer"
        },
        "absolute_tolerance": {
          "const": 0,
          "default": 0,
          "description": "Absolute tolerance does not apply to discrete outputs",
          "title": "Absolute Tolerance",
          "type": "integer"
        },
        "output_ids": {
          "const": "any_discrete_output",
          "default": "any_discrete_output",
          "description": "Limits the output tensor IDs to all outputs of integer or boolean data type.",
          "title": "Output Ids",
          "type": "string"
        }
      },
      "title": "model.v0_5.DiscreteReproducibilityTolerance",
      "type": "object"
    },
    "EnvironmentalImpact": {
      "additionalProperties": true,
      "description": "Environmental considerations for model training and deployment.\n\nCarbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).",
      "properties": {
        "hardware_type": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "GPU/CPU specifications",
          "title": "Hardware Type"
        },
        "hours_used": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Total compute hours",
          "title": "Hours Used"
        },
        "cloud_provider": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "If applicable",
          "title": "Cloud Provider"
        },
        "compute_region": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Geographic location",
          "title": "Compute Region"
        },
        "co2_emitted": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "kg CO2 equivalent\n\nCarbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).",
          "title": "Co2 Emitted"
        }
      },
      "title": "model.v0_5.EnvironmentalImpact",
      "type": "object"
    },
    "Evaluation": {
      "additionalProperties": true,
      "properties": {
        "model_id": {
          "anyOf": [
            {
              "minLength": 1,
              "title": "ModelId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Model being evaluated.",
          "title": "Model Id"
        },
        "dataset_id": {
          "description": "Dataset used for evaluation.",
          "minLength": 1,
          "title": "DatasetId",
          "type": "string"
        },
        "dataset_source": {
          "description": "Source of the dataset.",
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "title": "HttpUrl",
          "type": "string"
        },
        "dataset_role": {
          "description": "Role of the dataset used for evaluation.\n\n- `train`: dataset was (part of) the training data\n- `validation`: dataset was (part of) the validation data used during training, e.g. used for model selection or hyperparameter tuning\n- `test`: dataset was (part of) the designated test data; not used during training or validation, but acquired from the same source/distribution as training data\n- `independent`: dataset is entirely independent test data; not used during training or validation, and acquired from a different source/distribution than training data\n- `unknown`: role of the dataset is unknown; choose this if you are not certain if (a subset) of the data was seen by the model during training.",
          "enum": [
            "train",
            "validation",
            "test",
            "independent",
            "unknown"
          ],
          "title": "Dataset Role",
          "type": "string"
        },
        "sample_count": {
          "description": "Number of evaluated samples.",
          "title": "Sample Count",
          "type": "integer"
        },
        "evaluation_factors": {
          "description": "(Abbreviations of) each evaluation factor.\n\nEvaluation factors are criteria along which model performance is evaluated, e.g. different image conditions\nlike 'low SNR', 'high cell density', or different biological conditions like 'cell type A', 'cell type B'.\nAn 'overall' factor may be included to summarize performance across all conditions.",
          "items": {
            "maxLength": 16,
            "type": "string"
          },
          "title": "Evaluation Factors",
          "type": "array"
        },
        "evaluation_factors_long": {
          "description": "Descriptions (long form) of each evaluation factor.",
          "items": {
            "type": "string"
          },
          "title": "Evaluation Factors Long",
          "type": "array"
        },
        "metrics": {
          "description": "(Abbreviations of) metrics used for evaluation.",
          "items": {
            "maxLength": 16,
            "type": "string"
          },
          "title": "Metrics",
          "type": "array"
        },
        "metrics_long": {
          "description": "Description of each metric used.",
          "items": {
            "type": "string"
          },
          "title": "Metrics Long",
          "type": "array"
        },
        "results": {
          "description": "Results for each metric (rows; outer list) and each evaluation factor (columns; inner list).",
          "items": {
            "items": {
              "anyOf": [
                {
                  "type": "string"
                },
                {
                  "type": "number"
                },
                {
                  "type": "integer"
                }
              ]
            },
            "type": "array"
          },
          "title": "Results",
          "type": "array"
        },
        "results_summary": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Interpretation of results for general audience.\n\nConsider:\n    - Overall model performance\n    - Comparison to existing methods\n    - Limitations and areas for improvement",
          "title": "Results Summary"
        }
      },
      "required": [
        "dataset_id",
        "dataset_source",
        "dataset_role",
        "sample_count",
        "evaluation_factors",
        "evaluation_factors_long",
        "metrics",
        "metrics_long",
        "results"
      ],
      "title": "model.v0_5.Evaluation",
      "type": "object"
    },
    "ReproducibilityTolerance": {
      "additionalProperties": true,
      "properties": {
        "mismatched_elements_per_million": {
          "default": 100,
          "description": "Maximum number of mismatched elements/pixels per million to tolerate.",
          "maximum": 30000,
          "minimum": 0,
          "title": "Mismatched Elements Per Million",
          "type": "integer"
        },
        "weights_formats": {
          "default": [],
          "description": "Limits the weights formats these details apply to.",
          "items": {
            "enum": [
              "keras_hdf5",
              "keras_v3",
              "onnx",
              "pytorch_state_dict",
              "tensorflow_js",
              "tensorflow_saved_model_bundle",
              "torchscript"
            ],
            "type": "string"
          },
          "title": "Weights Formats",
          "type": "array"
        },
        "relative_tolerance": {
          "default": 0.001,
          "description": "Maximum relative tolerance of reproduced test tensor.\nNeeds to be 0 for discrete (integer and boolean) outputs.",
          "maximum": 0.01,
          "minimum": 0,
          "title": "Relative Tolerance",
          "type": "number"
        },
        "absolute_tolerance": {
          "default": 0.001,
          "description": "Maximum absolute tolerance of reproduced test tensor.\n\nThe `absolute_tolerance` may not be greater than 1% of the maximum absolute value of the test tensor\n(this serves to prevent false positives in the presence of very small numbers).\nNeeds to be 0 for discrete (integer and boolean) outputs.",
          "minimum": 0,
          "title": "Absolute Tolerance",
          "type": "number"
        },
        "output_ids": {
          "default": [],
          "description": "Limits the output tensor IDs these reproducibility details apply to.\n\nIf empty, the reproducibility details apply to all outputs.",
          "items": {
            "maxLength": 32,
            "minLength": 1,
            "title": "TensorId",
            "type": "string"
          },
          "title": "Output Ids",
          "type": "array"
        }
      },
      "title": "model.v0_5.ReproducibilityTolerance",
      "type": "object"
    },
    "TrainingDetails": {
      "additionalProperties": true,
      "properties": {
        "training_preprocessing": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Detailed image preprocessing steps during model training:\n\nMention:\n- *Normalization methods*\n- *Augmentation strategies*\n- *Resizing/resampling procedures*\n- *Artifact handling*",
          "title": "Training Preprocessing"
        },
        "training_epochs": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Number of training epochs.",
          "title": "Training Epochs"
        },
        "training_batch_size": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Batch size used in training.",
          "title": "Training Batch Size"
        },
        "initial_learning_rate": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Initial learning rate used in training.",
          "title": "Initial Learning Rate"
        },
        "learning_rate_schedule": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Learning rate schedule used in training.",
          "title": "Learning Rate Schedule"
        },
        "loss_function": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Loss function used in training, e.g. nn.MSELoss.",
          "title": "Loss Function"
        },
        "loss_function_kwargs": {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "description": "key word arguments for the `loss_function`",
          "title": "Loss Function Kwargs",
          "type": "object"
        },
        "optimizer": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "optimizer, e.g. torch.optim.Adam",
          "title": "Optimizer"
        },
        "optimizer_kwargs": {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "description": "key word arguments for the `optimizer`",
          "title": "Optimizer Kwargs",
          "type": "object"
        },
        "regularization": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Regularization techniques used during training, e.g. drop-out or weight decay.",
          "title": "Regularization"
        },
        "training_duration": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Total training duration in hours.",
          "title": "Training Duration"
        }
      },
      "title": "model.v0_5.TrainingDetails",
      "type": "object"
    },
    "YamlValue": {
      "anyOf": [
        {
          "type": "boolean"
        },
        {
          "format": "date",
          "type": "string"
        },
        {
          "format": "date-time",
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        },
        {
          "type": "string"
        },
        {
          "items": {
            "$ref": "#/$defs/YamlValue"
          },
          "type": "array"
        },
        {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "type": "object"
        },
        {
          "type": "null"
        }
      ]
    }
  },
  "additionalProperties": true,
  "properties": {
    "reproducibility_tolerance": {
      "description": "Tolerances to allow when reproducing the model's test outputs\nfrom the model's test inputs.\nOnly the first entry matching tensor id and weights format is considered.\nFor defaults see [`ContinuousReproducibilityTolerance`][ContinuousReproducibilityTolerance]/[`DiscreteReproducibilityTolerance`][DiscreteReproducibilityTolerance]",
      "items": {
        "anyOf": [
          {
            "$ref": "#/$defs/ReproducibilityTolerance"
          },
          {
            "$ref": "#/$defs/ContinuousReproducibilityTolerance"
          },
          {
            "$ref": "#/$defs/DiscreteReproducibilityTolerance"
          }
        ]
      },
      "title": "Reproducibility Tolerance",
      "type": "array"
    },
    "funded_by": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Funding agency, grant number if applicable",
      "title": "Funded By"
    },
    "architecture_type": {
      "anyOf": [
        {
          "maxLength": 32,
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Model architecture type, e.g., 3D U-Net, ResNet, transformer",
      "title": "Architecture Type"
    },
    "architecture_description": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Text description of model architecture.",
      "title": "Architecture Description"
    },
    "modality": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Input modality, e.g., fluorescence microscopy, electron microscopy",
      "title": "Modality"
    },
    "target_structure": {
      "description": "Biological structure(s) the model is designed to analyze, e.g., nuclei, mitochondria, cells",
      "items": {
        "type": "string"
      },
      "title": "Target Structure",
      "type": "array"
    },
    "task": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Bioimage-specific task type, e.g., segmentation, classification, detection, denoising",
      "title": "Task"
    },
    "new_version": {
      "anyOf": [
        {
          "minLength": 1,
          "title": "ModelId",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "A new version of this model exists with a different model id.",
      "title": "New Version"
    },
    "out_of_scope_use": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Describe how the model may be misused in bioimage analysis contexts and what users should **not** do with the model.",
      "title": "Out Of Scope Use"
    },
    "bias_risks_limitations": {
      "$ref": "#/$defs/BiasRisksLimitations",
      "description": "Description of known bias, risks, and technical limitations for in-scope model use."
    },
    "model_parameter_count": {
      "anyOf": [
        {
          "type": "integer"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Total number of model parameters.",
      "title": "Model Parameter Count"
    },
    "training": {
      "$ref": "#/$defs/TrainingDetails",
      "description": "Details on how the model was trained."
    },
    "inference_time": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Average inference time per image/tile. Specify hardware and image size. Multiple examples can be given.",
      "title": "Inference Time"
    },
    "memory_requirements_inference": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "GPU memory needed for inference. Multiple examples with different image size can be given.",
      "title": "Memory Requirements Inference"
    },
    "memory_requirements_training": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "GPU memory needed for training. Multiple examples with different image/batch sizes can be given.",
      "title": "Memory Requirements Training"
    },
    "evaluations": {
      "description": "Quantitative model evaluations.\n\nNote:\n    At the moment we recommend to include only a single test dataset\n    (with evaluation factors that may mark subsets of the dataset)\n    to avoid confusion and make the presentation of results cleaner.",
      "items": {
        "$ref": "#/$defs/Evaluation"
      },
      "title": "Evaluations",
      "type": "array"
    },
    "environmental_impact": {
      "$ref": "#/$defs/EnvironmentalImpact",
      "description": "Environmental considerations for model training and deployment"
    }
  },
  "title": "model.v0_5.BioimageioConfig",
  "type": "object"
}

Fields:

• reproducibility_tolerance (Sequence[ReproducibilityTolerance | ContinuousReproducibilityTolerance | DiscreteReproducibilityTolerance])
• funded_by (str | None)
• architecture_type (Annotated[str, MaxLen(32)] | None)
• architecture_description (str | None)
• modality (str | None)
• target_structure (list[str])
• task (str | None)
• new_version (ModelId | None)
• out_of_scope_use (str | None)
• bias_risks_limitations (BiasRisksLimitations)
• model_parameter_count (int | None)
• training (TrainingDetails)
• inference_time (str | None)
• memory_requirements_inference (str | None)
• memory_requirements_training (str | None)
• evaluations (list[Evaluation])
• environmental_impact (EnvironmentalImpact)

architecture_description pydantic-field

architecture_description: str | None = None

Text description of model architecture.

architecture_type pydantic-field

architecture_type: Annotated[str, MaxLen(32)] | None = None

Model architecture type, e.g., 3D U-Net, ResNet, transformer

bias_risks_limitations pydantic-field

bias_risks_limitations: BiasRisksLimitations

Description of known bias, risks, and technical limitations for in-scope model use.

environmental_impact pydantic-field

environmental_impact: EnvironmentalImpact

Environmental considerations for model training and deployment

evaluations pydantic-field

evaluations: list[Evaluation]

Quantitative model evaluations.

Note

At the moment we recommend to include only a single test dataset (with evaluation factors that may mark subsets of the dataset) to avoid confusion and make the presentation of results cleaner.

funded_by pydantic-field

funded_by: str | None = None

Funding agency, grant number if applicable

inference_time pydantic-field

inference_time: str | None = None

Average inference time per image/tile. Specify hardware and image size. Multiple examples can be given.

memory_requirements_inference pydantic-field

memory_requirements_inference: str | None = None

GPU memory needed for inference. Multiple examples with different image size can be given.

memory_requirements_training pydantic-field

memory_requirements_training: str | None = None

GPU memory needed for training. Multiple examples with different image/batch sizes can be given.

modality pydantic-field

modality: str | None = None

Input modality, e.g., fluorescence microscopy, electron microscopy

model_parameter_count pydantic-field

model_parameter_count: int | None = None

Total number of model parameters.

new_version pydantic-field

new_version: ModelId | None = None

A new version of this model exists with a different model id.

out_of_scope_use pydantic-field

out_of_scope_use: str | None = None

Describe how the model may be misused in bioimage analysis contexts and what users should not do with the model.

reproducibility_tolerance pydantic-field

reproducibility_tolerance: Sequence[
    ReproducibilityTolerance
    | ContinuousReproducibilityTolerance
    | DiscreteReproducibilityTolerance
]

Tolerances to allow when reproducing the model's test outputs from the model's test inputs. Only the first entry matching tensor id and weights format is considered. For defaults see ContinuousReproducibilityTolerance/DiscreteReproducibilityTolerance

target_structure pydantic-field

target_structure: list[str]

Biological structure(s) the model is designed to analyze, e.g., nuclei, mitochondria, cells

task pydantic-field

task: str | None = None

Bioimage-specific task type, e.g., segmentation, classification, detection, denoising

training pydantic-field

training: TrainingDetails

Details on how the model was trained.

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

CallableFromDepencency

Bases: ValidatedStringWithInnerNode[CallableFromDepencencyNode]

              flowchart TD
              bioimageio.spec.model.v0_5.CallableFromDepencency[CallableFromDepencency]
              bioimageio.spec._internal.validated_string_with_inner_node.ValidatedStringWithInnerNode[ValidatedStringWithInnerNode]
              bioimageio.spec._internal.validated_string.ValidatedString[ValidatedString]

                              bioimageio.spec._internal.validated_string_with_inner_node.ValidatedStringWithInnerNode --> bioimageio.spec.model.v0_5.CallableFromDepencency
                                bioimageio.spec._internal.validated_string.ValidatedString --> bioimageio.spec._internal.validated_string_with_inner_node.ValidatedStringWithInnerNode
                



              click bioimageio.spec.model.v0_5.CallableFromDepencency href "" "bioimageio.spec.model.v0_5.CallableFromDepencency"
              click bioimageio.spec._internal.validated_string_with_inner_node.ValidatedStringWithInnerNode href "" "bioimageio.spec._internal.validated_string_with_inner_node.ValidatedStringWithInnerNode"
              click bioimageio.spec._internal.validated_string.ValidatedString href "" "bioimageio.spec._internal.validated_string.ValidatedString"
            

Used by:

•  v0_4  CustomCallable

Methods:

Name	Description
__get_pydantic_core_schema__
__get_pydantic_json_schema__
__new__

Attributes:

Name	Type	Description
callable_name		The callable Python identifier implemented in module module_name.
module_name		The Python module that implements callable_name.
root_model	type[RootModel[Any]]	the pydantic root model to validate the string

callable_name property

callable_name

The callable Python identifier implemented in module module_name.

module_name property

module_name

The Python module that implements callable_name.

root_model class-attribute

root_model: type[RootModel[Any]] = RootModel[
    Annotated[
        str,
        StringConstraints(
            strip_whitespace=True, pattern="^.+\\..+$"
        ),
    ]
]

the pydantic root model to validate the string

__get_pydantic_core_schema__ classmethod

__get_pydantic_core_schema__(
    source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_core_schema__(
    cls, source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema:
    return no_info_after_validator_function(cls, handler(str))

__get_pydantic_json_schema__ classmethod

__get_pydantic_json_schema__(
    core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_json_schema__(
    cls, core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue:
    json_schema = cls.root_model.model_json_schema(mode=handler.mode)
    json_schema["title"] = cls.__name__.strip("_")
    if cls.__doc__:
        json_schema["description"] = cls.__doc__

    return json_schema

__new__

__new__(object: object)

Source code in src/bioimageio/spec/_internal/validated_string.py

def __new__(cls, object: object):
    _validated = cls.root_model.model_validate(object).root
    self = super().__new__(cls, _validated)
    self._validated = _validated
    return self._after_validator()

CellposeFlowDynamicsDescr pydantic-model

Bases: NodeWithExplicitlySetFields

Cellpose flow dynamics postprocessing as described in: - Carsen Stringer and Marius Pachitariu. Cellpose: a generalist algorithm for cellular segmentation. Nature Methods, 2021.

Note: Only available if the cellpose package is installed.

Show JSON schema:

{
  "$defs": {
    "CellposeFlowDynamicsKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [CellposeFlowDynamicsDescr][]",
      "properties": {
        "cellprob_threshold": {
          "title": "Cellprob Threshold",
          "type": "number"
        },
        "flow_threshold": {
          "title": "Flow Threshold",
          "type": "number"
        },
        "do_3D": {
          "title": "Do 3D",
          "type": "boolean"
        },
        "min_size": {
          "default": 15,
          "description": "Minimum size of objects to keep, in pixels. Default is 15, which is the default in Cellpose. Set to 0 to disable filtering by size.",
          "title": "Min Size",
          "type": "integer"
        },
        "output_dtype": {
          "default": "uint16",
          "enum": [
            "uint16",
            "uint32"
          ],
          "title": "Output Dtype",
          "type": "string"
        }
      },
      "required": [
        "cellprob_threshold",
        "flow_threshold",
        "do_3D"
      ],
      "title": "model.v0_5.CellposeFlowDynamicsKwargs",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "description": "Cellpose flow dynamics postprocessing as described in:\n- Carsen Stringer and Marius Pachitariu. [*Cellpose: a generalist algorithm for cellular segmentation*](https://www.nature.com/articles/s41592-020-01018-x). Nature Methods, 2021.\n\nNote: Only available if the `cellpose` package is installed.",
  "properties": {
    "id": {
      "const": "cellpose_flow_dynamics",
      "title": "Id",
      "type": "string"
    },
    "kwargs": {
      "$ref": "#/$defs/CellposeFlowDynamicsKwargs"
    }
  },
  "required": [
    "id",
    "kwargs"
  ],
  "title": "model.v0_5.CellposeFlowDynamicsDescr",
  "type": "object"
}

Fields:

• id (Literal['cellpose_flow_dynamics'])
• kwargs (CellposeFlowDynamicsKwargs)

id pydantic-field

id: Literal["cellpose_flow_dynamics"] = (
    "cellpose_flow_dynamics"
)

implemented_id class-attribute

implemented_id: Literal["cellpose_flow_dynamics"] = (
    "cellpose_flow_dynamics"
)

kwargs pydantic-field

kwargs: CellposeFlowDynamicsKwargs

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

CellposeFlowDynamicsKwargs pydantic-model

Bases: KwargsNode

key word arguments for CellposeFlowDynamicsDescr

Show JSON schema:

{
  "additionalProperties": false,
  "description": "key word arguments for [CellposeFlowDynamicsDescr][]",
  "properties": {
    "cellprob_threshold": {
      "title": "Cellprob Threshold",
      "type": "number"
    },
    "flow_threshold": {
      "title": "Flow Threshold",
      "type": "number"
    },
    "do_3D": {
      "title": "Do 3D",
      "type": "boolean"
    },
    "min_size": {
      "default": 15,
      "description": "Minimum size of objects to keep, in pixels. Default is 15, which is the default in Cellpose. Set to 0 to disable filtering by size.",
      "title": "Min Size",
      "type": "integer"
    },
    "output_dtype": {
      "default": "uint16",
      "enum": [
        "uint16",
        "uint32"
      ],
      "title": "Output Dtype",
      "type": "string"
    }
  },
  "required": [
    "cellprob_threshold",
    "flow_threshold",
    "do_3D"
  ],
  "title": "model.v0_5.CellposeFlowDynamicsKwargs",
  "type": "object"
}

Fields:

• cellprob_threshold (float)
• flow_threshold (float)
• do_3D (bool)
• min_size (int)
• output_dtype (Literal['uint16', 'uint32'])

cellprob_threshold pydantic-field

cellprob_threshold: float

do_3D pydantic-field

do_3D: bool

flow_threshold pydantic-field

flow_threshold: float

min_size pydantic-field

min_size: int = 15

Minimum size of objects to keep, in pixels. Default is 15, which is the default in Cellpose. Set to 0 to disable filtering by size.

output_dtype pydantic-field

output_dtype: Literal['uint16', 'uint32'] = 'uint16'

__contains__

__contains__(item: str) -> bool

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __contains__(self, item: str) -> bool:
    return item in self.__class__.model_fields

__getitem__

__getitem__(item: str) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __getitem__(self, item: str) -> Any:
    if item in self.__class__.model_fields:
        return getattr(self, item)
    else:
        raise KeyError(item)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get

get(item: str, default: Any = None) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def get(self, item: str, default: Any = None) -> Any:
    return self[item] if item in self else default  # ruff: ignore[SIM401]

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

ChannelAxis pydantic-model

Bases: AxisBase

Show JSON schema:

{
  "additionalProperties": false,
  "properties": {
    "id": {
      "default": "channel",
      "maxLength": 16,
      "minLength": 1,
      "title": "AxisId",
      "type": "string"
    },
    "description": {
      "default": "",
      "description": "A short description of this axis beyond its type and id.",
      "maxLength": 128,
      "title": "Description",
      "type": "string"
    },
    "type": {
      "const": "channel",
      "title": "Type",
      "type": "string"
    },
    "channel_names": {
      "description": "Name/label for each channel. The number of channels is given by `len(channel_names)`.",
      "items": {
        "type": "string"
      },
      "minItems": 1,
      "title": "Channel Names",
      "type": "array"
    }
  },
  "required": [
    "type",
    "channel_names"
  ],
  "title": "model.v0_5.ChannelAxis",
  "type": "object"
}

Fields:

• description (Annotated[str, MaxLen(128)])
• type (Literal['channel'])
• id (NonBatchAxisId)
• channel_names (NotEmpty[list[str]])

channel_names pydantic-field

channel_names: NotEmpty[list[str]]

Name/label for each channel. The number of channels is given by len(channel_names).

concatenable property

concatenable

description pydantic-field

description: Annotated[str, MaxLen(128)] = ''

A short description of this axis beyond its type and id.

id pydantic-field

id: NonBatchAxisId = CHANNEL_AXIS_ID

An axis id unique across all axes of one tensor.

implemented_type class-attribute

implemented_type: Literal['channel'] = 'channel'

scale property

scale: float

size property

size: int

type pydantic-field

type: Literal['channel'] = 'channel'

unit property

unit

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

CiteEntry pydantic-model

Bases: Node

A citation that should be referenced in work using this resource.

Show JSON schema:

{
  "additionalProperties": false,
  "description": "A citation that should be referenced in work using this resource.",
  "properties": {
    "text": {
      "description": "free text description",
      "title": "Text",
      "type": "string"
    },
    "doi": {
      "anyOf": [
        {
          "description": "A digital object identifier, see https://www.doi.org/",
          "pattern": "^10\\.[0-9]{4}.+$",
          "title": "Doi",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "A digital object identifier (DOI) is the prefered citation reference.\nSee https://www.doi.org/ for details.\nNote:\n    Either **doi** or **url** have to be specified.",
      "title": "Doi"
    },
    "url": {
      "anyOf": [
        {
          "description": "A URL with the HTTP or HTTPS scheme.",
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "title": "HttpUrl",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "URL to cite (preferably specify a **doi** instead/also).\nNote:\n    Either **doi** or **url** have to be specified.",
      "title": "Url"
    }
  },
  "required": [
    "text"
  ],
  "title": "generic.v0_3.CiteEntry",
  "type": "object"
}

Fields:

• text (str)
• doi (Doi | None)
• url (HttpUrl | None)

Validators:

• _check_doi_or_url

doi pydantic-field

doi: Doi | None = None

A digital object identifier (DOI) is the prefered citation reference. See https://www.doi.org/ for details. Note: Either doi or url have to be specified.

text pydantic-field

text: str

free text description

url pydantic-field

url: HttpUrl | None = None

URL to cite (preferably specify a doi instead/also). Note: Either doi or url have to be specified.

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

ClipDescr pydantic-model

Bases: NodeWithExplicitlySetFields

Set tensor values below min to min and above max to max.

See ScaleRangeDescr for examples.

Show JSON schema:

{
  "$defs": {
    "ClipKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [ClipDescr][]",
      "properties": {
        "min": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Minimum value for clipping.\n\nExclusive with [min_percentile][]",
          "title": "Min"
        },
        "min_percentile": {
          "anyOf": [
            {
              "exclusiveMaximum": 100,
              "minimum": 0,
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Minimum percentile for clipping.\n\nExclusive with [min][].\n\nIn range [0, 100).",
          "title": "Min Percentile"
        },
        "max": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Maximum value for clipping.\n\nExclusive with `max_percentile`.",
          "title": "Max"
        },
        "max_percentile": {
          "anyOf": [
            {
              "exclusiveMinimum": 1,
              "maximum": 100,
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Maximum percentile for clipping.\n\nExclusive with `max`.\n\nIn range (1, 100].",
          "title": "Max Percentile"
        },
        "axes": {
          "anyOf": [
            {
              "items": {
                "maxLength": 16,
                "minLength": 1,
                "title": "AxisId",
                "type": "string"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The subset of axes to determine percentiles jointly,\n\ni.e. axes to reduce to compute min/max from `min_percentile`/`max_percentile`.\nFor example to clip 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x')\nresulting in a tensor of equal shape with clipped values per channel, specify `axes=('batch', 'x', 'y')`.\nTo clip samples independently, leave out the 'batch' axis.\n\nOnly valid if `min_percentile` and/or `max_percentile` are set.\n\nDefault: Compute percentiles over all axes jointly.",
          "examples": [
            [
              "batch",
              "x",
              "y"
            ]
          ],
          "title": "Axes"
        }
      },
      "title": "model.v0_5.ClipKwargs",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "description": "Set tensor values below min to min and above max to max.\n\nSee `ScaleRangeDescr` for examples.",
  "properties": {
    "id": {
      "const": "clip",
      "title": "Id",
      "type": "string"
    },
    "kwargs": {
      "$ref": "#/$defs/ClipKwargs"
    }
  },
  "required": [
    "id",
    "kwargs"
  ],
  "title": "model.v0_5.ClipDescr",
  "type": "object"
}

Fields:

• id (Literal['clip'])
• kwargs (ClipKwargs)

id pydantic-field

id: Literal['clip'] = 'clip'

implemented_id class-attribute

implemented_id: Literal['clip'] = 'clip'

kwargs pydantic-field

kwargs: ClipKwargs

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

ClipKwargs pydantic-model

Bases: KwargsNode

key word arguments for ClipDescr

Show JSON schema:

{
  "additionalProperties": false,
  "description": "key word arguments for [ClipDescr][]",
  "properties": {
    "min": {
      "anyOf": [
        {
          "type": "number"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Minimum value for clipping.\n\nExclusive with [min_percentile][]",
      "title": "Min"
    },
    "min_percentile": {
      "anyOf": [
        {
          "exclusiveMaximum": 100,
          "minimum": 0,
          "type": "number"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Minimum percentile for clipping.\n\nExclusive with [min][].\n\nIn range [0, 100).",
      "title": "Min Percentile"
    },
    "max": {
      "anyOf": [
        {
          "type": "number"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Maximum value for clipping.\n\nExclusive with `max_percentile`.",
      "title": "Max"
    },
    "max_percentile": {
      "anyOf": [
        {
          "exclusiveMinimum": 1,
          "maximum": 100,
          "type": "number"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Maximum percentile for clipping.\n\nExclusive with `max`.\n\nIn range (1, 100].",
      "title": "Max Percentile"
    },
    "axes": {
      "anyOf": [
        {
          "items": {
            "maxLength": 16,
            "minLength": 1,
            "title": "AxisId",
            "type": "string"
          },
          "type": "array"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The subset of axes to determine percentiles jointly,\n\ni.e. axes to reduce to compute min/max from `min_percentile`/`max_percentile`.\nFor example to clip 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x')\nresulting in a tensor of equal shape with clipped values per channel, specify `axes=('batch', 'x', 'y')`.\nTo clip samples independently, leave out the 'batch' axis.\n\nOnly valid if `min_percentile` and/or `max_percentile` are set.\n\nDefault: Compute percentiles over all axes jointly.",
      "examples": [
        [
          "batch",
          "x",
          "y"
        ]
      ],
      "title": "Axes"
    }
  },
  "title": "model.v0_5.ClipKwargs",
  "type": "object"
}

Fields:

• min (float | None)
• min_percentile (Annotated[float, Interval(ge=0, lt=100)] | None)
• max (float | None)
• max_percentile (Annotated[float, Interval(gt=1, le=100)] | None)
• axes (Annotated[Sequence[AxisId] | None, Field(examples=[('batch', 'x', 'y')])])

Validators:

• _validate

axes pydantic-field

axes: Annotated[
    Sequence[AxisId] | None,
    Field(examples=[("batch", "x", "y")]),
] = None

The subset of axes to determine percentiles jointly,

i.e. axes to reduce to compute min/max from min_percentile/max_percentile. For example to clip 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x') resulting in a tensor of equal shape with clipped values per channel, specify axes=('batch', 'x', 'y'). To clip samples independently, leave out the 'batch' axis.

Only valid if min_percentile and/or max_percentile are set.

Default: Compute percentiles over all axes jointly.

max pydantic-field

max: float | None = None

Maximum value for clipping.

Exclusive with max_percentile.

max_percentile pydantic-field

max_percentile: (
    Annotated[float, Interval(gt=1, le=100)] | None
) = None

Maximum percentile for clipping.

Exclusive with max.

In range (1, 100].

min pydantic-field

min: float | None = None

Minimum value for clipping.

Exclusive with min_percentile

min_percentile pydantic-field

min_percentile: (
    Annotated[float, Interval(ge=0, lt=100)] | None
) = None

Minimum percentile for clipping.

Exclusive with min.

In range [0, 100).

__contains__

__contains__(item: str) -> bool

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __contains__(self, item: str) -> bool:
    return item in self.__class__.model_fields

__getitem__

__getitem__(item: str) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __getitem__(self, item: str) -> Any:
    if item in self.__class__.model_fields:
        return getattr(self, item)
    else:
        raise KeyError(item)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get

get(item: str, default: Any = None) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def get(self, item: str, default: Any = None) -> Any:
    return self[item] if item in self else default  # ruff: ignore[SIM401]

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

Config pydantic-model

Bases: Node

Show JSON schema:

{
  "$defs": {
    "BiasRisksLimitations": {
      "additionalProperties": true,
      "description": "Known biases, risks, technical limitations, and recommendations for model use.",
      "properties": {
        "known_biases": {
          "default": "In general bioimage models may suffer from biases caused by:\n\n- Imaging protocol dependencies\n- Use of a specific cell type\n- Species-specific training data limitations\n\n",
          "description": "Biases in training data or model behavior.",
          "title": "Known Biases",
          "type": "string"
        },
        "risks": {
          "default": "Common risks in bioimage analysis include:\n\n- Erroneously assuming generalization to unseen experimental conditions\n- Trusting (overconfident) model outputs without validation\n- Misinterpretation of results\n\n",
          "description": "Potential risks in the context of bioimage analysis.",
          "title": "Risks",
          "type": "string"
        },
        "limitations": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Technical limitations and failure modes.",
          "title": "Limitations"
        },
        "recommendations": {
          "default": "Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model.",
          "description": "Mitigation strategies regarding `known_biases`, `risks`, and `limitations`, as well as applicable best practices.\n\nConsider:\n- How to use a validation dataset?\n- How to manually validate?\n- Feasibility of domain adaptation for different experimental setups?",
          "title": "Recommendations",
          "type": "string"
        }
      },
      "title": "model.v0_5.BiasRisksLimitations",
      "type": "object"
    },
    "BioimageioConfig": {
      "additionalProperties": true,
      "properties": {
        "reproducibility_tolerance": {
          "description": "Tolerances to allow when reproducing the model's test outputs\nfrom the model's test inputs.\nOnly the first entry matching tensor id and weights format is considered.\nFor defaults see [`ContinuousReproducibilityTolerance`][ContinuousReproducibilityTolerance]/[`DiscreteReproducibilityTolerance`][DiscreteReproducibilityTolerance]",
          "items": {
            "anyOf": [
              {
                "$ref": "#/$defs/ReproducibilityTolerance"
              },
              {
                "$ref": "#/$defs/ContinuousReproducibilityTolerance"
              },
              {
                "$ref": "#/$defs/DiscreteReproducibilityTolerance"
              }
            ]
          },
          "title": "Reproducibility Tolerance",
          "type": "array"
        },
        "funded_by": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Funding agency, grant number if applicable",
          "title": "Funded By"
        },
        "architecture_type": {
          "anyOf": [
            {
              "maxLength": 32,
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Model architecture type, e.g., 3D U-Net, ResNet, transformer",
          "title": "Architecture Type"
        },
        "architecture_description": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Text description of model architecture.",
          "title": "Architecture Description"
        },
        "modality": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Input modality, e.g., fluorescence microscopy, electron microscopy",
          "title": "Modality"
        },
        "target_structure": {
          "description": "Biological structure(s) the model is designed to analyze, e.g., nuclei, mitochondria, cells",
          "items": {
            "type": "string"
          },
          "title": "Target Structure",
          "type": "array"
        },
        "task": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Bioimage-specific task type, e.g., segmentation, classification, detection, denoising",
          "title": "Task"
        },
        "new_version": {
          "anyOf": [
            {
              "minLength": 1,
              "title": "ModelId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "A new version of this model exists with a different model id.",
          "title": "New Version"
        },
        "out_of_scope_use": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Describe how the model may be misused in bioimage analysis contexts and what users should **not** do with the model.",
          "title": "Out Of Scope Use"
        },
        "bias_risks_limitations": {
          "$ref": "#/$defs/BiasRisksLimitations",
          "description": "Description of known bias, risks, and technical limitations for in-scope model use."
        },
        "model_parameter_count": {
          "anyOf": [
            {
              "type": "integer"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Total number of model parameters.",
          "title": "Model Parameter Count"
        },
        "training": {
          "$ref": "#/$defs/TrainingDetails",
          "description": "Details on how the model was trained."
        },
        "inference_time": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Average inference time per image/tile. Specify hardware and image size. Multiple examples can be given.",
          "title": "Inference Time"
        },
        "memory_requirements_inference": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "GPU memory needed for inference. Multiple examples with different image size can be given.",
          "title": "Memory Requirements Inference"
        },
        "memory_requirements_training": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "GPU memory needed for training. Multiple examples with different image/batch sizes can be given.",
          "title": "Memory Requirements Training"
        },
        "evaluations": {
          "description": "Quantitative model evaluations.\n\nNote:\n    At the moment we recommend to include only a single test dataset\n    (with evaluation factors that may mark subsets of the dataset)\n    to avoid confusion and make the presentation of results cleaner.",
          "items": {
            "$ref": "#/$defs/Evaluation"
          },
          "title": "Evaluations",
          "type": "array"
        },
        "environmental_impact": {
          "$ref": "#/$defs/EnvironmentalImpact",
          "description": "Environmental considerations for model training and deployment"
        }
      },
      "title": "model.v0_5.BioimageioConfig",
      "type": "object"
    },
    "ContinuousReproducibilityTolerance": {
      "additionalProperties": true,
      "properties": {
        "mismatched_elements_per_million": {
          "default": 100,
          "description": "Maximum number of mismatched elements/pixels per million to tolerate.",
          "maximum": 30000,
          "minimum": 0,
          "title": "Mismatched Elements Per Million",
          "type": "integer"
        },
        "weights_formats": {
          "default": [],
          "description": "Limits the weights formats these details apply to.",
          "items": {
            "enum": [
              "keras_hdf5",
              "keras_v3",
              "onnx",
              "pytorch_state_dict",
              "tensorflow_js",
              "tensorflow_saved_model_bundle",
              "torchscript"
            ],
            "type": "string"
          },
          "title": "Weights Formats",
          "type": "array"
        },
        "relative_tolerance": {
          "default": 0.001,
          "description": "Maximum relative tolerance of reproduced test tensor.\nNeeds to be 0 for discrete (integer and boolean) outputs.",
          "maximum": 0.01,
          "minimum": 0,
          "title": "Relative Tolerance",
          "type": "number"
        },
        "absolute_tolerance": {
          "default": 0.001,
          "description": "Maximum absolute tolerance of reproduced test tensor.\n\nThe `absolute_tolerance` may not be greater than 1% of the maximum absolute value of the test tensor\n(this serves to prevent false positives in the presence of very small numbers).\nNeeds to be 0 for discrete (integer and boolean) outputs.",
          "minimum": 0,
          "title": "Absolute Tolerance",
          "type": "number"
        },
        "output_ids": {
          "const": "any_continuous_output",
          "default": "any_continuous_output",
          "description": "Limits the output tensor IDs to all outputs of float data type.",
          "title": "Output Ids",
          "type": "string"
        }
      },
      "title": "model.v0_5.ContinuousReproducibilityTolerance",
      "type": "object"
    },
    "DiscreteReproducibilityTolerance": {
      "additionalProperties": true,
      "properties": {
        "mismatched_elements_per_million": {
          "default": 1000,
          "description": "Maximum number of mismatched elements/pixels per million to tolerate.\n\nNote: Increased default mismatched elements per million compared to\n[`ReproducibilityTolerance`][ReproducibilityTolerance]/[`ContinuousReproducibilityTolerance`][ContinuousReproducibilityTolerance],\nsince a single pixel mismatch can be more significant than for continuous outputs with numeric tolerance.",
          "maximum": 30000,
          "minimum": 0,
          "title": "Mismatched Elements Per Million",
          "type": "integer"
        },
        "weights_formats": {
          "default": [],
          "description": "Limits the weights formats these details apply to.",
          "items": {
            "enum": [
              "keras_hdf5",
              "keras_v3",
              "onnx",
              "pytorch_state_dict",
              "tensorflow_js",
              "tensorflow_saved_model_bundle",
              "torchscript"
            ],
            "type": "string"
          },
          "title": "Weights Formats",
          "type": "array"
        },
        "relative_tolerance": {
          "const": 0,
          "default": 0,
          "description": "Relative tolerance does not apply to discrete outputs",
          "title": "Relative Tolerance",
          "type": "integer"
        },
        "absolute_tolerance": {
          "const": 0,
          "default": 0,
          "description": "Absolute tolerance does not apply to discrete outputs",
          "title": "Absolute Tolerance",
          "type": "integer"
        },
        "output_ids": {
          "const": "any_discrete_output",
          "default": "any_discrete_output",
          "description": "Limits the output tensor IDs to all outputs of integer or boolean data type.",
          "title": "Output Ids",
          "type": "string"
        }
      },
      "title": "model.v0_5.DiscreteReproducibilityTolerance",
      "type": "object"
    },
    "EnvironmentalImpact": {
      "additionalProperties": true,
      "description": "Environmental considerations for model training and deployment.\n\nCarbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).",
      "properties": {
        "hardware_type": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "GPU/CPU specifications",
          "title": "Hardware Type"
        },
        "hours_used": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Total compute hours",
          "title": "Hours Used"
        },
        "cloud_provider": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "If applicable",
          "title": "Cloud Provider"
        },
        "compute_region": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Geographic location",
          "title": "Compute Region"
        },
        "co2_emitted": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "kg CO2 equivalent\n\nCarbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).",
          "title": "Co2 Emitted"
        }
      },
      "title": "model.v0_5.EnvironmentalImpact",
      "type": "object"
    },
    "Evaluation": {
      "additionalProperties": true,
      "properties": {
        "model_id": {
          "anyOf": [
            {
              "minLength": 1,
              "title": "ModelId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Model being evaluated.",
          "title": "Model Id"
        },
        "dataset_id": {
          "description": "Dataset used for evaluation.",
          "minLength": 1,
          "title": "DatasetId",
          "type": "string"
        },
        "dataset_source": {
          "description": "Source of the dataset.",
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "title": "HttpUrl",
          "type": "string"
        },
        "dataset_role": {
          "description": "Role of the dataset used for evaluation.\n\n- `train`: dataset was (part of) the training data\n- `validation`: dataset was (part of) the validation data used during training, e.g. used for model selection or hyperparameter tuning\n- `test`: dataset was (part of) the designated test data; not used during training or validation, but acquired from the same source/distribution as training data\n- `independent`: dataset is entirely independent test data; not used during training or validation, and acquired from a different source/distribution than training data\n- `unknown`: role of the dataset is unknown; choose this if you are not certain if (a subset) of the data was seen by the model during training.",
          "enum": [
            "train",
            "validation",
            "test",
            "independent",
            "unknown"
          ],
          "title": "Dataset Role",
          "type": "string"
        },
        "sample_count": {
          "description": "Number of evaluated samples.",
          "title": "Sample Count",
          "type": "integer"
        },
        "evaluation_factors": {
          "description": "(Abbreviations of) each evaluation factor.\n\nEvaluation factors are criteria along which model performance is evaluated, e.g. different image conditions\nlike 'low SNR', 'high cell density', or different biological conditions like 'cell type A', 'cell type B'.\nAn 'overall' factor may be included to summarize performance across all conditions.",
          "items": {
            "maxLength": 16,
            "type": "string"
          },
          "title": "Evaluation Factors",
          "type": "array"
        },
        "evaluation_factors_long": {
          "description": "Descriptions (long form) of each evaluation factor.",
          "items": {
            "type": "string"
          },
          "title": "Evaluation Factors Long",
          "type": "array"
        },
        "metrics": {
          "description": "(Abbreviations of) metrics used for evaluation.",
          "items": {
            "maxLength": 16,
            "type": "string"
          },
          "title": "Metrics",
          "type": "array"
        },
        "metrics_long": {
          "description": "Description of each metric used.",
          "items": {
            "type": "string"
          },
          "title": "Metrics Long",
          "type": "array"
        },
        "results": {
          "description": "Results for each metric (rows; outer list) and each evaluation factor (columns; inner list).",
          "items": {
            "items": {
              "anyOf": [
                {
                  "type": "string"
                },
                {
                  "type": "number"
                },
                {
                  "type": "integer"
                }
              ]
            },
            "type": "array"
          },
          "title": "Results",
          "type": "array"
        },
        "results_summary": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Interpretation of results for general audience.\n\nConsider:\n    - Overall model performance\n    - Comparison to existing methods\n    - Limitations and areas for improvement",
          "title": "Results Summary"
        }
      },
      "required": [
        "dataset_id",
        "dataset_source",
        "dataset_role",
        "sample_count",
        "evaluation_factors",
        "evaluation_factors_long",
        "metrics",
        "metrics_long",
        "results"
      ],
      "title": "model.v0_5.Evaluation",
      "type": "object"
    },
    "ReproducibilityTolerance": {
      "additionalProperties": true,
      "properties": {
        "mismatched_elements_per_million": {
          "default": 100,
          "description": "Maximum number of mismatched elements/pixels per million to tolerate.",
          "maximum": 30000,
          "minimum": 0,
          "title": "Mismatched Elements Per Million",
          "type": "integer"
        },
        "weights_formats": {
          "default": [],
          "description": "Limits the weights formats these details apply to.",
          "items": {
            "enum": [
              "keras_hdf5",
              "keras_v3",
              "onnx",
              "pytorch_state_dict",
              "tensorflow_js",
              "tensorflow_saved_model_bundle",
              "torchscript"
            ],
            "type": "string"
          },
          "title": "Weights Formats",
          "type": "array"
        },
        "relative_tolerance": {
          "default": 0.001,
          "description": "Maximum relative tolerance of reproduced test tensor.\nNeeds to be 0 for discrete (integer and boolean) outputs.",
          "maximum": 0.01,
          "minimum": 0,
          "title": "Relative Tolerance",
          "type": "number"
        },
        "absolute_tolerance": {
          "default": 0.001,
          "description": "Maximum absolute tolerance of reproduced test tensor.\n\nThe `absolute_tolerance` may not be greater than 1% of the maximum absolute value of the test tensor\n(this serves to prevent false positives in the presence of very small numbers).\nNeeds to be 0 for discrete (integer and boolean) outputs.",
          "minimum": 0,
          "title": "Absolute Tolerance",
          "type": "number"
        },
        "output_ids": {
          "default": [],
          "description": "Limits the output tensor IDs these reproducibility details apply to.\n\nIf empty, the reproducibility details apply to all outputs.",
          "items": {
            "maxLength": 32,
            "minLength": 1,
            "title": "TensorId",
            "type": "string"
          },
          "title": "Output Ids",
          "type": "array"
        }
      },
      "title": "model.v0_5.ReproducibilityTolerance",
      "type": "object"
    },
    "TrainingDetails": {
      "additionalProperties": true,
      "properties": {
        "training_preprocessing": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Detailed image preprocessing steps during model training:\n\nMention:\n- *Normalization methods*\n- *Augmentation strategies*\n- *Resizing/resampling procedures*\n- *Artifact handling*",
          "title": "Training Preprocessing"
        },
        "training_epochs": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Number of training epochs.",
          "title": "Training Epochs"
        },
        "training_batch_size": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Batch size used in training.",
          "title": "Training Batch Size"
        },
        "initial_learning_rate": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Initial learning rate used in training.",
          "title": "Initial Learning Rate"
        },
        "learning_rate_schedule": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Learning rate schedule used in training.",
          "title": "Learning Rate Schedule"
        },
        "loss_function": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Loss function used in training, e.g. nn.MSELoss.",
          "title": "Loss Function"
        },
        "loss_function_kwargs": {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "description": "key word arguments for the `loss_function`",
          "title": "Loss Function Kwargs",
          "type": "object"
        },
        "optimizer": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "optimizer, e.g. torch.optim.Adam",
          "title": "Optimizer"
        },
        "optimizer_kwargs": {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "description": "key word arguments for the `optimizer`",
          "title": "Optimizer Kwargs",
          "type": "object"
        },
        "regularization": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Regularization techniques used during training, e.g. drop-out or weight decay.",
          "title": "Regularization"
        },
        "training_duration": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Total training duration in hours.",
          "title": "Training Duration"
        }
      },
      "title": "model.v0_5.TrainingDetails",
      "type": "object"
    },
    "YamlValue": {
      "anyOf": [
        {
          "type": "boolean"
        },
        {
          "format": "date",
          "type": "string"
        },
        {
          "format": "date-time",
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        },
        {
          "type": "string"
        },
        {
          "items": {
            "$ref": "#/$defs/YamlValue"
          },
          "type": "array"
        },
        {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "type": "object"
        },
        {
          "type": "null"
        }
      ]
    }
  },
  "additionalProperties": true,
  "properties": {
    "bioimageio": {
      "$ref": "#/$defs/BioimageioConfig"
    },
    "stardist": {
      "$ref": "#/$defs/YamlValue",
      "default": null
    }
  },
  "title": "model.v0_5.Config",
  "type": "object"
}

Fields:

• bioimageio (BioimageioConfig)
• stardist (YamlValue)

bioimageio pydantic-field

bioimageio: BioimageioConfig

stardist pydantic-field

stardist: YamlValue = None

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

ConstantPadding pydantic-model

Bases: Node

Show JSON schema:

{
  "additionalProperties": false,
  "properties": {
    "mode": {
      "const": "constant",
      "default": "constant",
      "title": "Mode",
      "type": "string"
    },
    "value": {
      "anyOf": [
        {
          "type": "integer"
        },
        {
          "type": "number"
        }
      ],
      "default": 0,
      "title": "Value"
    }
  },
  "title": "model.v0_5.ConstantPadding",
  "type": "object"
}

Fields:

• mode (Literal['constant'])
• value (int | float)

mode pydantic-field

mode: Literal['constant'] = 'constant'

value pydantic-field

value: int | float = 0

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

ContinuousReproducibilityTolerance pydantic-model

Bases: _ReproducibilityToleranceBase[RelativeTolerance, AbsoluteTolerance]

Show JSON schema:

{
  "additionalProperties": true,
  "properties": {
    "mismatched_elements_per_million": {
      "default": 100,
      "description": "Maximum number of mismatched elements/pixels per million to tolerate.",
      "maximum": 30000,
      "minimum": 0,
      "title": "Mismatched Elements Per Million",
      "type": "integer"
    },
    "weights_formats": {
      "default": [],
      "description": "Limits the weights formats these details apply to.",
      "items": {
        "enum": [
          "keras_hdf5",
          "keras_v3",
          "onnx",
          "pytorch_state_dict",
          "tensorflow_js",
          "tensorflow_saved_model_bundle",
          "torchscript"
        ],
        "type": "string"
      },
      "title": "Weights Formats",
      "type": "array"
    },
    "relative_tolerance": {
      "default": 0.001,
      "description": "Maximum relative tolerance of reproduced test tensor.\nNeeds to be 0 for discrete (integer and boolean) outputs.",
      "maximum": 0.01,
      "minimum": 0,
      "title": "Relative Tolerance",
      "type": "number"
    },
    "absolute_tolerance": {
      "default": 0.001,
      "description": "Maximum absolute tolerance of reproduced test tensor.\n\nThe `absolute_tolerance` may not be greater than 1% of the maximum absolute value of the test tensor\n(this serves to prevent false positives in the presence of very small numbers).\nNeeds to be 0 for discrete (integer and boolean) outputs.",
      "minimum": 0,
      "title": "Absolute Tolerance",
      "type": "number"
    },
    "output_ids": {
      "const": "any_continuous_output",
      "default": "any_continuous_output",
      "description": "Limits the output tensor IDs to all outputs of float data type.",
      "title": "Output Ids",
      "type": "string"
    }
  },
  "title": "model.v0_5.ContinuousReproducibilityTolerance",
  "type": "object"
}

Fields:

• mismatched_elements_per_million (MismatchedElementsPerMillion)
• weights_formats (Sequence[WeightsFormat])
• output_ids (Literal['any_continuous_output'])
• relative_tolerance (RelativeTolerance)
• absolute_tolerance (AbsoluteTolerance)

absolute_tolerance pydantic-field

absolute_tolerance: AbsoluteTolerance = 0.001

Maximum absolute tolerance of reproduced test tensor.

The absolute_tolerance may not be greater than 1% of the maximum absolute value of the test tensor (this serves to prevent false positives in the presence of very small numbers). Needs to be 0 for discrete (integer and boolean) outputs.

mismatched_elements_per_million pydantic-field

mismatched_elements_per_million: MismatchedElementsPerMillion = 100

Maximum number of mismatched elements/pixels per million to tolerate.

output_ids pydantic-field

output_ids: Literal["any_continuous_output"] = (
    "any_continuous_output"
)

Limits the output tensor IDs to all outputs of float data type.

relative_tolerance pydantic-field

relative_tolerance: RelativeTolerance = 0.001

Maximum relative tolerance of reproduced test tensor. Needs to be 0 for discrete (integer and boolean) outputs.

weights_formats pydantic-field

weights_formats: Sequence[WeightsFormat] = ()

Limits the weights formats these details apply to.

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

serialize

serialize(
    handler: SerializerFunctionWrapHandler,
) -> dict[str, Any]

explicitly serialize tolerance values

(even if they are default values and exclude_defaults=True).

Source code in src/bioimageio/spec/model/v0_5.py

@model_serializer(mode="wrap")
def serialize(
    self,
    handler: SerializerFunctionWrapHandler,
) -> dict[str, Any]:
    """explicitly serialize tolerance values

    (even if they are default values and `exclude_defaults=True`).
    """
    data = handler(self)
    data["relative_tolerance"] = self.relative_tolerance
    data["absolute_tolerance"] = self.absolute_tolerance
    data["mismatched_elements_per_million"] = self.mismatched_elements_per_million
    return data

CustomProcessingDescr pydantic-model

Bases: NodeWithExplicitlySetFields, FileDescr

Custom (post)processing op — source file shipped inline with the model.

Supports (post)processing that cannot be expressed by the built-in named operations (watershed, connected components, etc.) using a simple Python callable interface.

The op is implemented in a .py file packaged alongside the model weights. Two styles are supported:

Callable class — kwargs go to __init__, tensors arrive in __call__:

.. code-block:: python

# my_postprocess.py
import numpy as np

class my_postprocess:
    def __init__(self, threshold: float = 0.5) -> None:
        self.threshold = threshold
    def __call__(self, *arrays: np.ndarray) -> np.ndarray:
        # arrays = model output tensors in rdf.yaml declaration order
        return (arrays[0] > self.threshold).astype(np.uint8)

Factory function — alternative closure style, identical runtime behaviour:

.. code-block:: python

# my_postprocess.py
import numpy as np

def my_postprocess(threshold: float = 0.5):
    def run(*arrays: np.ndarray) -> np.ndarray:
        return (arrays[0] > threshold).astype(np.uint8)
    return run

Reference it in rdf.yaml with the source file included in the package:

.. code-block:: yaml

postprocessing:
  - id: custom
    callable: my_postprocess   # class or function name in source
    source: my_postprocess.py  # packaged alongside weights
    sha256: <hash>             # sha256 of the source file
    kwargs:                    # forwarded to __init__ / factory
      threshold: 0.5

Security: source files are SHA-256 verified before execution. Execution requires explicit opt-in in bioimageio.core and curator review before Zoo publication.

Show JSON schema:

{
  "$defs": {
    "RelativeFilePath": {
      "description": "A path relative to the `rdf.yaml` file (also if the RDF source is a URL).",
      "format": "path",
      "title": "RelativeFilePath",
      "type": "string"
    },
    "YamlValue": {
      "anyOf": [
        {
          "type": "boolean"
        },
        {
          "format": "date",
          "type": "string"
        },
        {
          "format": "date-time",
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        },
        {
          "type": "string"
        },
        {
          "items": {
            "$ref": "#/$defs/YamlValue"
          },
          "type": "array"
        },
        {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "type": "object"
        },
        {
          "type": "null"
        }
      ]
    }
  },
  "additionalProperties": false,
  "description": "Custom (post)processing op \u2014 source file shipped inline with the model.\n\nSupports (post)processing that cannot be expressed by the built-in named\noperations (watershed, connected components, etc.)\nusing a simple Python callable interface.\n\nThe op is implemented in a ``.py`` file packaged alongside the model weights.\nTwo styles are supported:\n\n*Callable class* \u2014 kwargs go to ``__init__``, tensors arrive in ``__call__``:\n\n.. code-block:: python\n\n    # my_postprocess.py\n    import numpy as np\n\n    class my_postprocess:\n        def __init__(self, threshold: float = 0.5) -> None:\n            self.threshold = threshold\n        def __call__(self, *arrays: np.ndarray) -> np.ndarray:\n            # arrays = model output tensors in rdf.yaml declaration order\n            return (arrays[0] > self.threshold).astype(np.uint8)\n\n*Factory function* \u2014 alternative closure style, identical runtime behaviour:\n\n.. code-block:: python\n\n    # my_postprocess.py\n    import numpy as np\n\n    def my_postprocess(threshold: float = 0.5):\n        def run(*arrays: np.ndarray) -> np.ndarray:\n            return (arrays[0] > threshold).astype(np.uint8)\n        return run\n\nReference it in ``rdf.yaml`` with the source file included in the package:\n\n.. code-block:: yaml\n\n    postprocessing:\n      - id: custom\n        callable: my_postprocess   # class or function name in source\n        source: my_postprocess.py  # packaged alongside weights\n        sha256: <hash>             # sha256 of the source file\n        kwargs:                    # forwarded to __init__ / factory\n          threshold: 0.5\n\n**Security:** source files are SHA-256 verified before execution.\nExecution requires explicit opt-in in bioimageio.core and curator\nreview before Zoo publication.",
  "properties": {
    "source": {
      "anyOf": [
        {
          "description": "A URL with the HTTP or HTTPS scheme.",
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "title": "HttpUrl",
          "type": "string"
        },
        {
          "$ref": "#/$defs/RelativeFilePath"
        },
        {
          "format": "file-path",
          "title": "FilePath",
          "type": "string"
        }
      ],
      "description": "Python source file (included when packaging the model).",
      "title": "FileSource"
    },
    "sha256": {
      "anyOf": [
        {
          "description": "A SHA-256 hash value",
          "maxLength": 64,
          "minLength": 64,
          "title": "Sha256",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "SHA256 hash value of the **source** file.",
      "title": "Sha256"
    },
    "id": {
      "const": "custom",
      "title": "Id",
      "type": "string"
    },
    "callable": {
      "description": "Name of the callable class or factory function defined in ``source``.\n\nAt runtime: ``op = callable(**kwargs)``, then ``result = op(*output_tensors)``\nper image.  Both a class with ``__call__`` and a factory function returning\na callable satisfy this protocol.",
      "examples": [
        "my_postprocess_factory",
        "MyPostprocessClass"
      ],
      "title": "Callable",
      "type": "string"
    },
    "kwargs": {
      "additionalProperties": {
        "$ref": "#/$defs/YamlValue"
      },
      "description": "Keyword arguments forwarded to the callable (``__init__`` or factory).",
      "title": "Kwargs",
      "type": "object"
    }
  },
  "required": [
    "source",
    "id",
    "callable"
  ],
  "title": "model.v0_5.CustomProcessingDescr",
  "type": "object"
}

Fields:

• sha256 (Sha256 | None)
• id (Literal['custom'])
• callable (Annotated[str, Field(examples=['my_postprocess_factory', 'MyPostprocessClass'])])
• source (Annotated[FileSource, AfterValidator(wo_special_file_name)])
• kwargs (dict[str, YamlValue])

callable pydantic-field

callable: Annotated[
    str,
    Field(
        examples=[
            "my_postprocess_factory",
            "MyPostprocessClass",
        ]
    ),
]

Name of the callable class or factory function defined in source.

At runtime: op = callable(**kwargs), then result = op(*output_tensors) per image. Both a class with __call__ and a factory function returning a callable satisfy this protocol.

id pydantic-field

id: Literal['custom'] = 'custom'

implemented_id class-attribute

implemented_id: Literal['custom'] = 'custom'

kwargs pydantic-field

kwargs: dict[str, YamlValue]

Keyword arguments forwarded to the callable (__init__ or factory).

sha256 pydantic-field

sha256: Sha256 | None = None

SHA256 hash value of the source file.

source pydantic-field

source: Annotated[
    FileSource, AfterValidator(wo_special_file_name)
]

Python source file (included when packaging the model).

suffix property

suffix: str

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

download

download(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

alias for .get_reader

Source code in src/bioimageio/spec/_internal/io.py

def download(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """alias for `.get_reader`"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

get_reader

get_reader(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

open the file source (download if needed)

Source code in src/bioimageio/spec/_internal/io.py

def get_reader(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """open the file source (download if needed)"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

validate_sha256

validate_sha256(force_recompute: bool = False) -> None

validate the sha256 hash value of the source file

Source code in src/bioimageio/spec/_internal/io.py

def validate_sha256(self, force_recompute: bool = False) -> None:
    """validate the sha256 hash value of the **source** file"""
    context = get_validation_context()
    src_str = str(self.source)
    if force_recompute:
        actual_sha = None
    else:
        actual_sha = context.known_files.get(src_str)

    if actual_sha is None:
        if context.perform_io_checks or force_recompute:
            reader = get_reader(self.source, sha256=self.sha256)
            if force_recompute:
                actual_sha = get_sha256(reader)
            else:
                actual_sha = reader.sha256

            context.known_files[src_str] = actual_sha
        elif context.known_files and src_str not in context.known_files:
            # perform_io_checks is False, but known files were given,
            # so we expect all file references to be in there
            raise ValueError(f"File {src_str} not found in `known_files`.")

    if actual_sha is None or self.sha256 == actual_sha:
        return
    elif self.sha256 is None or context.update_hashes:
        self.sha256 = actual_sha
    elif self.sha256 != actual_sha:
        raise ValueError(
            f"Sha256 mismatch for {self.source}. Expected {self.sha256}, got "
            + f"{actual_sha}. Update expected `sha256` or point to the matching "
            + "file."
        )

DataDependentSize pydantic-model

Bases: Node

Show JSON schema:

{
  "additionalProperties": false,
  "properties": {
    "min": {
      "default": 1,
      "exclusiveMinimum": 0,
      "title": "Min",
      "type": "integer"
    },
    "max": {
      "anyOf": [
        {
          "exclusiveMinimum": 1,
          "type": "integer"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Max"
    }
  },
  "title": "model.v0_5.DataDependentSize",
  "type": "object"
}

Fields:

• min (Annotated[int, Gt(0)])
• max (Annotated[int | None, Gt(1)])

Validators:

• _validate_max_gt_min

max pydantic-field

max: Annotated[int | None, Gt(1)] = None

min pydantic-field

min: Annotated[int, Gt(0)] = 1

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

validate_size

validate_size(size: int, msg_prefix: str = '') -> int

Source code in src/bioimageio/spec/model/v0_5.py

def validate_size(self, size: int, msg_prefix: str = "") -> int:
    if size < self.min:
        raise ValueError(f"{msg_prefix}size {size} < {self.min}")

    if self.max is not None and size > self.max:
        raise ValueError(f"{msg_prefix}size {size} > {self.max}")

    return size

DatasetDescr pydantic-model

Bases: GenericDescrBase

A bioimage.io dataset resource description file (dataset RDF) describes a dataset relevant to bioimage processing.

Show JSON schema:

{
  "$defs": {
    "Author": {
      "additionalProperties": false,
      "properties": {
        "affiliation": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Affiliation",
          "title": "Affiliation"
        },
        "email": {
          "anyOf": [
            {
              "format": "email",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Email",
          "title": "Email"
        },
        "orcid": {
          "anyOf": [
            {
              "description": "An ORCID identifier, see https://orcid.org/",
              "title": "OrcidId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "An [ORCID iD](https://support.orcid.org/hc/en-us/sections/360001495313-What-is-ORCID\n) in hyphenated groups of 4 digits, (and [valid](\nhttps://support.orcid.org/hc/en-us/articles/360006897674-Structure-of-the-ORCID-Identifier\n) as per ISO 7064 11,2.)",
          "examples": [
            "0000-0001-2345-6789"
          ],
          "title": "Orcid"
        },
        "name": {
          "title": "Name",
          "type": "string"
        },
        "github_user": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Github User"
        }
      },
      "required": [
        "name"
      ],
      "title": "generic.v0_3.Author",
      "type": "object"
    },
    "BadgeDescr": {
      "additionalProperties": false,
      "description": "A custom badge",
      "properties": {
        "label": {
          "description": "badge label to display on hover",
          "examples": [
            "Open in Colab"
          ],
          "title": "Label",
          "type": "string"
        },
        "icon": {
          "anyOf": [
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "badge icon (included in bioimage.io package if not a URL)",
          "examples": [
            "https://colab.research.google.com/assets/colab-badge.svg"
          ],
          "title": "Icon"
        },
        "url": {
          "description": "target URL",
          "examples": [
            "https://colab.research.google.com/github/HenriquesLab/ZeroCostDL4Mic/blob/master/Colab_notebooks/U-net_2D_ZeroCostDL4Mic.ipynb"
          ],
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "title": "HttpUrl",
          "type": "string"
        }
      },
      "required": [
        "label",
        "url"
      ],
      "title": "generic.v0_2.BadgeDescr",
      "type": "object"
    },
    "BioimageioConfig": {
      "additionalProperties": true,
      "description": "bioimage.io internal metadata.",
      "properties": {},
      "title": "generic.v0_3.BioimageioConfig",
      "type": "object"
    },
    "CiteEntry": {
      "additionalProperties": false,
      "description": "A citation that should be referenced in work using this resource.",
      "properties": {
        "text": {
          "description": "free text description",
          "title": "Text",
          "type": "string"
        },
        "doi": {
          "anyOf": [
            {
              "description": "A digital object identifier, see https://www.doi.org/",
              "pattern": "^10\\.[0-9]{4}.+$",
              "title": "Doi",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "A digital object identifier (DOI) is the prefered citation reference.\nSee https://www.doi.org/ for details.\nNote:\n    Either **doi** or **url** have to be specified.",
          "title": "Doi"
        },
        "url": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "URL to cite (preferably specify a **doi** instead/also).\nNote:\n    Either **doi** or **url** have to be specified.",
          "title": "Url"
        }
      },
      "required": [
        "text"
      ],
      "title": "generic.v0_3.CiteEntry",
      "type": "object"
    },
    "Config": {
      "additionalProperties": true,
      "description": "A place to store additional metadata (often tool specific).\n\nSuch additional metadata is typically set programmatically by the respective tool\nor by people with specific insights into the tool.\nIf you want to store additional metadata that does not match any of the other\nfields, think of a key unlikely to collide with anyone elses use-case/tool and save\nit here.\n\nPlease consider creating [an issue in the bioimageio.spec repository](https://github.com/bioimage-io/spec-bioimage-io/issues/new?template=Blank+issue)\nif you are not sure if an existing field could cover your use case\nor if you think such a field should exist.",
      "properties": {
        "bioimageio": {
          "$ref": "#/$defs/BioimageioConfig"
        }
      },
      "title": "generic.v0_3.Config",
      "type": "object"
    },
    "FileDescr": {
      "additionalProperties": false,
      "description": "A file description",
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "File source",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        }
      },
      "required": [
        "source"
      ],
      "title": "_internal.io.FileDescr",
      "type": "object"
    },
    "Maintainer": {
      "additionalProperties": false,
      "properties": {
        "affiliation": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Affiliation",
          "title": "Affiliation"
        },
        "email": {
          "anyOf": [
            {
              "format": "email",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Email",
          "title": "Email"
        },
        "orcid": {
          "anyOf": [
            {
              "description": "An ORCID identifier, see https://orcid.org/",
              "title": "OrcidId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "An [ORCID iD](https://support.orcid.org/hc/en-us/sections/360001495313-What-is-ORCID\n) in hyphenated groups of 4 digits, (and [valid](\nhttps://support.orcid.org/hc/en-us/articles/360006897674-Structure-of-the-ORCID-Identifier\n) as per ISO 7064 11,2.)",
          "examples": [
            "0000-0001-2345-6789"
          ],
          "title": "Orcid"
        },
        "name": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Name"
        },
        "github_user": {
          "title": "Github User",
          "type": "string"
        }
      },
      "required": [
        "github_user"
      ],
      "title": "generic.v0_3.Maintainer",
      "type": "object"
    },
    "RelativeFilePath": {
      "description": "A path relative to the `rdf.yaml` file (also if the RDF source is a URL).",
      "format": "path",
      "title": "RelativeFilePath",
      "type": "string"
    },
    "Uploader": {
      "additionalProperties": false,
      "properties": {
        "email": {
          "description": "Email",
          "format": "email",
          "title": "Email",
          "type": "string"
        },
        "name": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "name",
          "title": "Name"
        }
      },
      "required": [
        "email"
      ],
      "title": "generic.v0_2.Uploader",
      "type": "object"
    },
    "Version": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        }
      ],
      "description": "wraps a packaging.version.Version instance for validation in pydantic models",
      "title": "Version"
    }
  },
  "additionalProperties": false,
  "description": "A bioimage.io dataset resource description file (dataset RDF) describes a dataset relevant to bioimage\nprocessing.",
  "properties": {
    "name": {
      "description": "A human-friendly name of the resource description.\nMay only contains letters, digits, underscore, minus, parentheses and spaces.",
      "maxLength": 128,
      "minLength": 5,
      "title": "Name",
      "type": "string"
    },
    "description": {
      "default": "",
      "description": "A string containing a brief description.",
      "maxLength": 1024,
      "title": "Description",
      "type": "string"
    },
    "covers": {
      "description": "Cover images. Please use an image smaller than 500KB and an aspect ratio width to height of 2:1 or 1:1.\nThe supported image formats are: ('.gif', '.jpeg', '.jpg', '.png', '.svg')",
      "examples": [
        [
          {
            "source": "cover.png"
          }
        ]
      ],
      "items": {
        "$ref": "#/$defs/FileDescr"
      },
      "title": "Covers",
      "type": "array"
    },
    "documentation": {
      "anyOf": [
        {
          "$ref": "#/$defs/FileDescr",
          "examples": [
            {
              "source": "README.md"
            }
          ]
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Additional model documentation.\nThe recommended documentation source file name is `README.md`. An `.md` suffix is mandatory."
    },
    "id_emoji": {
      "anyOf": [
        {
          "examples": [
            "\ud83e\udd88",
            "\ud83e\udda5"
          ],
          "maxLength": 2,
          "minLength": 1,
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "UTF-8 emoji for display alongside the `id`.",
      "title": "Id Emoji"
    },
    "authors": {
      "description": "The authors are the creators of this resource description and the primary points of contact.",
      "items": {
        "$ref": "#/$defs/Author"
      },
      "title": "Authors",
      "type": "array"
    },
    "attachments": {
      "description": "file attachments",
      "items": {
        "$ref": "#/$defs/FileDescr"
      },
      "title": "Attachments",
      "type": "array"
    },
    "cite": {
      "description": "citations",
      "items": {
        "$ref": "#/$defs/CiteEntry"
      },
      "title": "Cite",
      "type": "array"
    },
    "license": {
      "anyOf": [
        {
          "enum": [
            "0BSD",
            "3D-Slicer-1.0",
            "AAL",
            "Abstyles",
            "AdaCore-doc",
            "Adobe-2006",
            "Adobe-Display-PostScript",
            "Adobe-Glyph",
            "Adobe-Utopia",
            "ADSL",
            "AFL-1.1",
            "AFL-1.2",
            "AFL-2.0",
            "AFL-2.1",
            "AFL-3.0",
            "Afmparse",
            "AGPL-1.0-only",
            "AGPL-1.0-or-later",
            "AGPL-3.0-only",
            "AGPL-3.0-or-later",
            "Aladdin",
            "AMD-newlib",
            "AMDPLPA",
            "AML",
            "AML-glslang",
            "AMPAS",
            "ANTLR-PD",
            "ANTLR-PD-fallback",
            "any-OSI",
            "any-OSI-perl-modules",
            "Apache-1.0",
            "Apache-1.1",
            "Apache-2.0",
            "APAFML",
            "APL-1.0",
            "App-s2p",
            "APSL-1.0",
            "APSL-1.1",
            "APSL-1.2",
            "APSL-2.0",
            "Arphic-1999",
            "Artistic-1.0",
            "Artistic-1.0-cl8",
            "Artistic-1.0-Perl",
            "Artistic-2.0",
            "Artistic-dist",
            "Aspell-RU",
            "ASWF-Digital-Assets-1.0",
            "ASWF-Digital-Assets-1.1",
            "Baekmuk",
            "Bahyph",
            "Barr",
            "bcrypt-Solar-Designer",
            "Beerware",
            "Bitstream-Charter",
            "Bitstream-Vera",
            "BitTorrent-1.0",
            "BitTorrent-1.1",
            "blessing",
            "BlueOak-1.0.0",
            "Boehm-GC",
            "Boehm-GC-without-fee",
            "Borceux",
            "Brian-Gladman-2-Clause",
            "Brian-Gladman-3-Clause",
            "BSD-1-Clause",
            "BSD-2-Clause",
            "BSD-2-Clause-Darwin",
            "BSD-2-Clause-first-lines",
            "BSD-2-Clause-Patent",
            "BSD-2-Clause-pkgconf-disclaimer",
            "BSD-2-Clause-Views",
            "BSD-3-Clause",
            "BSD-3-Clause-acpica",
            "BSD-3-Clause-Attribution",
            "BSD-3-Clause-Clear",
            "BSD-3-Clause-flex",
            "BSD-3-Clause-HP",
            "BSD-3-Clause-LBNL",
            "BSD-3-Clause-Modification",
            "BSD-3-Clause-No-Military-License",
            "BSD-3-Clause-No-Nuclear-License",
            "BSD-3-Clause-No-Nuclear-License-2014",
            "BSD-3-Clause-No-Nuclear-Warranty",
            "BSD-3-Clause-Open-MPI",
            "BSD-3-Clause-Sun",
            "BSD-4-Clause",
            "BSD-4-Clause-Shortened",
            "BSD-4-Clause-UC",
            "BSD-4.3RENO",
            "BSD-4.3TAHOE",
            "BSD-Advertising-Acknowledgement",
            "BSD-Attribution-HPND-disclaimer",
            "BSD-Inferno-Nettverk",
            "BSD-Protection",
            "BSD-Source-beginning-file",
            "BSD-Source-Code",
            "BSD-Systemics",
            "BSD-Systemics-W3Works",
            "BSL-1.0",
            "BUSL-1.1",
            "bzip2-1.0.6",
            "C-UDA-1.0",
            "CAL-1.0",
            "CAL-1.0-Combined-Work-Exception",
            "Caldera",
            "Caldera-no-preamble",
            "Catharon",
            "CATOSL-1.1",
            "CC-BY-1.0",
            "CC-BY-2.0",
            "CC-BY-2.5",
            "CC-BY-2.5-AU",
            "CC-BY-3.0",
            "CC-BY-3.0-AT",
            "CC-BY-3.0-AU",
            "CC-BY-3.0-DE",
            "CC-BY-3.0-IGO",
            "CC-BY-3.0-NL",
            "CC-BY-3.0-US",
            "CC-BY-4.0",
            "CC-BY-NC-1.0",
            "CC-BY-NC-2.0",
            "CC-BY-NC-2.5",
            "CC-BY-NC-3.0",
            "CC-BY-NC-3.0-DE",
            "CC-BY-NC-4.0",
            "CC-BY-NC-ND-1.0",
            "CC-BY-NC-ND-2.0",
            "CC-BY-NC-ND-2.5",
            "CC-BY-NC-ND-3.0",
            "CC-BY-NC-ND-3.0-DE",
            "CC-BY-NC-ND-3.0-IGO",
            "CC-BY-NC-ND-4.0",
            "CC-BY-NC-SA-1.0",
            "CC-BY-NC-SA-2.0",
            "CC-BY-NC-SA-2.0-DE",
            "CC-BY-NC-SA-2.0-FR",
            "CC-BY-NC-SA-2.0-UK",
            "CC-BY-NC-SA-2.5",
            "CC-BY-NC-SA-3.0",
            "CC-BY-NC-SA-3.0-DE",
            "CC-BY-NC-SA-3.0-IGO",
            "CC-BY-NC-SA-4.0",
            "CC-BY-ND-1.0",
            "CC-BY-ND-2.0",
            "CC-BY-ND-2.5",
            "CC-BY-ND-3.0",
            "CC-BY-ND-3.0-DE",
            "CC-BY-ND-4.0",
            "CC-BY-SA-1.0",
            "CC-BY-SA-2.0",
            "CC-BY-SA-2.0-UK",
            "CC-BY-SA-2.1-JP",
            "CC-BY-SA-2.5",
            "CC-BY-SA-3.0",
            "CC-BY-SA-3.0-AT",
            "CC-BY-SA-3.0-DE",
            "CC-BY-SA-3.0-IGO",
            "CC-BY-SA-4.0",
            "CC-PDDC",
            "CC-PDM-1.0",
            "CC-SA-1.0",
            "CC0-1.0",
            "CDDL-1.0",
            "CDDL-1.1",
            "CDL-1.0",
            "CDLA-Permissive-1.0",
            "CDLA-Permissive-2.0",
            "CDLA-Sharing-1.0",
            "CECILL-1.0",
            "CECILL-1.1",
            "CECILL-2.0",
            "CECILL-2.1",
            "CECILL-B",
            "CECILL-C",
            "CERN-OHL-1.1",
            "CERN-OHL-1.2",
            "CERN-OHL-P-2.0",
            "CERN-OHL-S-2.0",
            "CERN-OHL-W-2.0",
            "CFITSIO",
            "check-cvs",
            "checkmk",
            "ClArtistic",
            "Clips",
            "CMU-Mach",
            "CMU-Mach-nodoc",
            "CNRI-Jython",
            "CNRI-Python",
            "CNRI-Python-GPL-Compatible",
            "COIL-1.0",
            "Community-Spec-1.0",
            "Condor-1.1",
            "copyleft-next-0.3.0",
            "copyleft-next-0.3.1",
            "Cornell-Lossless-JPEG",
            "CPAL-1.0",
            "CPL-1.0",
            "CPOL-1.02",
            "Cronyx",
            "Crossword",
            "CryptoSwift",
            "CrystalStacker",
            "CUA-OPL-1.0",
            "Cube",
            "curl",
            "cve-tou",
            "D-FSL-1.0",
            "DEC-3-Clause",
            "diffmark",
            "DL-DE-BY-2.0",
            "DL-DE-ZERO-2.0",
            "DOC",
            "DocBook-DTD",
            "DocBook-Schema",
            "DocBook-Stylesheet",
            "DocBook-XML",
            "Dotseqn",
            "DRL-1.0",
            "DRL-1.1",
            "DSDP",
            "dtoa",
            "dvipdfm",
            "ECL-1.0",
            "ECL-2.0",
            "EFL-1.0",
            "EFL-2.0",
            "eGenix",
            "Elastic-2.0",
            "Entessa",
            "EPICS",
            "EPL-1.0",
            "EPL-2.0",
            "ErlPL-1.1",
            "etalab-2.0",
            "EUDatagrid",
            "EUPL-1.0",
            "EUPL-1.1",
            "EUPL-1.2",
            "Eurosym",
            "Fair",
            "FBM",
            "FDK-AAC",
            "Ferguson-Twofish",
            "Frameworx-1.0",
            "FreeBSD-DOC",
            "FreeImage",
            "FSFAP",
            "FSFAP-no-warranty-disclaimer",
            "FSFUL",
            "FSFULLR",
            "FSFULLRSD",
            "FSFULLRWD",
            "FSL-1.1-ALv2",
            "FSL-1.1-MIT",
            "FTL",
            "Furuseth",
            "fwlw",
            "Game-Programming-Gems",
            "GCR-docs",
            "GD",
            "generic-xts",
            "GFDL-1.1-invariants-only",
            "GFDL-1.1-invariants-or-later",
            "GFDL-1.1-no-invariants-only",
            "GFDL-1.1-no-invariants-or-later",
            "GFDL-1.1-only",
            "GFDL-1.1-or-later",
            "GFDL-1.2-invariants-only",
            "GFDL-1.2-invariants-or-later",
            "GFDL-1.2-no-invariants-only",
            "GFDL-1.2-no-invariants-or-later",
            "GFDL-1.2-only",
            "GFDL-1.2-or-later",
            "GFDL-1.3-invariants-only",
            "GFDL-1.3-invariants-or-later",
            "GFDL-1.3-no-invariants-only",
            "GFDL-1.3-no-invariants-or-later",
            "GFDL-1.3-only",
            "GFDL-1.3-or-later",
            "Giftware",
            "GL2PS",
            "Glide",
            "Glulxe",
            "GLWTPL",
            "gnuplot",
            "GPL-1.0-only",
            "GPL-1.0-or-later",
            "GPL-2.0-only",
            "GPL-2.0-or-later",
            "GPL-3.0-only",
            "GPL-3.0-or-later",
            "Graphics-Gems",
            "gSOAP-1.3b",
            "gtkbook",
            "Gutmann",
            "HaskellReport",
            "HDF5",
            "hdparm",
            "HIDAPI",
            "Hippocratic-2.1",
            "HP-1986",
            "HP-1989",
            "HPND",
            "HPND-DEC",
            "HPND-doc",
            "HPND-doc-sell",
            "HPND-export-US",
            "HPND-export-US-acknowledgement",
            "HPND-export-US-modify",
            "HPND-export2-US",
            "HPND-Fenneberg-Livingston",
            "HPND-INRIA-IMAG",
            "HPND-Intel",
            "HPND-Kevlin-Henney",
            "HPND-Markus-Kuhn",
            "HPND-merchantability-variant",
            "HPND-MIT-disclaimer",
            "HPND-Netrek",
            "HPND-Pbmplus",
            "HPND-sell-MIT-disclaimer-xserver",
            "HPND-sell-regexpr",
            "HPND-sell-variant",
            "HPND-sell-variant-MIT-disclaimer",
            "HPND-sell-variant-MIT-disclaimer-rev",
            "HPND-UC",
            "HPND-UC-export-US",
            "HTMLTIDY",
            "IBM-pibs",
            "ICU",
            "IEC-Code-Components-EULA",
            "IJG",
            "IJG-short",
            "ImageMagick",
            "iMatix",
            "Imlib2",
            "Info-ZIP",
            "Inner-Net-2.0",
            "InnoSetup",
            "Intel",
            "Intel-ACPI",
            "Interbase-1.0",
            "IPA",
            "IPL-1.0",
            "ISC",
            "ISC-Veillard",
            "Jam",
            "JasPer-2.0",
            "jove",
            "JPL-image",
            "JPNIC",
            "JSON",
            "Kastrup",
            "Kazlib",
            "Knuth-CTAN",
            "LAL-1.2",
            "LAL-1.3",
            "Latex2e",
            "Latex2e-translated-notice",
            "Leptonica",
            "LGPL-2.0-only",
            "LGPL-2.0-or-later",
            "LGPL-2.1-only",
            "LGPL-2.1-or-later",
            "LGPL-3.0-only",
            "LGPL-3.0-or-later",
            "LGPLLR",
            "Libpng",
            "libpng-1.6.35",
            "libpng-2.0",
            "libselinux-1.0",
            "libtiff",
            "libutil-David-Nugent",
            "LiLiQ-P-1.1",
            "LiLiQ-R-1.1",
            "LiLiQ-Rplus-1.1",
            "Linux-man-pages-1-para",
            "Linux-man-pages-copyleft",
            "Linux-man-pages-copyleft-2-para",
            "Linux-man-pages-copyleft-var",
            "Linux-OpenIB",
            "LOOP",
            "LPD-document",
            "LPL-1.0",
            "LPL-1.02",
            "LPPL-1.0",
            "LPPL-1.1",
            "LPPL-1.2",
            "LPPL-1.3a",
            "LPPL-1.3c",
            "lsof",
            "Lucida-Bitmap-Fonts",
            "LZMA-SDK-9.11-to-9.20",
            "LZMA-SDK-9.22",
            "Mackerras-3-Clause",
            "Mackerras-3-Clause-acknowledgment",
            "magaz",
            "mailprio",
            "MakeIndex",
            "man2html",
            "Martin-Birgmeier",
            "McPhee-slideshow",
            "metamail",
            "Minpack",
            "MIPS",
            "MirOS",
            "MIT",
            "MIT-0",
            "MIT-advertising",
            "MIT-Click",
            "MIT-CMU",
            "MIT-enna",
            "MIT-feh",
            "MIT-Festival",
            "MIT-Khronos-old",
            "MIT-Modern-Variant",
            "MIT-open-group",
            "MIT-testregex",
            "MIT-Wu",
            "MITNFA",
            "MMIXware",
            "Motosoto",
            "MPEG-SSG",
            "mpi-permissive",
            "mpich2",
            "MPL-1.0",
            "MPL-1.1",
            "MPL-2.0",
            "MPL-2.0-no-copyleft-exception",
            "mplus",
            "MS-LPL",
            "MS-PL",
            "MS-RL",
            "MTLL",
            "MulanPSL-1.0",
            "MulanPSL-2.0",
            "Multics",
            "Mup",
            "NAIST-2003",
            "NASA-1.3",
            "Naumen",
            "NBPL-1.0",
            "NCBI-PD",
            "NCGL-UK-2.0",
            "NCL",
            "NCSA",
            "NetCDF",
            "Newsletr",
            "NGPL",
            "ngrep",
            "NICTA-1.0",
            "NIST-PD",
            "NIST-PD-fallback",
            "NIST-Software",
            "NLOD-1.0",
            "NLOD-2.0",
            "NLPL",
            "Nokia",
            "NOSL",
            "Noweb",
            "NPL-1.0",
            "NPL-1.1",
            "NPOSL-3.0",
            "NRL",
            "NTIA-PD",
            "NTP",
            "NTP-0",
            "O-UDA-1.0",
            "OAR",
            "OCCT-PL",
            "OCLC-2.0",
            "ODbL-1.0",
            "ODC-By-1.0",
            "OFFIS",
            "OFL-1.0",
            "OFL-1.0-no-RFN",
            "OFL-1.0-RFN",
            "OFL-1.1",
            "OFL-1.1-no-RFN",
            "OFL-1.1-RFN",
            "OGC-1.0",
            "OGDL-Taiwan-1.0",
            "OGL-Canada-2.0",
            "OGL-UK-1.0",
            "OGL-UK-2.0",
            "OGL-UK-3.0",
            "OGTSL",
            "OLDAP-1.1",
            "OLDAP-1.2",
            "OLDAP-1.3",
            "OLDAP-1.4",
            "OLDAP-2.0",
            "OLDAP-2.0.1",
            "OLDAP-2.1",
            "OLDAP-2.2",
            "OLDAP-2.2.1",
            "OLDAP-2.2.2",
            "OLDAP-2.3",
            "OLDAP-2.4",
            "OLDAP-2.5",
            "OLDAP-2.6",
            "OLDAP-2.7",
            "OLDAP-2.8",
            "OLFL-1.3",
            "OML",
            "OpenPBS-2.3",
            "OpenSSL",
            "OpenSSL-standalone",
            "OpenVision",
            "OPL-1.0",
            "OPL-UK-3.0",
            "OPUBL-1.0",
            "OSET-PL-2.1",
            "OSL-1.0",
            "OSL-1.1",
            "OSL-2.0",
            "OSL-2.1",
            "OSL-3.0",
            "PADL",
            "Parity-6.0.0",
            "Parity-7.0.0",
            "PDDL-1.0",
            "PHP-3.0",
            "PHP-3.01",
            "Pixar",
            "pkgconf",
            "Plexus",
            "pnmstitch",
            "PolyForm-Noncommercial-1.0.0",
            "PolyForm-Small-Business-1.0.0",
            "PostgreSQL",
            "PPL",
            "PSF-2.0",
            "psfrag",
            "psutils",
            "Python-2.0",
            "Python-2.0.1",
            "python-ldap",
            "Qhull",
            "QPL-1.0",
            "QPL-1.0-INRIA-2004",
            "radvd",
            "Rdisc",
            "RHeCos-1.1",
            "RPL-1.1",
            "RPL-1.5",
            "RPSL-1.0",
            "RSA-MD",
            "RSCPL",
            "Ruby",
            "Ruby-pty",
            "SAX-PD",
            "SAX-PD-2.0",
            "Saxpath",
            "SCEA",
            "SchemeReport",
            "Sendmail",
            "Sendmail-8.23",
            "Sendmail-Open-Source-1.1",
            "SGI-B-1.0",
            "SGI-B-1.1",
            "SGI-B-2.0",
            "SGI-OpenGL",
            "SGP4",
            "SHL-0.5",
            "SHL-0.51",
            "SimPL-2.0",
            "SISSL",
            "SISSL-1.2",
            "SL",
            "Sleepycat",
            "SMAIL-GPL",
            "SMLNJ",
            "SMPPL",
            "SNIA",
            "snprintf",
            "SOFA",
            "softSurfer",
            "Soundex",
            "Spencer-86",
            "Spencer-94",
            "Spencer-99",
            "SPL-1.0",
            "ssh-keyscan",
            "SSH-OpenSSH",
            "SSH-short",
            "SSLeay-standalone",
            "SSPL-1.0",
            "SugarCRM-1.1.3",
            "SUL-1.0",
            "Sun-PPP",
            "Sun-PPP-2000",
            "SunPro",
            "SWL",
            "swrule",
            "Symlinks",
            "TAPR-OHL-1.0",
            "TCL",
            "TCP-wrappers",
            "TermReadKey",
            "TGPPL-1.0",
            "ThirdEye",
            "threeparttable",
            "TMate",
            "TORQUE-1.1",
            "TOSL",
            "TPDL",
            "TPL-1.0",
            "TrustedQSL",
            "TTWL",
            "TTYP0",
            "TU-Berlin-1.0",
            "TU-Berlin-2.0",
            "Ubuntu-font-1.0",
            "UCAR",
            "UCL-1.0",
            "ulem",
            "UMich-Merit",
            "Unicode-3.0",
            "Unicode-DFS-2015",
            "Unicode-DFS-2016",
            "Unicode-TOU",
            "UnixCrypt",
            "Unlicense",
            "Unlicense-libtelnet",
            "Unlicense-libwhirlpool",
            "UPL-1.0",
            "URT-RLE",
            "Vim",
            "VOSTROM",
            "VSL-1.0",
            "W3C",
            "W3C-19980720",
            "W3C-20150513",
            "w3m",
            "Watcom-1.0",
            "Widget-Workshop",
            "Wsuipa",
            "WTFPL",
            "wwl",
            "X11",
            "X11-distribute-modifications-variant",
            "X11-swapped",
            "Xdebug-1.03",
            "Xerox",
            "Xfig",
            "XFree86-1.1",
            "xinetd",
            "xkeyboard-config-Zinoviev",
            "xlock",
            "Xnet",
            "xpp",
            "XSkat",
            "xzoom",
            "YPL-1.0",
            "YPL-1.1",
            "Zed",
            "Zeeff",
            "Zend-2.0",
            "Zimbra-1.3",
            "Zimbra-1.4",
            "Zlib",
            "zlib-acknowledgement",
            "ZPL-1.1",
            "ZPL-2.0",
            "ZPL-2.1"
          ],
          "title": "LicenseId",
          "type": "string"
        },
        {
          "enum": [
            "AGPL-1.0",
            "AGPL-3.0",
            "BSD-2-Clause-FreeBSD",
            "BSD-2-Clause-NetBSD",
            "bzip2-1.0.5",
            "eCos-2.0",
            "GFDL-1.1",
            "GFDL-1.2",
            "GFDL-1.3",
            "GPL-1.0",
            "GPL-1.0+",
            "GPL-2.0",
            "GPL-2.0+",
            "GPL-2.0-with-autoconf-exception",
            "GPL-2.0-with-bison-exception",
            "GPL-2.0-with-classpath-exception",
            "GPL-2.0-with-font-exception",
            "GPL-2.0-with-GCC-exception",
            "GPL-3.0",
            "GPL-3.0+",
            "GPL-3.0-with-autoconf-exception",
            "GPL-3.0-with-GCC-exception",
            "LGPL-2.0",
            "LGPL-2.0+",
            "LGPL-2.1",
            "LGPL-2.1+",
            "LGPL-3.0",
            "LGPL-3.0+",
            "Net-SNMP",
            "Nunit",
            "StandardML-NJ",
            "wxWindows"
          ],
          "title": "DeprecatedLicenseId",
          "type": "string"
        },
        {
          "$ref": "#/$defs/FileDescr"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "A [SPDX license identifier](https://spdx.org/licenses/) or a custom license file.",
      "examples": [
        "CC0-1.0",
        "MIT",
        "BSD-2-Clause"
      ],
      "title": "License"
    },
    "git_repo": {
      "anyOf": [
        {
          "description": "A URL with the HTTP or HTTPS scheme.",
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "title": "HttpUrl",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "A URL to the Git repository where the resource is being developed.",
      "examples": [
        "https://github.com/bioimage-io/spec-bioimage-io/tree/main/example_descriptions/models/unet2d_nuclei_broad"
      ],
      "title": "Git Repo"
    },
    "icon": {
      "anyOf": [
        {
          "maxLength": 2,
          "minLength": 1,
          "type": "string"
        },
        {
          "$ref": "#/$defs/FileDescr"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "An icon for illustration, e.g. on bioimage.io",
      "title": "Icon"
    },
    "links": {
      "description": "IDs of other bioimage.io resources",
      "examples": [
        [
          "ilastik/ilastik",
          "deepimagej/deepimagej",
          "zero/notebook_u-net_3d_zerocostdl4mic"
        ]
      ],
      "items": {
        "type": "string"
      },
      "title": "Links",
      "type": "array"
    },
    "uploader": {
      "anyOf": [
        {
          "$ref": "#/$defs/Uploader"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The person who uploaded the model (e.g. to bioimage.io)"
    },
    "maintainers": {
      "description": "Maintainers of this resource.\nIf not specified, `authors` are maintainers and at least some of them has to specify their `github_user` name",
      "items": {
        "$ref": "#/$defs/Maintainer"
      },
      "title": "Maintainers",
      "type": "array"
    },
    "tags": {
      "description": "Associated tags",
      "examples": [
        [
          "unet2d",
          "pytorch",
          "nucleus",
          "segmentation",
          "dsb2018"
        ]
      ],
      "items": {
        "type": "string"
      },
      "title": "Tags",
      "type": "array"
    },
    "version": {
      "anyOf": [
        {
          "$ref": "#/$defs/Version"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The version of the resource following SemVer 2.0."
    },
    "version_comment": {
      "anyOf": [
        {
          "maxLength": 512,
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "A comment on the version of the resource.",
      "title": "Version Comment"
    },
    "format_version": {
      "const": "0.3.4",
      "description": "The **format** version of this resource specification",
      "title": "Format Version",
      "type": "string"
    },
    "badges": {
      "description": "badges associated with this resource",
      "items": {
        "$ref": "#/$defs/BadgeDescr"
      },
      "title": "Badges",
      "type": "array"
    },
    "config": {
      "$ref": "#/$defs/Config",
      "description": "A field for custom configuration that can contain any keys not present in the RDF spec.\nThis means you should not store, for example, a GitHub repo URL in `config` since there is a `git_repo` field.\nKeys in `config` may be very specific to a tool or consumer software. To avoid conflicting definitions,\nit is recommended to wrap added configuration into a sub-field named with the specific domain or tool name,\nfor example:\n```yaml\nconfig:\n    giraffe_neckometer:  # here is the domain name\n        length: 3837283\n        address:\n            home: zoo\n    imagej:              # config specific to ImageJ\n        macro_dir: path/to/macro/file\n```\nIf possible, please use [`snake_case`](https://en.wikipedia.org/wiki/Snake_case) for keys in `config`.\nYou may want to list linked files additionally under `attachments` to include them when packaging a resource.\n(Packaging a resource means downloading/copying important linked files and creating a ZIP archive that contains\nan altered rdf.yaml file with local references to the downloaded files.)"
    },
    "type": {
      "const": "dataset",
      "title": "Type",
      "type": "string"
    },
    "id": {
      "anyOf": [
        {
          "minLength": 1,
          "title": "DatasetId",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "bioimage.io-wide unique resource identifier\nassigned by bioimage.io; version **un**specific.",
      "title": "Id"
    },
    "parent": {
      "anyOf": [
        {
          "minLength": 1,
          "title": "DatasetId",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The description from which this one is derived",
      "title": "Parent"
    },
    "source": {
      "anyOf": [
        {
          "description": "A URL with the HTTP or HTTPS scheme.",
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "title": "HttpUrl",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "\"URL to the source of the dataset.",
      "title": "Source"
    }
  },
  "required": [
    "name",
    "format_version",
    "type"
  ],
  "title": "dataset 0.3.4",
  "type": "object"
}

Fields:

• _validation_summary (ValidationSummary | None)
• name (Annotated[Annotated[str, RestrictCharacters(string.ascii_letters + string.digits + '_+- ()')], MinLen(5), MaxLen(128), warn(MaxLen(64), 'Name longer than 64 characters.', INFO)])
• description (FAIR[Annotated[str, MaxLen(1024), warn(MaxLen(512), 'Description longer than 512 characters.')]])
• covers (list[_FileDescr_cover])
• documentation (FAIR[FileDescr_documentation | None])
• id_emoji (Annotated[str, Len(min_length=1, max_length=2), Field(examples=['🦈', '🦥'])] | None)
• authors (FAIR[list[Author]])
• attachments (list[FileDescr_package])
• cite (FAIR[list[CiteEntry]])
• license (FAIR[Annotated[LicenseId | DeprecatedLicenseId | None | FileDescr_package, Field(union_mode='left_to_right', examples=['CC0-1.0', 'MIT', 'BSD-2-Clause'])]])
• git_repo (Annotated[HttpUrl | None, Field(examples=['https://github.com/bioimage-io/spec-bioimage-io/tree/main/example_descriptions/models/unet2d_nuclei_broad'])])
• icon (Annotated[str, Len(min_length=1, max_length=2)] | FileDescr_package | None)
• links (Annotated[list[str], Field(examples=[('ilastik/ilastik', 'deepimagej/deepimagej', 'zero/notebook_u-net_3d_zerocostdl4mic')])])
• uploader (Uploader | None)
• maintainers (list[Maintainer])
• tags (FAIR[Annotated[list[str], Field(examples=[('unet2d', 'pytorch', 'nucleus', 'segmentation', 'dsb2018')])]])
• version (Version | None)
• version_comment (Annotated[str, MaxLen(512)] | None)
• format_version (Literal['0.3.4'])
• badges (list[BadgeDescr])
• config (Config)
• type (Literal['dataset'])
• id (DatasetId | None)
• parent (DatasetId | None)
• source (FAIR[HttpUrl | None])

Validators:

• _convert_from_older_format
• _convert

attachments pydantic-field

attachments: list[FileDescr_package]

file attachments

authors pydantic-field

authors: FAIR[list[Author]]

The authors are the creators of this resource description and the primary points of contact.

badges pydantic-field

badges: list[BadgeDescr]

badges associated with this resource

cite pydantic-field

cite: FAIR[list[CiteEntry]]

citations

config pydantic-field

config: Config

A field for custom configuration that can contain any keys not present in the RDF spec. This means you should not store, for example, a GitHub repo URL in config since there is a git_repo field. Keys in config may be very specific to a tool or consumer software. To avoid conflicting definitions, it is recommended to wrap added configuration into a sub-field named with the specific domain or tool name, for example:

config:
    giraffe_neckometer:  # here is the domain name
        length: 3837283
        address:
            home: zoo
    imagej:              # config specific to ImageJ
        macro_dir: path/to/macro/file

If possible, please use snake_case for keys in config. You may want to list linked files additionally under attachments to include them when packaging a resource. (Packaging a resource means downloading/copying important linked files and creating a ZIP archive that contains an altered rdf.yaml file with local references to the downloaded files.)

covers pydantic-field

covers: list[_FileDescr_cover]

Cover images.

description pydantic-field

description: FAIR[
    Annotated[
        str,
        MaxLen(1024),
        warn(
            MaxLen(512),
            "Description longer than 512 characters.",
        ),
    ]
] = ""

A string containing a brief description.

documentation pydantic-field

documentation: FAIR[FileDescr_documentation | None] = None

Additional model documentation. The recommended documentation source file name is README.md. An .md suffix is mandatory.

file_name property

file_name: FileName | None

File name of the bioimageio.yaml file the description was loaded from.

format_version pydantic-field

format_version: Literal['0.3.4'] = '0.3.4'

git_repo pydantic-field

git_repo: Annotated[
    HttpUrl | None,
    Field(
        examples=[
            "https://github.com/bioimage-io/spec-bioimage-io/tree/main/example_descriptions/models/unet2d_nuclei_broad"
        ]
    ),
] = None

A URL to the Git repository where the resource is being developed.

icon pydantic-field

icon: (
    Annotated[str, Len(min_length=1, max_length=2)]
    | FileDescr_package
    | None
) = None

An icon for illustration, e.g. on bioimage.io

id pydantic-field

id: DatasetId | None = None

bioimage.io-wide unique resource identifier assigned by bioimage.io; version unspecific.

id_emoji pydantic-field

id_emoji: (
    Annotated[
        str,
        Len(min_length=1, max_length=2),
        Field(examples=["🦈", "🦥"]),
    ]
    | None
) = None

UTF-8 emoji for display alongside the id.

implemented_format_version class-attribute

implemented_format_version: Literal['0.3.4'] = '0.3.4'

implemented_format_version_tuple class-attribute

implemented_format_version_tuple: tuple[int, int, int]

implemented_type class-attribute

implemented_type: Literal['dataset'] = 'dataset'

license pydantic-field

license: FAIR[
    Annotated[
        LicenseId
        | DeprecatedLicenseId
        | None
        | FileDescr_package,
        Field(
            union_mode="left_to_right",
            examples=["CC0-1.0", "MIT", "BSD-2-Clause"],
        ),
    ]
] = None

A SPDX license identifier or a custom license file.

links pydantic-field

links: Annotated[
    list[str],
    Field(
        examples=[
            (
                "ilastik/ilastik",
                "deepimagej/deepimagej",
                "zero/notebook_u-net_3d_zerocostdl4mic",
            )
        ]
    ),
]

IDs of other bioimage.io resources

maintainers pydantic-field

maintainers: list[Maintainer]

Maintainers of this resource. If not specified, authors are maintainers and at least some of them has to specify their github_user name

name pydantic-field

name: Annotated[
    Annotated[
        str,
        RestrictCharacters(
            string.ascii_letters + string.digits + "_+- ()"
        ),
    ],
    MinLen(5),
    MaxLen(128),
    warn(
        MaxLen(64), "Name longer than 64 characters.", INFO
    ),
]

A human-friendly name of the resource description. May only contains letters, digits, underscore, minus, parentheses and spaces.

parent pydantic-field

parent: DatasetId | None = None

The description from which this one is derived

root property

root: RootHttpUrl | DirectoryPath | ZipFile

The URL/Path prefix to resolve any relative paths with.

source pydantic-field

source: FAIR[HttpUrl | None] = None

"URL to the source of the dataset.

tags pydantic-field

tags: FAIR[
    Annotated[
        list[str],
        Field(
            examples=[
                (
                    "unet2d",
                    "pytorch",
                    "nucleus",
                    "segmentation",
                    "dsb2018",
                )
            ]
        ),
    ]
]

Associated tags

type pydantic-field

type: Literal['dataset'] = 'dataset'

uploader pydantic-field

uploader: Uploader | None = None

The person who uploaded the model (e.g. to bioimage.io)

validation_summary property

validation_summary: ValidationSummary

version pydantic-field

version: Version | None = None

The version of the resource following SemVer 2.0.

version_comment pydantic-field

version_comment: Annotated[str, MaxLen(512)] | None = None

A comment on the version of the resource.

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any)

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any):
    super().__pydantic_init_subclass__(**kwargs)
    # set classvar implemented_format_version_tuple
    if "format_version" in cls.model_fields:
        if "." not in cls.implemented_format_version:
            cls.implemented_format_version_tuple = (0, 0, 0)
        else:
            fv_tuple = get_format_version_tuple(cls.implemented_format_version)
            assert fv_tuple is not None, (
                f"failed to cast '{cls.implemented_format_version}' to tuple"
            )
            cls.implemented_format_version_tuple = fv_tuple

convert_from_old_format_wo_validation classmethod

convert_from_old_format_wo_validation(
    data: BioimageioYamlContent,
) -> None

Convert metadata following an older format version to this classes' format without validating the result.

Source code in src/bioimageio/spec/generic/v0_3.py

@classmethod
def convert_from_old_format_wo_validation(cls, data: BioimageioYamlContent) -> None:
    """Convert metadata following an older format version to this classes' format
    without validating the result.
    """
    convert_from_older_format(data)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get_package_content

get_package_content(
    local_files_only: bool = False,
) -> dict[FileName, FileDescr | BioimageioYamlContent]

Returns package content without creating the package.

Source code in src/bioimageio/spec/_internal/common_nodes.py

def get_package_content(
    self, local_files_only: bool = False
) -> dict[FileName, FileDescr | BioimageioYamlContent]:
    """Returns package content without creating the package."""
    content: dict[FileName, FileDescr] = {}
    with PackagingContext(
        bioimageio_yaml_file_name=BIOIMAGEIO_YAML,
        file_sources=content,
        local_files_only=local_files_only,
    ):
        rdf_content: BioimageioYamlContent = self.model_dump(
            mode="json", exclude_unset=True
        )

    _ = rdf_content.pop("rdf_source", None)

    return {**content, BIOIMAGEIO_YAML: rdf_content}

load classmethod

load(
    data: IncompleteDescrView,
    context: ValidationContext | None = None,
) -> Self | InvalidDescr

factory method to create a resource description object

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def load(
    cls,
    data: IncompleteDescrView,
    context: ValidationContext | None = None,
) -> Self | InvalidDescr:
    """factory method to create a resource description object"""

    context = context or get_validation_context()
    if context.perform_io_checks:
        file_descrs = extract_file_descrs(data)
        populate_cache(file_descrs)  # TODO: add progress bar

    with context.replace(log_warnings=context.warning_level <= INFO):
        rd, errors, val_warnings = cls._load_impl(deepcopy_incomplete_descr(data))

    if context.warning_level > INFO:
        all_warnings_context = context.replace(
            warning_level=INFO, log_warnings=False, raise_errors=False
        )
        # raise all validation warnings by reloading
        with all_warnings_context:
            _, _, val_warnings = cls._load_impl(deepcopy_incomplete_descr(data))

    format_status = "failed" if errors else "passed"
    rd.validation_summary.add_detail(
        ValidationDetail(
            errors=errors,
            name=(
                "bioimageio.spec format validation"
                f" {rd.type} {cls.implemented_format_version}"
            ),
            status=format_status,
            warnings=val_warnings,
        ),
        update_status=False,  # this special validation detail needs manual format updating below
    )
    assert format_status != "failed" or isinstance(rd, InvalidDescr)

    return rd

load_from_kwargs classmethod

load_from_kwargs(
    context: ValidationContext | None = None,
    *args: P.args,
    **kwargs: P.kwargs,
) -> T | InvalidDescr

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def load_from_kwargs(
    cls: Callable[P, T],
    context: ValidationContext | None = None,
    *args: P.args,
    **kwargs: P.kwargs,
) -> T | InvalidDescr:
    sig = signature(cls)
    bound = sig.bind_partial(*args, **kwargs)
    return cls.load(dict(bound.arguments), context=context)  # pyright: ignore[reportFunctionMemberAccess]

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

package

package(
    dest: ZipFile | IO[bytes] | Path | str | None = None,
    /,
    local_files_only: bool = False,
) -> ZipFile

package the described resource as a zip archive

Parameters:

Name	Type	Description	Default
dest	ZipFile | IO[bytes] | Path | str | None	(path/bytes stream of) destination zipfile	None

Source code in src/bioimageio/spec/_internal/common_nodes.py

def package(
    self,
    dest: ZipFile | IO[bytes] | Path | str | None = None,
    /,
    local_files_only: bool = False,
) -> ZipFile:
    """package the described resource as a zip archive

    Args:
        dest: (path/bytes stream of) destination zipfile
    """
    if dest is None:
        dest = BytesIO()

    if isinstance(dest, ZipFile):
        zip = dest
        if "r" in zip.mode:
            raise ValueError(
                f"zip file {dest} opened in '{zip.mode}' mode,"
                + " but write access is needed for packaging."
            )
    else:
        zip = ZipFile(dest, mode="w")

    if zip.filename is None:
        zip.filename = (
            str(getattr(self, "id", getattr(self, "name", "bioimageio"))) + ".zip"
        )

    content = self.get_package_content(local_files_only=local_files_only)
    write_content_to_zip(content, zip)
    return zip

warn_about_tag_categories classmethod

warn_about_tag_categories(
    value: list[str], info: ValidationInfo
) -> list[str]

Source code in src/bioimageio/spec/generic/v0_3.py

@as_warning
@field_validator("tags")
@classmethod
def warn_about_tag_categories(
    cls, value: list[str], info: ValidationInfo
) -> list[str]:
    categories = TAG_CATEGORIES.get(info.data["type"], {})
    missing_categories: list[dict[str, Sequence[str]]] = []
    for cat, entries in categories.items():
        if not any(e in value for e in entries):
            missing_categories.append({cat: entries})

    if missing_categories:
        raise ValueError(
            f"Missing tags from bioimage.io categories: {missing_categories}"
        )

    return value

DatasetId

Bases: ResourceId

              flowchart TD
              bioimageio.spec.model.v0_5.DatasetId[DatasetId]
              bioimageio.spec.generic.v0_3.ResourceId[ResourceId]
              bioimageio.spec._internal.validated_string.ValidatedString[ValidatedString]

                              bioimageio.spec.generic.v0_3.ResourceId --> bioimageio.spec.model.v0_5.DatasetId
                                bioimageio.spec._internal.validated_string.ValidatedString --> bioimageio.spec.generic.v0_3.ResourceId
                



              click bioimageio.spec.model.v0_5.DatasetId href "" "bioimageio.spec.model.v0_5.DatasetId"
              click bioimageio.spec.generic.v0_3.ResourceId href "" "bioimageio.spec.generic.v0_3.ResourceId"
              click bioimageio.spec._internal.validated_string.ValidatedString href "" "bioimageio.spec._internal.validated_string.ValidatedString"
            

Methods:

Name	Description
__get_pydantic_core_schema__
__get_pydantic_json_schema__
__new__

Attributes:

Name	Type	Description
root_model	type[RootModel[Any]]	the pydantic root model to validate the string

root_model class-attribute

root_model: type[RootModel[Any]] = RootModel[
    Annotated[
        NotEmpty[str],
        RestrictCharacters(
            string.ascii_lowercase + string.digits + "_-/."
        ),
        annotated_types.Predicate(
            lambda s: (
                not (
                    s.startswith("/") or s.endswith("/")
                )
            )
        ),
    ]
]

the pydantic root model to validate the string

__get_pydantic_core_schema__ classmethod

__get_pydantic_core_schema__(
    source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_core_schema__(
    cls, source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema:
    return no_info_after_validator_function(cls, handler(str))

__get_pydantic_json_schema__ classmethod

__get_pydantic_json_schema__(
    core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_json_schema__(
    cls, core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue:
    json_schema = cls.root_model.model_json_schema(mode=handler.mode)
    json_schema["title"] = cls.__name__.strip("_")
    if cls.__doc__:
        json_schema["description"] = cls.__doc__

    return json_schema

__new__

__new__(object: object)

Source code in src/bioimageio/spec/_internal/validated_string.py

def __new__(cls, object: object):
    _validated = cls.root_model.model_validate(object).root
    self = super().__new__(cls, _validated)
    self._validated = _validated
    return self._after_validator()

Datetime

Bases: RootModel[Annotated[datetime, BeforeValidator(_validate_datetime), PrettyPlainSerializer(_serialize_datetime_json, when_used='json-unless-none')]]

              flowchart TD
              bioimageio.spec.model.v0_5.Datetime[Datetime]

              

              click bioimageio.spec.model.v0_5.Datetime href "" "bioimageio.spec.model.v0_5.Datetime"
            

Timestamp in ISO 8601 format with a few restrictions listed here.

Methods:

Name	Description
now

now classmethod

now()

Source code in src/bioimageio/spec/_internal/types.py

@classmethod
def now(cls):
    return cls(datetime.now(UTC))

DeprecatedLicenseId

Bases: ValidatedString

              flowchart TD
              bioimageio.spec.model.v0_5.DeprecatedLicenseId[DeprecatedLicenseId]
              bioimageio.spec._internal.validated_string.ValidatedString[ValidatedString]

                              bioimageio.spec._internal.validated_string.ValidatedString --> bioimageio.spec.model.v0_5.DeprecatedLicenseId
                


              click bioimageio.spec.model.v0_5.DeprecatedLicenseId href "" "bioimageio.spec.model.v0_5.DeprecatedLicenseId"
              click bioimageio.spec._internal.validated_string.ValidatedString href "" "bioimageio.spec._internal.validated_string.ValidatedString"
            

Methods:

Name	Description
__get_pydantic_core_schema__
__get_pydantic_json_schema__
__new__

Attributes:

Name	Type	Description
root_model	type[RootModel[Any]]	the pydantic root model to validate the string

root_model class-attribute

root_model: type[RootModel[Any]] = RootModel[
    DeprecatedLicenseIdLiteral
]

the pydantic root model to validate the string

__get_pydantic_core_schema__ classmethod

__get_pydantic_core_schema__(
    source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_core_schema__(
    cls, source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema:
    return no_info_after_validator_function(cls, handler(str))

__get_pydantic_json_schema__ classmethod

__get_pydantic_json_schema__(
    core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_json_schema__(
    cls, core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue:
    json_schema = cls.root_model.model_json_schema(mode=handler.mode)
    json_schema["title"] = cls.__name__.strip("_")
    if cls.__doc__:
        json_schema["description"] = cls.__doc__

    return json_schema

__new__

__new__(object: object)

Source code in src/bioimageio/spec/_internal/validated_string.py

def __new__(cls, object: object):
    _validated = cls.root_model.model_validate(object).root
    self = super().__new__(cls, _validated)
    self._validated = _validated
    return self._after_validator()

DiscreteReproducibilityTolerance pydantic-model

Bases: _ReproducibilityToleranceBase[Literal[0], Literal[0]]

Show JSON schema:

{
  "additionalProperties": true,
  "properties": {
    "mismatched_elements_per_million": {
      "default": 1000,
      "description": "Maximum number of mismatched elements/pixels per million to tolerate.\n\nNote: Increased default mismatched elements per million compared to\n[`ReproducibilityTolerance`][ReproducibilityTolerance]/[`ContinuousReproducibilityTolerance`][ContinuousReproducibilityTolerance],\nsince a single pixel mismatch can be more significant than for continuous outputs with numeric tolerance.",
      "maximum": 30000,
      "minimum": 0,
      "title": "Mismatched Elements Per Million",
      "type": "integer"
    },
    "weights_formats": {
      "default": [],
      "description": "Limits the weights formats these details apply to.",
      "items": {
        "enum": [
          "keras_hdf5",
          "keras_v3",
          "onnx",
          "pytorch_state_dict",
          "tensorflow_js",
          "tensorflow_saved_model_bundle",
          "torchscript"
        ],
        "type": "string"
      },
      "title": "Weights Formats",
      "type": "array"
    },
    "relative_tolerance": {
      "const": 0,
      "default": 0,
      "description": "Relative tolerance does not apply to discrete outputs",
      "title": "Relative Tolerance",
      "type": "integer"
    },
    "absolute_tolerance": {
      "const": 0,
      "default": 0,
      "description": "Absolute tolerance does not apply to discrete outputs",
      "title": "Absolute Tolerance",
      "type": "integer"
    },
    "output_ids": {
      "const": "any_discrete_output",
      "default": "any_discrete_output",
      "description": "Limits the output tensor IDs to all outputs of integer or boolean data type.",
      "title": "Output Ids",
      "type": "string"
    }
  },
  "title": "model.v0_5.DiscreteReproducibilityTolerance",
  "type": "object"
}

Fields:

• weights_formats (Sequence[WeightsFormat])
• output_ids (Literal['any_discrete_output'])
• relative_tolerance (Literal[0])
• absolute_tolerance (Literal[0])
• mismatched_elements_per_million (MismatchedElementsPerMillion)

absolute_tolerance pydantic-field

absolute_tolerance: Literal[0] = 0

Absolute tolerance does not apply to discrete outputs

mismatched_elements_per_million pydantic-field

mismatched_elements_per_million: MismatchedElementsPerMillion = 1000

Maximum number of mismatched elements/pixels per million to tolerate.

Note: Increased default mismatched elements per million compared to ReproducibilityTolerance/ContinuousReproducibilityTolerance, since a single pixel mismatch can be more significant than for continuous outputs with numeric tolerance.

output_ids pydantic-field

output_ids: Literal["any_discrete_output"] = (
    "any_discrete_output"
)

Limits the output tensor IDs to all outputs of integer or boolean data type.

relative_tolerance pydantic-field

relative_tolerance: Literal[0] = 0

Relative tolerance does not apply to discrete outputs

weights_formats pydantic-field

weights_formats: Sequence[WeightsFormat] = ()

Limits the weights formats these details apply to.

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

serialize

serialize(
    handler: SerializerFunctionWrapHandler,
) -> dict[str, Any]

explicitly serialize tolerance values

(even if they are default values and exclude_defaults=True).

Source code in src/bioimageio/spec/model/v0_5.py

@model_serializer(mode="wrap")
def serialize(
    self,
    handler: SerializerFunctionWrapHandler,
) -> dict[str, Any]:
    """explicitly serialize tolerance values

    (even if they are default values and `exclude_defaults=True`).
    """
    data = handler(self)
    data["relative_tolerance"] = self.relative_tolerance
    data["absolute_tolerance"] = self.absolute_tolerance
    data["mismatched_elements_per_million"] = self.mismatched_elements_per_million
    return data

Doi

Bases: ValidatedString

              flowchart TD
              bioimageio.spec.model.v0_5.Doi[Doi]
              bioimageio.spec._internal.validated_string.ValidatedString[ValidatedString]

                              bioimageio.spec._internal.validated_string.ValidatedString --> bioimageio.spec.model.v0_5.Doi
                


              click bioimageio.spec.model.v0_5.Doi href "" "bioimageio.spec.model.v0_5.Doi"
              click bioimageio.spec._internal.validated_string.ValidatedString href "" "bioimageio.spec._internal.validated_string.ValidatedString"
            

A digital object identifier, see https://www.doi.org/

Methods:

Name	Description
__get_pydantic_core_schema__
__get_pydantic_json_schema__
__new__

Attributes:

Name	Type	Description
root_model	type[RootModel[Any]]	the pydantic root model to validate the string

root_model class-attribute

root_model: type[RootModel[Any]] = RootModel[
    Annotated[str, StringConstraints(pattern=DOI_REGEX)]
]

the pydantic root model to validate the string

__get_pydantic_core_schema__ classmethod

__get_pydantic_core_schema__(
    source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_core_schema__(
    cls, source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema:
    return no_info_after_validator_function(cls, handler(str))

__get_pydantic_json_schema__ classmethod

__get_pydantic_json_schema__(
    core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_json_schema__(
    cls, core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue:
    json_schema = cls.root_model.model_json_schema(mode=handler.mode)
    json_schema["title"] = cls.__name__.strip("_")
    if cls.__doc__:
        json_schema["description"] = cls.__doc__

    return json_schema

__new__

__new__(object: object)

Source code in src/bioimageio/spec/_internal/validated_string.py

def __new__(cls, object: object):
    _validated = cls.root_model.model_validate(object).root
    self = super().__new__(cls, _validated)
    self._validated = _validated
    return self._after_validator()

EdgePadding pydantic-model

Bases: Node

Show JSON schema:

{
  "additionalProperties": false,
  "properties": {
    "mode": {
      "const": "edge",
      "default": "edge",
      "title": "Mode",
      "type": "string"
    }
  },
  "title": "model.v0_5.EdgePadding",
  "type": "object"
}

Fields:

• mode (Literal['edge'])

mode pydantic-field

mode: Literal['edge'] = 'edge'

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

EnsureDtypeDescr pydantic-model

Bases: NodeWithExplicitlySetFields

Cast the tensor data type to EnsureDtypeKwargs.dtype (if not matching).

This can for example be used to ensure the inner neural network model gets a different input tensor data type than the fully described bioimage.io model does.

Examples:

The described bioimage.io model (incl. preprocessing) accepts any float32-compatible tensor, normalizes it with percentiles and clipping and then casts it to uint8, which is what the neural network in this example expects. - in YAML

inputs:
- data:
    type: float32  # described bioimage.io model is compatible with any float32 input tensor
  preprocessing:
  - id: scale_range
      kwargs:
      axes: ['y', 'x']
      max_percentile: 99.8
      min_percentile: 5.0
  - id: clip
      kwargs:
      min: 0.0
      max: 1.0
  - id: ensure_dtype  # the neural network of the model requires uint8
      kwargs:
      dtype: uint8

- in Python: >>> preprocessing = [ ... ScaleRangeDescr( ... kwargs=ScaleRangeKwargs( ... axes= (AxisId('y'), AxisId('x')), ... max_percentile= 99.8, ... min_percentile= 5.0, ... ) ... ), ... ClipDescr(kwargs=ClipKwargs(min=0.0, max=1.0)), ... EnsureDtypeDescr(kwargs=EnsureDtypeKwargs(dtype="uint8")), ... ]

Show JSON schema:

{
  "$defs": {
    "EnsureDtypeKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [EnsureDtypeDescr][]",
      "properties": {
        "dtype": {
          "enum": [
            "float32",
            "float64",
            "uint8",
            "int8",
            "uint16",
            "int16",
            "uint32",
            "int32",
            "uint64",
            "int64",
            "bool"
          ],
          "title": "Dtype",
          "type": "string"
        }
      },
      "required": [
        "dtype"
      ],
      "title": "model.v0_5.EnsureDtypeKwargs",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "description": "Cast the tensor data type to `EnsureDtypeKwargs.dtype` (if not matching).\n\nThis can for example be used to ensure the inner neural network model gets a\ndifferent input tensor data type than the fully described bioimage.io model does.\n\nExamples:\n    The described bioimage.io model (incl. preprocessing) accepts any\n    float32-compatible tensor, normalizes it with percentiles and clipping and then\n    casts it to uint8, which is what the neural network in this example expects.\n    - in YAML\n        ```yaml\n        inputs:\n        - data:\n            type: float32  # described bioimage.io model is compatible with any float32 input tensor\n          preprocessing:\n          - id: scale_range\n              kwargs:\n              axes: ['y', 'x']\n              max_percentile: 99.8\n              min_percentile: 5.0\n          - id: clip\n              kwargs:\n              min: 0.0\n              max: 1.0\n          - id: ensure_dtype  # the neural network of the model requires uint8\n              kwargs:\n              dtype: uint8\n        ```\n    - in Python:\n        >>> preprocessing = [\n        ...     ScaleRangeDescr(\n        ...         kwargs=ScaleRangeKwargs(\n        ...           axes= (AxisId('y'), AxisId('x')),\n        ...           max_percentile= 99.8,\n        ...           min_percentile= 5.0,\n        ...         )\n        ...     ),\n        ...     ClipDescr(kwargs=ClipKwargs(min=0.0, max=1.0)),\n        ...     EnsureDtypeDescr(kwargs=EnsureDtypeKwargs(dtype=\"uint8\")),\n        ... ]",
  "properties": {
    "id": {
      "const": "ensure_dtype",
      "title": "Id",
      "type": "string"
    },
    "kwargs": {
      "$ref": "#/$defs/EnsureDtypeKwargs"
    }
  },
  "required": [
    "id",
    "kwargs"
  ],
  "title": "model.v0_5.EnsureDtypeDescr",
  "type": "object"
}

Fields:

• id (Literal['ensure_dtype'])
• kwargs (EnsureDtypeKwargs)

id pydantic-field

id: Literal['ensure_dtype'] = 'ensure_dtype'

implemented_id class-attribute

implemented_id: Literal['ensure_dtype'] = 'ensure_dtype'

kwargs pydantic-field

kwargs: EnsureDtypeKwargs

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

EnsureDtypeKwargs pydantic-model

Bases: KwargsNode

key word arguments for EnsureDtypeDescr

Show JSON schema:

{
  "additionalProperties": false,
  "description": "key word arguments for [EnsureDtypeDescr][]",
  "properties": {
    "dtype": {
      "enum": [
        "float32",
        "float64",
        "uint8",
        "int8",
        "uint16",
        "int16",
        "uint32",
        "int32",
        "uint64",
        "int64",
        "bool"
      ],
      "title": "Dtype",
      "type": "string"
    }
  },
  "required": [
    "dtype"
  ],
  "title": "model.v0_5.EnsureDtypeKwargs",
  "type": "object"
}

Fields:

• dtype (Literal['float32', 'float64', 'uint8', 'int8', 'uint16', 'int16', 'uint32', 'int32', 'uint64', 'int64', 'bool'])

dtype pydantic-field

dtype: Literal[
    "float32",
    "float64",
    "uint8",
    "int8",
    "uint16",
    "int16",
    "uint32",
    "int32",
    "uint64",
    "int64",
    "bool",
]

__contains__

__contains__(item: str) -> bool

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __contains__(self, item: str) -> bool:
    return item in self.__class__.model_fields

__getitem__

__getitem__(item: str) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __getitem__(self, item: str) -> Any:
    if item in self.__class__.model_fields:
        return getattr(self, item)
    else:
        raise KeyError(item)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get

get(item: str, default: Any = None) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def get(self, item: str, default: Any = None) -> Any:
    return self[item] if item in self else default  # ruff: ignore[SIM401]

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

EnvironmentalImpact pydantic-model

Bases: Node

Environmental considerations for model training and deployment.

Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019).

Show JSON schema:

{
  "additionalProperties": true,
  "description": "Environmental considerations for model training and deployment.\n\nCarbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).",
  "properties": {
    "hardware_type": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "GPU/CPU specifications",
      "title": "Hardware Type"
    },
    "hours_used": {
      "anyOf": [
        {
          "type": "number"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Total compute hours",
      "title": "Hours Used"
    },
    "cloud_provider": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "If applicable",
      "title": "Cloud Provider"
    },
    "compute_region": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Geographic location",
      "title": "Compute Region"
    },
    "co2_emitted": {
      "anyOf": [
        {
          "type": "number"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "kg CO2 equivalent\n\nCarbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).",
      "title": "Co2 Emitted"
    }
  },
  "title": "model.v0_5.EnvironmentalImpact",
  "type": "object"
}

Fields:

• hardware_type (str | None)
• hours_used (float | None)
• cloud_provider (str | None)
• compute_region (str | None)
• co2_emitted (float | None)

cloud_provider pydantic-field

cloud_provider: str | None = None

If applicable

co2_emitted pydantic-field

co2_emitted: float | None = None

kg CO2 equivalent

Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019).

compute_region pydantic-field

compute_region: str | None = None

Geographic location

hardware_type pydantic-field

hardware_type: str | None = None

GPU/CPU specifications

hours_used pydantic-field

hours_used: float | None = None

Total compute hours

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

format_md

format_md()

Filled Markdown template section following Hugging Face Model Card Template.

Source code in src/bioimageio/spec/model/v0_5.py

def format_md(self):
    """Filled Markdown template section following [Hugging Face Model Card Template](https://huggingface.co/docs/hub/en/model-card-annotated)."""
    if self == self.__class__():
        return ""

    ret = "# Environmental Impact\n\n"
    if self.hardware_type is not None:
        ret += f"- **Hardware Type:** {self.hardware_type}\n"
    if self.hours_used is not None:
        ret += f"- **Hours used:** {self.hours_used}\n"
    if self.cloud_provider is not None:
        ret += f"- **Cloud Provider:** {self.cloud_provider}\n"
    if self.compute_region is not None:
        ret += f"- **Compute Region:** {self.compute_region}\n"
    if self.co2_emitted is not None:
        ret += f"- **Carbon Emitted:** {self.co2_emitted} kg CO2e\n"

    return ret + "\n"

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

Evaluation pydantic-model

Bases: Node

Show JSON schema:

{
  "additionalProperties": true,
  "properties": {
    "model_id": {
      "anyOf": [
        {
          "minLength": 1,
          "title": "ModelId",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Model being evaluated.",
      "title": "Model Id"
    },
    "dataset_id": {
      "description": "Dataset used for evaluation.",
      "minLength": 1,
      "title": "DatasetId",
      "type": "string"
    },
    "dataset_source": {
      "description": "Source of the dataset.",
      "format": "uri",
      "maxLength": 2083,
      "minLength": 1,
      "title": "HttpUrl",
      "type": "string"
    },
    "dataset_role": {
      "description": "Role of the dataset used for evaluation.\n\n- `train`: dataset was (part of) the training data\n- `validation`: dataset was (part of) the validation data used during training, e.g. used for model selection or hyperparameter tuning\n- `test`: dataset was (part of) the designated test data; not used during training or validation, but acquired from the same source/distribution as training data\n- `independent`: dataset is entirely independent test data; not used during training or validation, and acquired from a different source/distribution than training data\n- `unknown`: role of the dataset is unknown; choose this if you are not certain if (a subset) of the data was seen by the model during training.",
      "enum": [
        "train",
        "validation",
        "test",
        "independent",
        "unknown"
      ],
      "title": "Dataset Role",
      "type": "string"
    },
    "sample_count": {
      "description": "Number of evaluated samples.",
      "title": "Sample Count",
      "type": "integer"
    },
    "evaluation_factors": {
      "description": "(Abbreviations of) each evaluation factor.\n\nEvaluation factors are criteria along which model performance is evaluated, e.g. different image conditions\nlike 'low SNR', 'high cell density', or different biological conditions like 'cell type A', 'cell type B'.\nAn 'overall' factor may be included to summarize performance across all conditions.",
      "items": {
        "maxLength": 16,
        "type": "string"
      },
      "title": "Evaluation Factors",
      "type": "array"
    },
    "evaluation_factors_long": {
      "description": "Descriptions (long form) of each evaluation factor.",
      "items": {
        "type": "string"
      },
      "title": "Evaluation Factors Long",
      "type": "array"
    },
    "metrics": {
      "description": "(Abbreviations of) metrics used for evaluation.",
      "items": {
        "maxLength": 16,
        "type": "string"
      },
      "title": "Metrics",
      "type": "array"
    },
    "metrics_long": {
      "description": "Description of each metric used.",
      "items": {
        "type": "string"
      },
      "title": "Metrics Long",
      "type": "array"
    },
    "results": {
      "description": "Results for each metric (rows; outer list) and each evaluation factor (columns; inner list).",
      "items": {
        "items": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "number"
            },
            {
              "type": "integer"
            }
          ]
        },
        "type": "array"
      },
      "title": "Results",
      "type": "array"
    },
    "results_summary": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Interpretation of results for general audience.\n\nConsider:\n    - Overall model performance\n    - Comparison to existing methods\n    - Limitations and areas for improvement",
      "title": "Results Summary"
    }
  },
  "required": [
    "dataset_id",
    "dataset_source",
    "dataset_role",
    "sample_count",
    "evaluation_factors",
    "evaluation_factors_long",
    "metrics",
    "metrics_long",
    "results"
  ],
  "title": "model.v0_5.Evaluation",
  "type": "object"
}

Fields:

• model_id (ModelId | None)
• dataset_id (DatasetId)
• dataset_source (HttpUrl)
• dataset_role (Literal['train', 'validation', 'test', 'independent', 'unknown'])
• sample_count (int)
• evaluation_factors (list[Annotated[str, MaxLen(16)]])
• evaluation_factors_long (list[str])
• metrics (list[Annotated[str, MaxLen(16)]])
• metrics_long (list[str])
• results (list[list[str | float | int]])
• results_summary (str | None)

Validators:

• _validate_list_lengths

dataset_id pydantic-field

dataset_id: DatasetId

Dataset used for evaluation.

dataset_role pydantic-field

dataset_role: Literal[
    "train", "validation", "test", "independent", "unknown"
]

Role of the dataset used for evaluation.

• train: dataset was (part of) the training data
• validation: dataset was (part of) the validation data used during training, e.g. used for model selection or hyperparameter tuning
• test: dataset was (part of) the designated test data; not used during training or validation, but acquired from the same source/distribution as training data
• independent: dataset is entirely independent test data; not used during training or validation, and acquired from a different source/distribution than training data
• unknown: role of the dataset is unknown; choose this if you are not certain if (a subset) of the data was seen by the model during training.

dataset_source pydantic-field

dataset_source: HttpUrl

Source of the dataset.

evaluation_factors pydantic-field

evaluation_factors: list[Annotated[str, MaxLen(16)]]

(Abbreviations of) each evaluation factor.

Evaluation factors are criteria along which model performance is evaluated, e.g. different image conditions like 'low SNR', 'high cell density', or different biological conditions like 'cell type A', 'cell type B'. An 'overall' factor may be included to summarize performance across all conditions.

evaluation_factors_long pydantic-field

evaluation_factors_long: list[str]

Descriptions (long form) of each evaluation factor.

metrics pydantic-field

metrics: list[Annotated[str, MaxLen(16)]]

(Abbreviations of) metrics used for evaluation.

metrics_long pydantic-field

metrics_long: list[str]

Description of each metric used.

model_id pydantic-field

model_id: ModelId | None = None

Model being evaluated.

results pydantic-field

results: list[list[str | float | int]]

Results for each metric (rows; outer list) and each evaluation factor (columns; inner list).

results_summary pydantic-field

results_summary: str | None = None

Interpretation of results for general audience.

Consider

• Overall model performance
• Comparison to existing methods
• Limitations and areas for improvement

sample_count pydantic-field

sample_count: int

Number of evaluated samples.

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

format_md

format_md()

Source code in src/bioimageio/spec/model/v0_5.py

    def format_md(self):
        results_header = ["Metric"] + self.evaluation_factors
        results_table_cells = [results_header, ["---"] * len(results_header)] + [
            [metric] + [str(r) for r in row]
            for metric, row in zip(self.metrics, self.results)
        ]

        results_table = "".join(
            "| " + " | ".join(row) + " |\n" for row in results_table_cells
        )
        factors = "".join(
            f"\n - {ef}: {efl}"
            for ef, efl in zip(self.evaluation_factors, self.evaluation_factors_long)
        )
        metrics = "".join(
            f"\n - {em}: {eml}" for em, eml in zip(self.metrics, self.metrics_long)
        )

        return f"""## Testing Data, Factors & Metrics

Evaluation of {self.model_id or "this"} model on the {self.dataset_id} dataset (dataset role: {self.dataset_role}).

### Testing Data

- **Source:** [{self.dataset_id}]({self.dataset_source})
- **Size:** {self.sample_count} evaluated samples

### Factors
{factors}

### Metrics
{metrics}

## Results

### Quantitative Results

{results_table}

### Summary

{self.results_summary or "missing"}

"""

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

FileDescr pydantic-model

Bases: Node

A file description

Show JSON schema:

{
  "$defs": {
    "RelativeFilePath": {
      "description": "A path relative to the `rdf.yaml` file (also if the RDF source is a URL).",
      "format": "path",
      "title": "RelativeFilePath",
      "type": "string"
    }
  },
  "additionalProperties": false,
  "description": "A file description",
  "properties": {
    "source": {
      "anyOf": [
        {
          "description": "A URL with the HTTP or HTTPS scheme.",
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "title": "HttpUrl",
          "type": "string"
        },
        {
          "$ref": "#/$defs/RelativeFilePath"
        },
        {
          "format": "file-path",
          "title": "FilePath",
          "type": "string"
        }
      ],
      "description": "File source",
      "title": "FileSource"
    },
    "sha256": {
      "anyOf": [
        {
          "description": "A SHA-256 hash value",
          "maxLength": 64,
          "minLength": 64,
          "title": "Sha256",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "SHA256 hash value of the **source** file.",
      "title": "Sha256"
    }
  },
  "required": [
    "source"
  ],
  "title": "_internal.io.FileDescr",
  "type": "object"
}

Fields:

• source (FileSource)
• sha256 (Sha256 | None)

sha256 pydantic-field

sha256: Sha256 | None = None

SHA256 hash value of the source file.

source pydantic-field

source: FileSource

File source

suffix property

suffix: str

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

download

download(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

alias for .get_reader

Source code in src/bioimageio/spec/_internal/io.py

def download(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """alias for `.get_reader`"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

get_reader

get_reader(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

open the file source (download if needed)

Source code in src/bioimageio/spec/_internal/io.py

def get_reader(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """open the file source (download if needed)"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

validate_sha256

validate_sha256(force_recompute: bool = False) -> None

validate the sha256 hash value of the source file

Source code in src/bioimageio/spec/_internal/io.py

def validate_sha256(self, force_recompute: bool = False) -> None:
    """validate the sha256 hash value of the **source** file"""
    context = get_validation_context()
    src_str = str(self.source)
    if force_recompute:
        actual_sha = None
    else:
        actual_sha = context.known_files.get(src_str)

    if actual_sha is None:
        if context.perform_io_checks or force_recompute:
            reader = get_reader(self.source, sha256=self.sha256)
            if force_recompute:
                actual_sha = get_sha256(reader)
            else:
                actual_sha = reader.sha256

            context.known_files[src_str] = actual_sha
        elif context.known_files and src_str not in context.known_files:
            # perform_io_checks is False, but known files were given,
            # so we expect all file references to be in there
            raise ValueError(f"File {src_str} not found in `known_files`.")

    if actual_sha is None or self.sha256 == actual_sha:
        return
    elif self.sha256 is None or context.update_hashes:
        self.sha256 = actual_sha
    elif self.sha256 != actual_sha:
        raise ValueError(
            f"Sha256 mismatch for {self.source}. Expected {self.sha256}, got "
            + f"{actual_sha}. Update expected `sha256` or point to the matching "
            + "file."
        )

FixedZeroMeanUnitVarianceAlongAxisKwargs pydantic-model

Bases: KwargsNode

key word arguments for FixedZeroMeanUnitVarianceDescr

Show JSON schema:

{
  "additionalProperties": false,
  "description": "key word arguments for [FixedZeroMeanUnitVarianceDescr][]",
  "properties": {
    "mean": {
      "description": "The mean value(s) to normalize with.",
      "items": {
        "type": "number"
      },
      "minItems": 1,
      "title": "Mean",
      "type": "array"
    },
    "std": {
      "description": "The standard deviation value(s) to normalize with.\nSize must match `mean` values.",
      "items": {
        "minimum": 1e-06,
        "type": "number"
      },
      "minItems": 1,
      "title": "Std",
      "type": "array"
    },
    "axis": {
      "description": "The axis of the mean/std values to normalize each entry along that dimension\nseparately.",
      "examples": [
        "channel",
        "index"
      ],
      "maxLength": 16,
      "minLength": 1,
      "title": "AxisId",
      "type": "string"
    }
  },
  "required": [
    "mean",
    "std",
    "axis"
  ],
  "title": "model.v0_5.FixedZeroMeanUnitVarianceAlongAxisKwargs",
  "type": "object"
}

Fields:

• mean (NotEmpty[list[float]])
• std (NotEmpty[list[Annotated[float, Ge(1e-06)]]])
• axis (Annotated[NonBatchAxisId, Field(examples=['channel', 'index'])])

Validators:

• _mean_and_std_match

axis pydantic-field

axis: Annotated[
    NonBatchAxisId, Field(examples=["channel", "index"])
]

The axis of the mean/std values to normalize each entry along that dimension separately.

mean pydantic-field

mean: NotEmpty[list[float]]

The mean value(s) to normalize with.

std pydantic-field

std: NotEmpty[list[Annotated[float, Ge(1e-06)]]]

The standard deviation value(s) to normalize with. Size must match mean values.

__contains__

__contains__(item: str) -> bool

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __contains__(self, item: str) -> bool:
    return item in self.__class__.model_fields

__getitem__

__getitem__(item: str) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __getitem__(self, item: str) -> Any:
    if item in self.__class__.model_fields:
        return getattr(self, item)
    else:
        raise KeyError(item)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get

get(item: str, default: Any = None) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def get(self, item: str, default: Any = None) -> Any:
    return self[item] if item in self else default  # ruff: ignore[SIM401]

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

FixedZeroMeanUnitVarianceDescr pydantic-model

Bases: NodeWithExplicitlySetFields

Subtract a given mean and divide by the standard deviation.

Normalize with fixed, precomputed values for FixedZeroMeanUnitVarianceKwargs.mean and FixedZeroMeanUnitVarianceKwargs.std Use FixedZeroMeanUnitVarianceAlongAxisKwargs for independent scaling along given axes.

Examples:

1. scalar value for whole tensor

   • in YAML

     preprocessing:
       - id: fixed_zero_mean_unit_variance
         kwargs:
           mean: 103.5
           std: 13.7
     

   • in Python

     preprocessing = [FixedZeroMeanUnitVarianceDescr( ... kwargs=FixedZeroMeanUnitVarianceKwargs(mean=103.5, std=13.7) ... )]

2. independently along an axis

   • in YAML

     preprocessing:
       - id: fixed_zero_mean_unit_variance
         kwargs:
           axis: channel
           mean: [101.5, 102.5, 103.5]
           std: [11.7, 12.7, 13.7]
     

   • in Python

     preprocessing = [FixedZeroMeanUnitVarianceDescr( ... kwargs=FixedZeroMeanUnitVarianceAlongAxisKwargs( ... axis=AxisId("channel"), ... mean=[101.5, 102.5, 103.5], ... std=[11.7, 12.7, 13.7], ... ) ... )]

Show JSON schema:

{
  "$defs": {
    "FixedZeroMeanUnitVarianceAlongAxisKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [FixedZeroMeanUnitVarianceDescr][]",
      "properties": {
        "mean": {
          "description": "The mean value(s) to normalize with.",
          "items": {
            "type": "number"
          },
          "minItems": 1,
          "title": "Mean",
          "type": "array"
        },
        "std": {
          "description": "The standard deviation value(s) to normalize with.\nSize must match `mean` values.",
          "items": {
            "minimum": 1e-06,
            "type": "number"
          },
          "minItems": 1,
          "title": "Std",
          "type": "array"
        },
        "axis": {
          "description": "The axis of the mean/std values to normalize each entry along that dimension\nseparately.",
          "examples": [
            "channel",
            "index"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        }
      },
      "required": [
        "mean",
        "std",
        "axis"
      ],
      "title": "model.v0_5.FixedZeroMeanUnitVarianceAlongAxisKwargs",
      "type": "object"
    },
    "FixedZeroMeanUnitVarianceKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [FixedZeroMeanUnitVarianceDescr][]",
      "properties": {
        "mean": {
          "description": "The mean value to normalize with.",
          "title": "Mean",
          "type": "number"
        },
        "std": {
          "description": "The standard deviation value to normalize with.",
          "minimum": 1e-06,
          "title": "Std",
          "type": "number"
        }
      },
      "required": [
        "mean",
        "std"
      ],
      "title": "model.v0_5.FixedZeroMeanUnitVarianceKwargs",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "description": "Subtract a given mean and divide by the standard deviation.\n\nNormalize with fixed, precomputed values for\n`FixedZeroMeanUnitVarianceKwargs.mean` and `FixedZeroMeanUnitVarianceKwargs.std`\nUse `FixedZeroMeanUnitVarianceAlongAxisKwargs` for independent scaling along given\naxes.\n\nExamples:\n1. scalar value for whole tensor\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: fixed_zero_mean_unit_variance\n        kwargs:\n          mean: 103.5\n          std: 13.7\n    ```\n    - in Python\n    >>> preprocessing = [FixedZeroMeanUnitVarianceDescr(\n    ...   kwargs=FixedZeroMeanUnitVarianceKwargs(mean=103.5, std=13.7)\n    ... )]\n\n2. independently along an axis\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: fixed_zero_mean_unit_variance\n        kwargs:\n          axis: channel\n          mean: [101.5, 102.5, 103.5]\n          std: [11.7, 12.7, 13.7]\n    ```\n    - in Python\n    >>> preprocessing = [FixedZeroMeanUnitVarianceDescr(\n    ...   kwargs=FixedZeroMeanUnitVarianceAlongAxisKwargs(\n    ...     axis=AxisId(\"channel\"),\n    ...     mean=[101.5, 102.5, 103.5],\n    ...     std=[11.7, 12.7, 13.7],\n    ...   )\n    ... )]",
  "properties": {
    "id": {
      "const": "fixed_zero_mean_unit_variance",
      "title": "Id",
      "type": "string"
    },
    "kwargs": {
      "anyOf": [
        {
          "$ref": "#/$defs/FixedZeroMeanUnitVarianceKwargs"
        },
        {
          "$ref": "#/$defs/FixedZeroMeanUnitVarianceAlongAxisKwargs"
        }
      ],
      "title": "Kwargs"
    }
  },
  "required": [
    "id",
    "kwargs"
  ],
  "title": "model.v0_5.FixedZeroMeanUnitVarianceDescr",
  "type": "object"
}

Fields:

• id (Literal['fixed_zero_mean_unit_variance'])
• kwargs (FixedZeroMeanUnitVarianceKwargs | FixedZeroMeanUnitVarianceAlongAxisKwargs)

id pydantic-field

id: Literal["fixed_zero_mean_unit_variance"] = (
    "fixed_zero_mean_unit_variance"
)

implemented_id class-attribute

implemented_id: Literal["fixed_zero_mean_unit_variance"] = (
    "fixed_zero_mean_unit_variance"
)

kwargs pydantic-field

kwargs: (
    FixedZeroMeanUnitVarianceKwargs
    | FixedZeroMeanUnitVarianceAlongAxisKwargs
)

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

FixedZeroMeanUnitVarianceKwargs pydantic-model

Bases: KwargsNode

key word arguments for FixedZeroMeanUnitVarianceDescr

Show JSON schema:

{
  "additionalProperties": false,
  "description": "key word arguments for [FixedZeroMeanUnitVarianceDescr][]",
  "properties": {
    "mean": {
      "description": "The mean value to normalize with.",
      "title": "Mean",
      "type": "number"
    },
    "std": {
      "description": "The standard deviation value to normalize with.",
      "minimum": 1e-06,
      "title": "Std",
      "type": "number"
    }
  },
  "required": [
    "mean",
    "std"
  ],
  "title": "model.v0_5.FixedZeroMeanUnitVarianceKwargs",
  "type": "object"
}

Fields:

• mean (float)
• std (Annotated[float, Ge(1e-06)])

mean pydantic-field

mean: float

The mean value to normalize with.

std pydantic-field

std: Annotated[float, Ge(1e-06)]

The standard deviation value to normalize with.

__contains__

__contains__(item: str) -> bool

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __contains__(self, item: str) -> bool:
    return item in self.__class__.model_fields

__getitem__

__getitem__(item: str) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __getitem__(self, item: str) -> Any:
    if item in self.__class__.model_fields:
        return getattr(self, item)
    else:
        raise KeyError(item)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get

get(item: str, default: Any = None) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def get(self, item: str, default: Any = None) -> Any:
    return self[item] if item in self else default  # ruff: ignore[SIM401]

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

HttpUrl

Bases: RootHttpUrl

              flowchart TD
              bioimageio.spec.model.v0_5.HttpUrl[HttpUrl]
              bioimageio.spec._internal.root_url.RootHttpUrl[RootHttpUrl]
              bioimageio.spec._internal.root_url._RootUrl[_RootUrl]
              bioimageio.spec._internal.validated_string.ValidatedString[ValidatedString]

                              bioimageio.spec._internal.root_url.RootHttpUrl --> bioimageio.spec.model.v0_5.HttpUrl
                                bioimageio.spec._internal.root_url._RootUrl --> bioimageio.spec._internal.root_url.RootHttpUrl
                                bioimageio.spec._internal.validated_string.ValidatedString --> bioimageio.spec._internal.root_url._RootUrl
                




              click bioimageio.spec.model.v0_5.HttpUrl href "" "bioimageio.spec.model.v0_5.HttpUrl"
              click bioimageio.spec._internal.root_url.RootHttpUrl href "" "bioimageio.spec._internal.root_url.RootHttpUrl"
              click bioimageio.spec._internal.root_url._RootUrl href "" "bioimageio.spec._internal.root_url._RootUrl"
              click bioimageio.spec._internal.validated_string.ValidatedString href "" "bioimageio.spec._internal.validated_string.ValidatedString"
            

A URL with the HTTP or HTTPS scheme.

Returned by:

• API Reference  spec
  •  get_resource_package_content
  •  upload

Subclassed by:

•  v0_2  RelativeFilePath
•  v0_2  RelativeFilePath
•  v0_2  RelativeFilePath
•  v0_2  RelativeFilePath
•  v0_3  RelativeFilePath
•  v0_3  RelativeFilePath
•  v0_3  RelativeFilePath
•  v0_3  RelativeFilePath
•  v0_4  RelativeFilePath
•  v0_5  RelativeFilePath
• API Reference  common  RelativeFilePath

Used by:

•  v0_2  NotebookSource
•  v0_3  NotebookSource
• API Reference  common  FileSource

Methods:

Name	Description
__get_pydantic_core_schema__
__get_pydantic_json_schema__
__new__
__truediv__
absolute	analog to absolute method of pathlib.
exists	True if URL is available

Attributes:

Name	Type	Description
host	str | None
parent	RootHttpUrl
parents	Iterable[RootHttpUrl]	iterate over all URL parents (max 100)
path	str | None
root_model	type[RootModel[Any]]	the pydantic root model to validate the string
scheme	str
suffix	str
suffixes	list[str]

host property

host: str | None

parent property

parent: RootHttpUrl

parents property

parents: Iterable[RootHttpUrl]

iterate over all URL parents (max 100)

path property

path: str | None

root_model class-attribute

root_model: type[RootModel[Any]] = RootModel[
    pydantic.HttpUrl
]

the pydantic root model to validate the string

scheme property

scheme: str

suffix property

suffix: str

suffixes property

suffixes: list[str]

__get_pydantic_core_schema__ classmethod

__get_pydantic_core_schema__(
    source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_core_schema__(
    cls, source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema:
    return no_info_after_validator_function(cls, handler(str))

__get_pydantic_json_schema__ classmethod

__get_pydantic_json_schema__(
    core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_json_schema__(
    cls, core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue:
    json_schema = cls.root_model.model_json_schema(mode=handler.mode)
    json_schema["title"] = cls.__name__.strip("_")
    if cls.__doc__:
        json_schema["description"] = cls.__doc__

    return json_schema

__new__

__new__(object: object)

Source code in src/bioimageio/spec/_internal/validated_string.py

def __new__(cls, object: object):
    _validated = cls.root_model.model_validate(object).root
    self = super().__new__(cls, _validated)
    self._validated = _validated
    return self._after_validator()

__truediv__

__truediv__(other: str) -> RootHttpUrl

Source code in src/bioimageio/spec/_internal/root_url.py

def __truediv__(self, other: str) -> RootHttpUrl:
    parsed = urlsplit(str(self))
    return RootHttpUrl(
        urlunsplit(
            (
                parsed.scheme,
                parsed.netloc,
                f"{parsed.path.strip('/')}/{other.strip('/')}",
                parsed.query,
                parsed.fragment,
            )
        )
    )

absolute

absolute()

analog to absolute method of pathlib.

Source code in src/bioimageio/spec/_internal/root_url.py

def absolute(self):
    """analog to `absolute` method of pathlib."""
    return self

exists

exists()

True if URL is available

Source code in src/bioimageio/spec/_internal/url.py

def exists(self):
    """True if URL is available"""
    if self._exists is None:
        ctxt = get_validation_context()
        try:
            with ctxt.replace(warning_level=warning_levels.WARNING):
                self._validated = _validate_url(self._validated)
        except Exception as e:
            if ctxt.log_warnings:
                logger.info(e)

            self._exists = False
        else:
            self._exists = True

    return self._exists

Identifier

Bases: ValidatedString

              flowchart TD
              bioimageio.spec.model.v0_5.Identifier[Identifier]
              bioimageio.spec._internal.validated_string.ValidatedString[ValidatedString]

                              bioimageio.spec._internal.validated_string.ValidatedString --> bioimageio.spec.model.v0_5.Identifier
                


              click bioimageio.spec.model.v0_5.Identifier href "" "bioimageio.spec.model.v0_5.Identifier"
              click bioimageio.spec._internal.validated_string.ValidatedString href "" "bioimageio.spec._internal.validated_string.ValidatedString"
            

Methods:

Name	Description
__get_pydantic_core_schema__
__get_pydantic_json_schema__
__new__

Attributes:

Name	Type	Description
root_model	type[RootModel[Any]]	the pydantic root model to validate the string

root_model class-attribute

root_model: type[RootModel[Any]] = RootModel[IdentifierAnno]

the pydantic root model to validate the string

__get_pydantic_core_schema__ classmethod

__get_pydantic_core_schema__(
    source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_core_schema__(
    cls, source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema:
    return no_info_after_validator_function(cls, handler(str))

__get_pydantic_json_schema__ classmethod

__get_pydantic_json_schema__(
    core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_json_schema__(
    cls, core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue:
    json_schema = cls.root_model.model_json_schema(mode=handler.mode)
    json_schema["title"] = cls.__name__.strip("_")
    if cls.__doc__:
        json_schema["description"] = cls.__doc__

    return json_schema

__new__

__new__(object: object)

Source code in src/bioimageio/spec/_internal/validated_string.py

def __new__(cls, object: object):
    _validated = cls.root_model.model_validate(object).root
    self = super().__new__(cls, _validated)
    self._validated = _validated
    return self._after_validator()

IndexAxisBase pydantic-model

Bases: AxisBase

Show JSON schema:

{
  "additionalProperties": false,
  "properties": {
    "id": {
      "default": "index",
      "maxLength": 16,
      "minLength": 1,
      "title": "AxisId",
      "type": "string"
    },
    "description": {
      "default": "",
      "description": "A short description of this axis beyond its type and id.",
      "maxLength": 128,
      "title": "Description",
      "type": "string"
    },
    "type": {
      "const": "index",
      "title": "Type",
      "type": "string"
    }
  },
  "required": [
    "type"
  ],
  "title": "model.v0_5.IndexAxisBase",
  "type": "object"
}

Fields:

• description (Annotated[str, MaxLen(128)])
• type (Literal['index'])
• id (NonBatchAxisId)

description pydantic-field

description: Annotated[str, MaxLen(128)] = ''

A short description of this axis beyond its type and id.

id pydantic-field

id: NonBatchAxisId = DEFAULT_INDEX_AXIS_ID

An axis id unique across all axes of one tensor.

implemented_type class-attribute

implemented_type: Literal['index'] = 'index'

scale property

scale: float

type pydantic-field

type: Literal['index'] = 'index'

unit property

unit

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

IndexInputAxis pydantic-model

Bases: IndexAxisBase, _WithInputAxisSize

Show JSON schema:

{
  "$defs": {
    "ParameterizedSize": {
      "additionalProperties": false,
      "description": "Describes a range of valid tensor axis sizes as `size = min + n*step`.\n\n- **min** and **step** are given by the model description.\n- All blocksize paramters n = 0,1,2,... yield a valid `size`.\n- A greater blocksize paramter n = 0,1,2,... results in a greater **size**.\n  This allows to adjust the axis size more generically.",
      "properties": {
        "min": {
          "exclusiveMinimum": 0,
          "title": "Min",
          "type": "integer"
        },
        "step": {
          "exclusiveMinimum": 0,
          "title": "Step",
          "type": "integer"
        }
      },
      "required": [
        "min",
        "step"
      ],
      "title": "model.v0_5.ParameterizedSize",
      "type": "object"
    },
    "SizeReference": {
      "additionalProperties": false,
      "description": "A tensor axis size (extent in pixels/frames) defined in relation to a reference axis.\n\n`axis.size = reference.size * reference.scale / axis.scale + offset`\n\nNote:\n1. The axis and the referenced axis need to have the same unit (or no unit).\n2. Batch axes may not be referenced.\n3. Fractions are rounded down.\n4. If the reference axis is `concatenable` the referencing axis is assumed to be\n    `concatenable` as well with the same block order.\n\nExample:\nAn unisotropic input image of w*h=100*49 pixels depicts a phsical space of 200*196mm\u00b2.\nLet's assume that we want to express the image height h in relation to its width w\ninstead of only accepting input images of exactly 100*49 pixels\n(for example to express a range of valid image shapes by parametrizing w, see `ParameterizedSize`).\n\n>>> w = SpaceInputAxis(id=AxisId(\"w\"), size=100, unit=\"millimeter\", scale=2)\n>>> h = SpaceInputAxis(\n...     id=AxisId(\"h\"),\n...     size=SizeReference(tensor_id=TensorId(\"input\"), axis_id=AxisId(\"w\"), offset=-1),\n...     unit=\"millimeter\",\n...     scale=4,\n... )\n>>> print(h.size.get_size(h, w))\n49\n\n\u21d2 h = w * w.scale / h.scale + offset = 100 * 2mm / 4mm - 1 = 49",
      "properties": {
        "tensor_id": {
          "description": "tensor id of the reference axis",
          "maxLength": 32,
          "minLength": 1,
          "title": "TensorId",
          "type": "string"
        },
        "axis_id": {
          "description": "axis id of the reference axis",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "offset": {
          "default": 0,
          "title": "Offset",
          "type": "integer"
        }
      },
      "required": [
        "tensor_id",
        "axis_id"
      ],
      "title": "model.v0_5.SizeReference",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "properties": {
    "size": {
      "anyOf": [
        {
          "exclusiveMinimum": 0,
          "type": "integer"
        },
        {
          "$ref": "#/$defs/ParameterizedSize"
        },
        {
          "$ref": "#/$defs/SizeReference"
        }
      ],
      "description": "The size/length of this axis can be specified as\n- fixed integer\n- parameterized series of valid sizes ([ParameterizedSize][])\n- reference to another axis with an optional offset ([SizeReference][])",
      "examples": [
        10,
        {
          "min": 32,
          "step": 16
        },
        {
          "axis_id": "a",
          "offset": 5,
          "tensor_id": "t"
        }
      ],
      "title": "Size"
    },
    "id": {
      "default": "index",
      "maxLength": 16,
      "minLength": 1,
      "title": "AxisId",
      "type": "string"
    },
    "description": {
      "default": "",
      "description": "A short description of this axis beyond its type and id.",
      "maxLength": 128,
      "title": "Description",
      "type": "string"
    },
    "type": {
      "const": "index",
      "title": "Type",
      "type": "string"
    },
    "concatenable": {
      "default": false,
      "description": "If a model has a `concatenable` input axis, it can be processed blockwise,\nsplitting a longer sample axis into blocks matching its input tensor description.\nOutput axes are concatenable if they have a [SizeReference][] to a concatenable\ninput axis.",
      "title": "Concatenable",
      "type": "boolean"
    }
  },
  "required": [
    "size",
    "type"
  ],
  "title": "model.v0_5.IndexInputAxis",
  "type": "object"
}

Fields:

• size (Annotated[Annotated[int, Gt(0)] | ParameterizedSize | SizeReference, Field(examples=[10, ParameterizedSize(min=32, step=16).model_dump(mode='json'), {'tensor_id': 't', 'axis_id': 'a', 'offset': 5}])])
• id (NonBatchAxisId)
• description (Annotated[str, MaxLen(128)])
• type (Literal['index'])
• concatenable (bool)

concatenable pydantic-field

concatenable: bool = False

If a model has a concatenable input axis, it can be processed blockwise, splitting a longer sample axis into blocks matching its input tensor description. Output axes are concatenable if they have a SizeReference to a concatenable input axis.

description pydantic-field

description: Annotated[str, MaxLen(128)] = ''

A short description of this axis beyond its type and id.

id pydantic-field

id: NonBatchAxisId = DEFAULT_INDEX_AXIS_ID

An axis id unique across all axes of one tensor.

implemented_type class-attribute

implemented_type: Literal['index'] = 'index'

scale property

scale: float

size pydantic-field

size: Annotated[
    Annotated[int, Gt(0)]
    | ParameterizedSize
    | SizeReference,
    Field(
        examples=[
            10,
            ParameterizedSize(min=32, step=16).model_dump(
                mode="json"
            ),
            {"tensor_id": "t", "axis_id": "a", "offset": 5},
        ]
    ),
]

The size/length of this axis can be specified as - fixed integer - parameterized series of valid sizes (ParameterizedSize) - reference to another axis with an optional offset (SizeReference)

type pydantic-field

type: Literal['index'] = 'index'

unit property

unit

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

IndexOutputAxis pydantic-model

Bases: IndexAxisBase

Show JSON schema:

{
  "$defs": {
    "DataDependentSize": {
      "additionalProperties": false,
      "properties": {
        "min": {
          "default": 1,
          "exclusiveMinimum": 0,
          "title": "Min",
          "type": "integer"
        },
        "max": {
          "anyOf": [
            {
              "exclusiveMinimum": 1,
              "type": "integer"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Max"
        }
      },
      "title": "model.v0_5.DataDependentSize",
      "type": "object"
    },
    "SizeReference": {
      "additionalProperties": false,
      "description": "A tensor axis size (extent in pixels/frames) defined in relation to a reference axis.\n\n`axis.size = reference.size * reference.scale / axis.scale + offset`\n\nNote:\n1. The axis and the referenced axis need to have the same unit (or no unit).\n2. Batch axes may not be referenced.\n3. Fractions are rounded down.\n4. If the reference axis is `concatenable` the referencing axis is assumed to be\n    `concatenable` as well with the same block order.\n\nExample:\nAn unisotropic input image of w*h=100*49 pixels depicts a phsical space of 200*196mm\u00b2.\nLet's assume that we want to express the image height h in relation to its width w\ninstead of only accepting input images of exactly 100*49 pixels\n(for example to express a range of valid image shapes by parametrizing w, see `ParameterizedSize`).\n\n>>> w = SpaceInputAxis(id=AxisId(\"w\"), size=100, unit=\"millimeter\", scale=2)\n>>> h = SpaceInputAxis(\n...     id=AxisId(\"h\"),\n...     size=SizeReference(tensor_id=TensorId(\"input\"), axis_id=AxisId(\"w\"), offset=-1),\n...     unit=\"millimeter\",\n...     scale=4,\n... )\n>>> print(h.size.get_size(h, w))\n49\n\n\u21d2 h = w * w.scale / h.scale + offset = 100 * 2mm / 4mm - 1 = 49",
      "properties": {
        "tensor_id": {
          "description": "tensor id of the reference axis",
          "maxLength": 32,
          "minLength": 1,
          "title": "TensorId",
          "type": "string"
        },
        "axis_id": {
          "description": "axis id of the reference axis",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "offset": {
          "default": 0,
          "title": "Offset",
          "type": "integer"
        }
      },
      "required": [
        "tensor_id",
        "axis_id"
      ],
      "title": "model.v0_5.SizeReference",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "properties": {
    "id": {
      "default": "index",
      "maxLength": 16,
      "minLength": 1,
      "title": "AxisId",
      "type": "string"
    },
    "description": {
      "default": "",
      "description": "A short description of this axis beyond its type and id.",
      "maxLength": 128,
      "title": "Description",
      "type": "string"
    },
    "type": {
      "const": "index",
      "title": "Type",
      "type": "string"
    },
    "size": {
      "anyOf": [
        {
          "exclusiveMinimum": 0,
          "type": "integer"
        },
        {
          "$ref": "#/$defs/SizeReference"
        },
        {
          "$ref": "#/$defs/DataDependentSize"
        }
      ],
      "description": "The size/length of this axis can be specified as\n- fixed integer\n- reference to another axis with an optional offset ([SizeReference][])\n- data dependent size using [DataDependentSize][] (size is only known after model inference)",
      "examples": [
        10,
        {
          "axis_id": "a",
          "offset": 5,
          "tensor_id": "t"
        }
      ],
      "title": "Size"
    }
  },
  "required": [
    "type",
    "size"
  ],
  "title": "model.v0_5.IndexOutputAxis",
  "type": "object"
}

Fields:

• id (NonBatchAxisId)
• description (Annotated[str, MaxLen(128)])
• type (Literal['index'])
• size (Annotated[Annotated[int, Gt(0)] | SizeReference | DataDependentSize, Field(examples=[10, {'tensor_id': 't', 'axis_id': 'a', 'offset': 5}])])

description pydantic-field

description: Annotated[str, MaxLen(128)] = ''

A short description of this axis beyond its type and id.

id pydantic-field

id: NonBatchAxisId = DEFAULT_INDEX_AXIS_ID

An axis id unique across all axes of one tensor.

implemented_type class-attribute

implemented_type: Literal['index'] = 'index'

scale property

scale: float

size pydantic-field

size: Annotated[
    Annotated[int, Gt(0)]
    | SizeReference
    | DataDependentSize,
    Field(
        examples=[
            10,
            {"tensor_id": "t", "axis_id": "a", "offset": 5},
        ]
    ),
]

The size/length of this axis can be specified as - fixed integer - reference to another axis with an optional offset (SizeReference) - data dependent size using DataDependentSize (size is only known after model inference)

type pydantic-field

type: Literal['index'] = 'index'

unit property

unit

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

InputTensorDescr pydantic-model

Bases: TensorDescrBase[InputAxis]

Show JSON schema:

{
  "$defs": {
    "BatchAxis": {
      "additionalProperties": false,
      "properties": {
        "id": {
          "default": "batch",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "batch",
          "title": "Type",
          "type": "string"
        },
        "size": {
          "anyOf": [
            {
              "const": 1,
              "type": "integer"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The batch size may be fixed to 1,\notherwise (the default) it may be chosen arbitrarily depending on available memory",
          "title": "Size"
        }
      },
      "required": [
        "type"
      ],
      "title": "model.v0_5.BatchAxis",
      "type": "object"
    },
    "BinarizeAlongAxisKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [BinarizeDescr][]",
      "properties": {
        "threshold": {
          "description": "The fixed threshold values along `axis`",
          "items": {
            "type": "number"
          },
          "minItems": 1,
          "title": "Threshold",
          "type": "array"
        },
        "axis": {
          "description": "The `threshold` axis",
          "examples": [
            "channel"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        }
      },
      "required": [
        "threshold",
        "axis"
      ],
      "title": "model.v0_5.BinarizeAlongAxisKwargs",
      "type": "object"
    },
    "BinarizeDescr": {
      "additionalProperties": false,
      "description": "Binarize the tensor with a fixed threshold.\n\nValues above [BinarizeKwargs.threshold][]/[BinarizeAlongAxisKwargs.threshold][]\nwill be set to one, values below the threshold to zero.\n\nExamples:\n- in YAML\n    ```yaml\n    postprocessing:\n      - id: binarize\n        kwargs:\n          axis: 'channel'\n          threshold: [0.25, 0.5, 0.75]\n    ```\n- in Python:\n\n    >>> postprocessing = [BinarizeDescr(\n    ...   kwargs=BinarizeAlongAxisKwargs(\n    ...       axis=AxisId('channel'),\n    ...       threshold=[0.25, 0.5, 0.75],\n    ...   )\n    ... )]",
      "properties": {
        "id": {
          "const": "binarize",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "anyOf": [
            {
              "$ref": "#/$defs/BinarizeKwargs"
            },
            {
              "$ref": "#/$defs/BinarizeAlongAxisKwargs"
            }
          ],
          "title": "Kwargs"
        }
      },
      "required": [
        "id",
        "kwargs"
      ],
      "title": "model.v0_5.BinarizeDescr",
      "type": "object"
    },
    "BinarizeKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [BinarizeDescr][]",
      "properties": {
        "threshold": {
          "description": "The fixed threshold",
          "title": "Threshold",
          "type": "number"
        }
      },
      "required": [
        "threshold"
      ],
      "title": "model.v0_5.BinarizeKwargs",
      "type": "object"
    },
    "ChannelAxis": {
      "additionalProperties": false,
      "properties": {
        "id": {
          "default": "channel",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "channel",
          "title": "Type",
          "type": "string"
        },
        "channel_names": {
          "description": "Name/label for each channel. The number of channels is given by `len(channel_names)`.",
          "items": {
            "type": "string"
          },
          "minItems": 1,
          "title": "Channel Names",
          "type": "array"
        }
      },
      "required": [
        "type",
        "channel_names"
      ],
      "title": "model.v0_5.ChannelAxis",
      "type": "object"
    },
    "ClipDescr": {
      "additionalProperties": false,
      "description": "Set tensor values below min to min and above max to max.\n\nSee `ScaleRangeDescr` for examples.",
      "properties": {
        "id": {
          "const": "clip",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "$ref": "#/$defs/ClipKwargs"
        }
      },
      "required": [
        "id",
        "kwargs"
      ],
      "title": "model.v0_5.ClipDescr",
      "type": "object"
    },
    "ClipKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [ClipDescr][]",
      "properties": {
        "min": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Minimum value for clipping.\n\nExclusive with [min_percentile][]",
          "title": "Min"
        },
        "min_percentile": {
          "anyOf": [
            {
              "exclusiveMaximum": 100,
              "minimum": 0,
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Minimum percentile for clipping.\n\nExclusive with [min][].\n\nIn range [0, 100).",
          "title": "Min Percentile"
        },
        "max": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Maximum value for clipping.\n\nExclusive with `max_percentile`.",
          "title": "Max"
        },
        "max_percentile": {
          "anyOf": [
            {
              "exclusiveMinimum": 1,
              "maximum": 100,
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Maximum percentile for clipping.\n\nExclusive with `max`.\n\nIn range (1, 100].",
          "title": "Max Percentile"
        },
        "axes": {
          "anyOf": [
            {
              "items": {
                "maxLength": 16,
                "minLength": 1,
                "title": "AxisId",
                "type": "string"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The subset of axes to determine percentiles jointly,\n\ni.e. axes to reduce to compute min/max from `min_percentile`/`max_percentile`.\nFor example to clip 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x')\nresulting in a tensor of equal shape with clipped values per channel, specify `axes=('batch', 'x', 'y')`.\nTo clip samples independently, leave out the 'batch' axis.\n\nOnly valid if `min_percentile` and/or `max_percentile` are set.\n\nDefault: Compute percentiles over all axes jointly.",
          "examples": [
            [
              "batch",
              "x",
              "y"
            ]
          ],
          "title": "Axes"
        }
      },
      "title": "model.v0_5.ClipKwargs",
      "type": "object"
    },
    "ConstantPadding": {
      "additionalProperties": false,
      "properties": {
        "mode": {
          "const": "constant",
          "default": "constant",
          "title": "Mode",
          "type": "string"
        },
        "value": {
          "anyOf": [
            {
              "type": "integer"
            },
            {
              "type": "number"
            }
          ],
          "default": 0,
          "title": "Value"
        }
      },
      "title": "model.v0_5.ConstantPadding",
      "type": "object"
    },
    "EdgePadding": {
      "additionalProperties": false,
      "properties": {
        "mode": {
          "const": "edge",
          "default": "edge",
          "title": "Mode",
          "type": "string"
        }
      },
      "title": "model.v0_5.EdgePadding",
      "type": "object"
    },
    "EnsureDtypeDescr": {
      "additionalProperties": false,
      "description": "Cast the tensor data type to `EnsureDtypeKwargs.dtype` (if not matching).\n\nThis can for example be used to ensure the inner neural network model gets a\ndifferent input tensor data type than the fully described bioimage.io model does.\n\nExamples:\n    The described bioimage.io model (incl. preprocessing) accepts any\n    float32-compatible tensor, normalizes it with percentiles and clipping and then\n    casts it to uint8, which is what the neural network in this example expects.\n    - in YAML\n        ```yaml\n        inputs:\n        - data:\n            type: float32  # described bioimage.io model is compatible with any float32 input tensor\n          preprocessing:\n          - id: scale_range\n              kwargs:\n              axes: ['y', 'x']\n              max_percentile: 99.8\n              min_percentile: 5.0\n          - id: clip\n              kwargs:\n              min: 0.0\n              max: 1.0\n          - id: ensure_dtype  # the neural network of the model requires uint8\n              kwargs:\n              dtype: uint8\n        ```\n    - in Python:\n        >>> preprocessing = [\n        ...     ScaleRangeDescr(\n        ...         kwargs=ScaleRangeKwargs(\n        ...           axes= (AxisId('y'), AxisId('x')),\n        ...           max_percentile= 99.8,\n        ...           min_percentile= 5.0,\n        ...         )\n        ...     ),\n        ...     ClipDescr(kwargs=ClipKwargs(min=0.0, max=1.0)),\n        ...     EnsureDtypeDescr(kwargs=EnsureDtypeKwargs(dtype=\"uint8\")),\n        ... ]",
      "properties": {
        "id": {
          "const": "ensure_dtype",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "$ref": "#/$defs/EnsureDtypeKwargs"
        }
      },
      "required": [
        "id",
        "kwargs"
      ],
      "title": "model.v0_5.EnsureDtypeDescr",
      "type": "object"
    },
    "EnsureDtypeKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [EnsureDtypeDescr][]",
      "properties": {
        "dtype": {
          "enum": [
            "float32",
            "float64",
            "uint8",
            "int8",
            "uint16",
            "int16",
            "uint32",
            "int32",
            "uint64",
            "int64",
            "bool"
          ],
          "title": "Dtype",
          "type": "string"
        }
      },
      "required": [
        "dtype"
      ],
      "title": "model.v0_5.EnsureDtypeKwargs",
      "type": "object"
    },
    "FileDescr": {
      "additionalProperties": false,
      "description": "A file description",
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "File source",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        }
      },
      "required": [
        "source"
      ],
      "title": "_internal.io.FileDescr",
      "type": "object"
    },
    "FixedZeroMeanUnitVarianceAlongAxisKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [FixedZeroMeanUnitVarianceDescr][]",
      "properties": {
        "mean": {
          "description": "The mean value(s) to normalize with.",
          "items": {
            "type": "number"
          },
          "minItems": 1,
          "title": "Mean",
          "type": "array"
        },
        "std": {
          "description": "The standard deviation value(s) to normalize with.\nSize must match `mean` values.",
          "items": {
            "minimum": 1e-06,
            "type": "number"
          },
          "minItems": 1,
          "title": "Std",
          "type": "array"
        },
        "axis": {
          "description": "The axis of the mean/std values to normalize each entry along that dimension\nseparately.",
          "examples": [
            "channel",
            "index"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        }
      },
      "required": [
        "mean",
        "std",
        "axis"
      ],
      "title": "model.v0_5.FixedZeroMeanUnitVarianceAlongAxisKwargs",
      "type": "object"
    },
    "FixedZeroMeanUnitVarianceDescr": {
      "additionalProperties": false,
      "description": "Subtract a given mean and divide by the standard deviation.\n\nNormalize with fixed, precomputed values for\n`FixedZeroMeanUnitVarianceKwargs.mean` and `FixedZeroMeanUnitVarianceKwargs.std`\nUse `FixedZeroMeanUnitVarianceAlongAxisKwargs` for independent scaling along given\naxes.\n\nExamples:\n1. scalar value for whole tensor\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: fixed_zero_mean_unit_variance\n        kwargs:\n          mean: 103.5\n          std: 13.7\n    ```\n    - in Python\n    >>> preprocessing = [FixedZeroMeanUnitVarianceDescr(\n    ...   kwargs=FixedZeroMeanUnitVarianceKwargs(mean=103.5, std=13.7)\n    ... )]\n\n2. independently along an axis\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: fixed_zero_mean_unit_variance\n        kwargs:\n          axis: channel\n          mean: [101.5, 102.5, 103.5]\n          std: [11.7, 12.7, 13.7]\n    ```\n    - in Python\n    >>> preprocessing = [FixedZeroMeanUnitVarianceDescr(\n    ...   kwargs=FixedZeroMeanUnitVarianceAlongAxisKwargs(\n    ...     axis=AxisId(\"channel\"),\n    ...     mean=[101.5, 102.5, 103.5],\n    ...     std=[11.7, 12.7, 13.7],\n    ...   )\n    ... )]",
      "properties": {
        "id": {
          "const": "fixed_zero_mean_unit_variance",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "anyOf": [
            {
              "$ref": "#/$defs/FixedZeroMeanUnitVarianceKwargs"
            },
            {
              "$ref": "#/$defs/FixedZeroMeanUnitVarianceAlongAxisKwargs"
            }
          ],
          "title": "Kwargs"
        }
      },
      "required": [
        "id",
        "kwargs"
      ],
      "title": "model.v0_5.FixedZeroMeanUnitVarianceDescr",
      "type": "object"
    },
    "FixedZeroMeanUnitVarianceKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [FixedZeroMeanUnitVarianceDescr][]",
      "properties": {
        "mean": {
          "description": "The mean value to normalize with.",
          "title": "Mean",
          "type": "number"
        },
        "std": {
          "description": "The standard deviation value to normalize with.",
          "minimum": 1e-06,
          "title": "Std",
          "type": "number"
        }
      },
      "required": [
        "mean",
        "std"
      ],
      "title": "model.v0_5.FixedZeroMeanUnitVarianceKwargs",
      "type": "object"
    },
    "IndexInputAxis": {
      "additionalProperties": false,
      "properties": {
        "size": {
          "anyOf": [
            {
              "exclusiveMinimum": 0,
              "type": "integer"
            },
            {
              "$ref": "#/$defs/ParameterizedSize"
            },
            {
              "$ref": "#/$defs/SizeReference"
            }
          ],
          "description": "The size/length of this axis can be specified as\n- fixed integer\n- parameterized series of valid sizes ([ParameterizedSize][])\n- reference to another axis with an optional offset ([SizeReference][])",
          "examples": [
            10,
            {
              "min": 32,
              "step": 16
            },
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ],
          "title": "Size"
        },
        "id": {
          "default": "index",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "index",
          "title": "Type",
          "type": "string"
        },
        "concatenable": {
          "default": false,
          "description": "If a model has a `concatenable` input axis, it can be processed blockwise,\nsplitting a longer sample axis into blocks matching its input tensor description.\nOutput axes are concatenable if they have a [SizeReference][] to a concatenable\ninput axis.",
          "title": "Concatenable",
          "type": "boolean"
        }
      },
      "required": [
        "size",
        "type"
      ],
      "title": "model.v0_5.IndexInputAxis",
      "type": "object"
    },
    "IntervalOrRatioDataDescr": {
      "additionalProperties": false,
      "properties": {
        "type": {
          "default": "float32",
          "enum": [
            "float32",
            "float64",
            "uint8",
            "int8",
            "uint16",
            "int16",
            "uint32",
            "int32",
            "uint64",
            "int64"
          ],
          "examples": [
            "float32",
            "float64",
            "uint8",
            "uint16"
          ],
          "title": "Type",
          "type": "string"
        },
        "range": {
          "default": [
            null,
            null
          ],
          "description": "Tuple `(minimum, maximum)` specifying the allowed range of the data in this tensor.\n`None` corresponds to min/max of what can be expressed by **type**.",
          "maxItems": 2,
          "minItems": 2,
          "prefixItems": [
            {
              "anyOf": [
                {
                  "type": "number"
                },
                {
                  "type": "null"
                }
              ]
            },
            {
              "anyOf": [
                {
                  "type": "number"
                },
                {
                  "type": "null"
                }
              ]
            }
          ],
          "title": "Range",
          "type": "array"
        },
        "unit": {
          "anyOf": [
            {
              "const": "arbitrary unit",
              "type": "string"
            },
            {
              "description": "An SI unit",
              "minLength": 1,
              "pattern": "^(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^[+-]?[1-9]\\d*)?((\u00b7(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^[+-]?[1-9]\\d*)?)|(/(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^+?[1-9]\\d*)?))*$",
              "title": "SiUnit",
              "type": "string"
            }
          ],
          "default": "arbitrary unit",
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "description": "Scale for data on an interval (or ratio) scale.",
          "title": "Scale",
          "type": "number"
        },
        "offset": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Offset for data on a ratio scale.",
          "title": "Offset"
        }
      },
      "title": "model.v0_5.IntervalOrRatioDataDescr",
      "type": "object"
    },
    "NominalOrOrdinalDataDescr": {
      "additionalProperties": false,
      "properties": {
        "values": {
          "anyOf": [
            {
              "items": {
                "type": "integer"
              },
              "minItems": 1,
              "type": "array"
            },
            {
              "items": {
                "type": "number"
              },
              "minItems": 1,
              "type": "array"
            },
            {
              "items": {
                "type": "boolean"
              },
              "minItems": 1,
              "type": "array"
            },
            {
              "items": {
                "type": "string"
              },
              "minItems": 1,
              "type": "array"
            }
          ],
          "description": "A fixed set of nominal or an ascending sequence of ordinal values.\nIn this case `data.type` is required to be an unsigend integer type, e.g. 'uint8'.\nString `values` are interpreted as labels for tensor values 0, ..., N.\nNote: as YAML 1.2 does not natively support a \"set\" datatype,\nnominal values should be given as a sequence (aka list/array) as well.",
          "title": "Values"
        },
        "type": {
          "default": "uint8",
          "enum": [
            "float32",
            "float64",
            "uint8",
            "int8",
            "uint16",
            "int16",
            "uint32",
            "int32",
            "uint64",
            "int64",
            "bool"
          ],
          "examples": [
            "float32",
            "uint8",
            "uint16",
            "int64",
            "bool"
          ],
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "const": "arbitrary unit",
              "type": "string"
            },
            {
              "description": "An SI unit",
              "minLength": 1,
              "pattern": "^(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^[+-]?[1-9]\\d*)?((\u00b7(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^[+-]?[1-9]\\d*)?)|(/(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^+?[1-9]\\d*)?))*$",
              "title": "SiUnit",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        }
      },
      "required": [
        "values"
      ],
      "title": "model.v0_5.NominalOrOrdinalDataDescr",
      "type": "object"
    },
    "ParameterizedSize": {
      "additionalProperties": false,
      "description": "Describes a range of valid tensor axis sizes as `size = min + n*step`.\n\n- **min** and **step** are given by the model description.\n- All blocksize paramters n = 0,1,2,... yield a valid `size`.\n- A greater blocksize paramter n = 0,1,2,... results in a greater **size**.\n  This allows to adjust the axis size more generically.",
      "properties": {
        "min": {
          "exclusiveMinimum": 0,
          "title": "Min",
          "type": "integer"
        },
        "step": {
          "exclusiveMinimum": 0,
          "title": "Step",
          "type": "integer"
        }
      },
      "required": [
        "min",
        "step"
      ],
      "title": "model.v0_5.ParameterizedSize",
      "type": "object"
    },
    "ReflectPadding": {
      "additionalProperties": false,
      "properties": {
        "mode": {
          "const": "reflect",
          "default": "reflect",
          "title": "Mode",
          "type": "string"
        }
      },
      "title": "model.v0_5.ReflectPadding",
      "type": "object"
    },
    "RelativeFilePath": {
      "description": "A path relative to the `rdf.yaml` file (also if the RDF source is a URL).",
      "format": "path",
      "title": "RelativeFilePath",
      "type": "string"
    },
    "ScaleLinearAlongAxisKwargs": {
      "additionalProperties": false,
      "description": "Key word arguments for [ScaleLinearDescr][]",
      "properties": {
        "axis": {
          "description": "The axis of gain and offset values.",
          "examples": [
            "channel"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "gain": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "items": {
                "type": "number"
              },
              "minItems": 1,
              "type": "array"
            }
          ],
          "default": 1.0,
          "description": "multiplicative factor",
          "title": "Gain"
        },
        "offset": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "items": {
                "type": "number"
              },
              "minItems": 1,
              "type": "array"
            }
          ],
          "default": 0.0,
          "description": "additive term",
          "title": "Offset"
        }
      },
      "required": [
        "axis"
      ],
      "title": "model.v0_5.ScaleLinearAlongAxisKwargs",
      "type": "object"
    },
    "ScaleLinearDescr": {
      "additionalProperties": false,
      "description": "Fixed linear scaling.\n\nExamples:\n  1. Scale with scalar gain and offset\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: scale_linear\n        kwargs:\n          gain: 2.0\n          offset: 3.0\n    ```\n    - in Python:\n\n    >>> preprocessing = [\n    ...     ScaleLinearDescr(kwargs=ScaleLinearKwargs(gain= 2.0, offset=3.0))\n    ... ]\n\n  2. Independent scaling along an axis\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: scale_linear\n        kwargs:\n          axis: 'channel'\n          gain: [1.0, 2.0, 3.0]\n    ```\n    - in Python:\n\n    >>> preprocessing = [\n    ...     ScaleLinearDescr(\n    ...         kwargs=ScaleLinearAlongAxisKwargs(\n    ...             axis=AxisId(\"channel\"),\n    ...             gain=[1.0, 2.0, 3.0],\n    ...         )\n    ...     )\n    ... ]",
      "properties": {
        "id": {
          "const": "scale_linear",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "anyOf": [
            {
              "$ref": "#/$defs/ScaleLinearKwargs"
            },
            {
              "$ref": "#/$defs/ScaleLinearAlongAxisKwargs"
            }
          ],
          "title": "Kwargs"
        }
      },
      "required": [
        "id",
        "kwargs"
      ],
      "title": "model.v0_5.ScaleLinearDescr",
      "type": "object"
    },
    "ScaleLinearKwargs": {
      "additionalProperties": false,
      "description": "Key word arguments for [ScaleLinearDescr][]",
      "properties": {
        "gain": {
          "default": 1.0,
          "description": "multiplicative factor",
          "title": "Gain",
          "type": "number"
        },
        "offset": {
          "default": 0.0,
          "description": "additive term",
          "title": "Offset",
          "type": "number"
        }
      },
      "title": "model.v0_5.ScaleLinearKwargs",
      "type": "object"
    },
    "ScaleRangeDescr": {
      "additionalProperties": false,
      "description": "Scale with percentiles.\n\nExamples:\n1. Scale linearly to map 5th percentile to 0 and 99.8th percentile to 1.0\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: scale_range\n        kwargs:\n          axes: ['y', 'x']\n          max_percentile: 99.8\n          min_percentile: 5.0\n    ```\n    - in Python\n\n    >>> preprocessing = [\n    ...     ScaleRangeDescr(\n    ...         kwargs=ScaleRangeKwargs(\n    ...           axes= (AxisId('y'), AxisId('x')),\n    ...           max_percentile= 99.8,\n    ...           min_percentile= 5.0,\n    ...         )\n    ...     )\n    ... ]\n\n  2. Combine the above scaling with additional clipping to clip values outside the range given by the percentiles.\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: scale_range\n        kwargs:\n          axes: ['y', 'x']\n          max_percentile: 99.8\n          min_percentile: 5.0\n       - id: clip\n         kwargs:\n          min: 0.0\n          max: 1.0\n    ```\n    - in Python\n\n    >>> preprocessing = [\n    ...   ScaleRangeDescr(\n    ...     kwargs=ScaleRangeKwargs(\n    ...       axes= (AxisId('y'), AxisId('x')),\n    ...       max_percentile= 99.8,\n    ...       min_percentile= 5.0,\n    ...     )\n    ...   ),\n    ...   ClipDescr(\n    ...     kwargs=ClipKwargs(\n    ...       min=0.0,\n    ...       max=1.0,\n    ...     )\n    ...   ),\n    ... ]",
      "properties": {
        "id": {
          "const": "scale_range",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "$ref": "#/$defs/ScaleRangeKwargs"
        }
      },
      "required": [
        "id"
      ],
      "title": "model.v0_5.ScaleRangeDescr",
      "type": "object"
    },
    "ScaleRangeKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [ScaleRangeDescr][]\n\nFor `min_percentile`=0.0 (the default) and `max_percentile`=100 (the default)\nthis processing step normalizes data to the [0, 1] intervall.\nFor other percentiles the normalized values will partially be outside the [0, 1]\nintervall. Use `ScaleRange` followed by `ClipDescr` if you want to limit the\nnormalized values to a range.",
      "properties": {
        "axes": {
          "anyOf": [
            {
              "items": {
                "maxLength": 16,
                "minLength": 1,
                "title": "AxisId",
                "type": "string"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The subset of axes to normalize jointly, i.e. axes to reduce to compute the min/max percentile value.\nFor example to normalize 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x')\nresulting in a tensor of equal shape normalized per channel, specify `axes=('batch', 'x', 'y')`.\nTo normalize samples independently, leave out the \"batch\" axis.\nDefault: Scale all axes jointly.",
          "examples": [
            [
              "batch",
              "x",
              "y"
            ]
          ],
          "title": "Axes"
        },
        "min_percentile": {
          "default": 0.0,
          "description": "The lower percentile used to determine the value to align with zero.",
          "exclusiveMaximum": 100,
          "minimum": 0,
          "title": "Min Percentile",
          "type": "number"
        },
        "max_percentile": {
          "default": 100.0,
          "description": "The upper percentile used to determine the value to align with one.\nHas to be bigger than `min_percentile`.\nThe range is 1 to 100 instead of 0 to 100 to avoid mistakenly\naccepting percentiles specified in the range 0.0 to 1.0.",
          "exclusiveMinimum": 1,
          "maximum": 100,
          "title": "Max Percentile",
          "type": "number"
        },
        "eps": {
          "default": 1e-06,
          "description": "Epsilon for numeric stability.\n`out = (tensor - v_lower) / (v_upper - v_lower + eps)`;\nwith `v_lower,v_upper` values at the respective percentiles.",
          "exclusiveMinimum": 0,
          "maximum": 0.1,
          "title": "Eps",
          "type": "number"
        },
        "reference_tensor": {
          "anyOf": [
            {
              "maxLength": 32,
              "minLength": 1,
              "title": "TensorId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "ID of the unprocessed input tensor to compute the percentiles from.\nDefault: The tensor itself.",
          "title": "Reference Tensor"
        }
      },
      "title": "model.v0_5.ScaleRangeKwargs",
      "type": "object"
    },
    "SigmoidDescr": {
      "additionalProperties": false,
      "description": "The logistic sigmoid function, a.k.a. expit function.\n\nExamples:\n- in YAML\n    ```yaml\n    postprocessing:\n      - id: sigmoid\n    ```\n- in Python:\n\n    >>> postprocessing = [SigmoidDescr()]",
      "properties": {
        "id": {
          "const": "sigmoid",
          "title": "Id",
          "type": "string"
        }
      },
      "required": [
        "id"
      ],
      "title": "model.v0_5.SigmoidDescr",
      "type": "object"
    },
    "SizeReference": {
      "additionalProperties": false,
      "description": "A tensor axis size (extent in pixels/frames) defined in relation to a reference axis.\n\n`axis.size = reference.size * reference.scale / axis.scale + offset`\n\nNote:\n1. The axis and the referenced axis need to have the same unit (or no unit).\n2. Batch axes may not be referenced.\n3. Fractions are rounded down.\n4. If the reference axis is `concatenable` the referencing axis is assumed to be\n    `concatenable` as well with the same block order.\n\nExample:\nAn unisotropic input image of w*h=100*49 pixels depicts a phsical space of 200*196mm\u00b2.\nLet's assume that we want to express the image height h in relation to its width w\ninstead of only accepting input images of exactly 100*49 pixels\n(for example to express a range of valid image shapes by parametrizing w, see `ParameterizedSize`).\n\n>>> w = SpaceInputAxis(id=AxisId(\"w\"), size=100, unit=\"millimeter\", scale=2)\n>>> h = SpaceInputAxis(\n...     id=AxisId(\"h\"),\n...     size=SizeReference(tensor_id=TensorId(\"input\"), axis_id=AxisId(\"w\"), offset=-1),\n...     unit=\"millimeter\",\n...     scale=4,\n... )\n>>> print(h.size.get_size(h, w))\n49\n\n\u21d2 h = w * w.scale / h.scale + offset = 100 * 2mm / 4mm - 1 = 49",
      "properties": {
        "tensor_id": {
          "description": "tensor id of the reference axis",
          "maxLength": 32,
          "minLength": 1,
          "title": "TensorId",
          "type": "string"
        },
        "axis_id": {
          "description": "axis id of the reference axis",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "offset": {
          "default": 0,
          "title": "Offset",
          "type": "integer"
        }
      },
      "required": [
        "tensor_id",
        "axis_id"
      ],
      "title": "model.v0_5.SizeReference",
      "type": "object"
    },
    "SoftmaxDescr": {
      "additionalProperties": false,
      "description": "The softmax function.\n\nExamples:\n- in YAML\n    ```yaml\n    postprocessing:\n      - id: softmax\n        kwargs:\n          axis: channel\n    ```\n- in Python:\n\n    >>> postprocessing = [SoftmaxDescr(kwargs=SoftmaxKwargs(axis=AxisId(\"channel\")))]",
      "properties": {
        "id": {
          "const": "softmax",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "$ref": "#/$defs/SoftmaxKwargs"
        }
      },
      "required": [
        "id"
      ],
      "title": "model.v0_5.SoftmaxDescr",
      "type": "object"
    },
    "SoftmaxKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [SoftmaxDescr][]",
      "properties": {
        "axis": {
          "default": "channel",
          "description": "The axis to apply the softmax function along.\nNote:\n    Defaults to 'channel' axis\n    (which may not exist, in which case\n    a different axis id has to be specified).",
          "examples": [
            "channel"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        }
      },
      "title": "model.v0_5.SoftmaxKwargs",
      "type": "object"
    },
    "SpaceInputAxis": {
      "additionalProperties": false,
      "properties": {
        "size": {
          "anyOf": [
            {
              "exclusiveMinimum": 0,
              "type": "integer"
            },
            {
              "$ref": "#/$defs/ParameterizedSize"
            },
            {
              "$ref": "#/$defs/SizeReference"
            }
          ],
          "description": "The size/length of this axis can be specified as\n- fixed integer\n- parameterized series of valid sizes ([ParameterizedSize][])\n- reference to another axis with an optional offset ([SizeReference][])",
          "examples": [
            10,
            {
              "min": 32,
              "step": 16
            },
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ],
          "title": "Size"
        },
        "id": {
          "default": "x",
          "examples": [
            "x",
            "y",
            "z"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "space",
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "enum": [
                "attometer",
                "angstrom",
                "centimeter",
                "decimeter",
                "exameter",
                "femtometer",
                "foot",
                "gigameter",
                "hectometer",
                "inch",
                "kilometer",
                "megameter",
                "meter",
                "micrometer",
                "mile",
                "millimeter",
                "nanometer",
                "parsec",
                "petameter",
                "picometer",
                "terameter",
                "yard",
                "yoctometer",
                "yottameter",
                "zeptometer",
                "zettameter"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "exclusiveMinimum": 0,
          "title": "Scale",
          "type": "number"
        },
        "concatenable": {
          "default": false,
          "description": "If a model has a `concatenable` input axis, it can be processed blockwise,\nsplitting a longer sample axis into blocks matching its input tensor description.\nOutput axes are concatenable if they have a [SizeReference][] to a concatenable\ninput axis.",
          "title": "Concatenable",
          "type": "boolean"
        }
      },
      "required": [
        "size",
        "type"
      ],
      "title": "model.v0_5.SpaceInputAxis",
      "type": "object"
    },
    "SymmetricPadding": {
      "additionalProperties": false,
      "properties": {
        "mode": {
          "const": "symmetric",
          "default": "symmetric",
          "title": "Mode",
          "type": "string"
        }
      },
      "title": "model.v0_5.SymmetricPadding",
      "type": "object"
    },
    "TimeInputAxis": {
      "additionalProperties": false,
      "properties": {
        "size": {
          "anyOf": [
            {
              "exclusiveMinimum": 0,
              "type": "integer"
            },
            {
              "$ref": "#/$defs/ParameterizedSize"
            },
            {
              "$ref": "#/$defs/SizeReference"
            }
          ],
          "description": "The size/length of this axis can be specified as\n- fixed integer\n- parameterized series of valid sizes ([ParameterizedSize][])\n- reference to another axis with an optional offset ([SizeReference][])",
          "examples": [
            10,
            {
              "min": 32,
              "step": 16
            },
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ],
          "title": "Size"
        },
        "id": {
          "default": "time",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "time",
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "enum": [
                "attosecond",
                "centisecond",
                "day",
                "decisecond",
                "exasecond",
                "femtosecond",
                "gigasecond",
                "hectosecond",
                "hour",
                "kilosecond",
                "megasecond",
                "microsecond",
                "millisecond",
                "minute",
                "nanosecond",
                "petasecond",
                "picosecond",
                "second",
                "terasecond",
                "yoctosecond",
                "yottasecond",
                "zeptosecond",
                "zettasecond"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "exclusiveMinimum": 0,
          "title": "Scale",
          "type": "number"
        },
        "concatenable": {
          "default": false,
          "description": "If a model has a `concatenable` input axis, it can be processed blockwise,\nsplitting a longer sample axis into blocks matching its input tensor description.\nOutput axes are concatenable if they have a [SizeReference][] to a concatenable\ninput axis.",
          "title": "Concatenable",
          "type": "boolean"
        }
      },
      "required": [
        "size",
        "type"
      ],
      "title": "model.v0_5.TimeInputAxis",
      "type": "object"
    },
    "ZeroMeanUnitVarianceDescr": {
      "additionalProperties": false,
      "description": "Subtract mean and divide by variance.\n\nExamples:\n    Subtract tensor mean and variance\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: zero_mean_unit_variance\n    ```\n    - in Python\n    >>> preprocessing = [ZeroMeanUnitVarianceDescr()]",
      "properties": {
        "id": {
          "const": "zero_mean_unit_variance",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "$ref": "#/$defs/ZeroMeanUnitVarianceKwargs"
        }
      },
      "required": [
        "id"
      ],
      "title": "model.v0_5.ZeroMeanUnitVarianceDescr",
      "type": "object"
    },
    "ZeroMeanUnitVarianceKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [ZeroMeanUnitVarianceDescr][]",
      "properties": {
        "axes": {
          "anyOf": [
            {
              "items": {
                "maxLength": 16,
                "minLength": 1,
                "title": "AxisId",
                "type": "string"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The subset of axes to normalize jointly, i.e. axes to reduce to compute mean/std.\nFor example to normalize 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x')\nresulting in a tensor of equal shape normalized per channel, specify `axes=('batch', 'x', 'y')`.\nTo normalize each sample independently leave out the 'batch' axis.\nDefault: Scale all axes jointly.",
          "examples": [
            [
              "batch",
              "x",
              "y"
            ]
          ],
          "title": "Axes"
        },
        "eps": {
          "default": 1e-06,
          "description": "epsilon for numeric stability: `out = (tensor - mean) / (std + eps)`.",
          "exclusiveMinimum": 0,
          "maximum": 0.1,
          "title": "Eps",
          "type": "number"
        }
      },
      "title": "model.v0_5.ZeroMeanUnitVarianceKwargs",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "properties": {
    "id": {
      "default": "input",
      "description": "Input tensor id.\nNo duplicates are allowed across all inputs and outputs.",
      "maxLength": 32,
      "minLength": 1,
      "title": "TensorId",
      "type": "string"
    },
    "description": {
      "default": "",
      "description": "free text description",
      "maxLength": 128,
      "title": "Description",
      "type": "string"
    },
    "axes": {
      "description": "tensor axes",
      "items": {
        "discriminator": {
          "mapping": {
            "batch": "#/$defs/BatchAxis",
            "channel": "#/$defs/ChannelAxis",
            "index": "#/$defs/IndexInputAxis",
            "space": "#/$defs/SpaceInputAxis",
            "time": "#/$defs/TimeInputAxis"
          },
          "propertyName": "type"
        },
        "oneOf": [
          {
            "$ref": "#/$defs/BatchAxis"
          },
          {
            "$ref": "#/$defs/ChannelAxis"
          },
          {
            "$ref": "#/$defs/IndexInputAxis"
          },
          {
            "$ref": "#/$defs/TimeInputAxis"
          },
          {
            "$ref": "#/$defs/SpaceInputAxis"
          }
        ]
      },
      "minItems": 1,
      "title": "Axes",
      "type": "array"
    },
    "test_tensor": {
      "anyOf": [
        {
          "$ref": "#/$defs/FileDescr"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "An example tensor to use for testing.\nUsing the model with the test input tensors is expected to yield the test output tensors.\nEach test tensor has be a an ndarray in the\n[numpy.lib file format](https://numpy.org/doc/stable/reference/generated/numpy.lib.format.html#module-numpy.lib.format).\nThe file extension must be '.npy'."
    },
    "sample_tensor": {
      "anyOf": [
        {
          "$ref": "#/$defs/FileDescr"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "A sample tensor to illustrate a possible input/output for the model,\nThe sample image primarily serves to inform a human user about an example use case\nand is typically stored as .hdf5, .png or .tiff.\nIt has to be readable by the [imageio library](https://imageio.readthedocs.io/en/stable/formats/index.html#supported-formats)\n(numpy's `.npy` format is not supported).\nThe image dimensionality has to match the number of axes specified in this tensor description."
    },
    "data": {
      "anyOf": [
        {
          "$ref": "#/$defs/NominalOrOrdinalDataDescr"
        },
        {
          "$ref": "#/$defs/IntervalOrRatioDataDescr"
        },
        {
          "items": {
            "anyOf": [
              {
                "$ref": "#/$defs/NominalOrOrdinalDataDescr"
              },
              {
                "$ref": "#/$defs/IntervalOrRatioDataDescr"
              }
            ]
          },
          "minItems": 1,
          "type": "array"
        }
      ],
      "default": {
        "type": "float32",
        "range": [
          null,
          null
        ],
        "unit": "arbitrary unit",
        "scale": 1.0,
        "offset": null
      },
      "description": "Description of the tensor's data values, optionally per channel.\nIf specified per channel, the data `type` needs to match across channels.",
      "title": "Data"
    },
    "output_of": {
      "anyOf": [
        {
          "minLength": 1,
          "title": "ModelId",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "If this input tensor is the output of another model, specify the model id here.\nThis model's input id must match the output id of the referenced model.",
      "title": "Output Of"
    },
    "optional": {
      "default": false,
      "description": "indicates that this tensor may be `None`",
      "title": "Optional",
      "type": "boolean"
    },
    "pad": {
      "anyOf": [
        {
          "$ref": "#/$defs/ConstantPadding"
        },
        {
          "$ref": "#/$defs/EdgePadding"
        },
        {
          "$ref": "#/$defs/ReflectPadding"
        },
        {
          "$ref": "#/$defs/SymmetricPadding"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Explicitly specify how to pad this input tensor.\n\nUse `axes[i].pad` to specify padding width.\n\nNote:\n  Non-blockwise sample prediction only applies padding for axes with a `pad` specification.",
      "title": "Pad"
    },
    "preprocessing": {
      "description": "Description of how this input should be preprocessed.\n\nnotes:\n- If preprocessing does not start with an 'ensure_dtype' entry, it is added\n  to ensure an input tensor's data type matches the input tensor's data description.\n- If preprocessing does not end with an 'ensure_dtype' or 'binarize' entry, an\n  'ensure_dtype' step is added to ensure preprocessing steps are not unintentionally\n  changing the data type.",
      "items": {
        "discriminator": {
          "mapping": {
            "binarize": "#/$defs/BinarizeDescr",
            "clip": "#/$defs/ClipDescr",
            "ensure_dtype": "#/$defs/EnsureDtypeDescr",
            "fixed_zero_mean_unit_variance": "#/$defs/FixedZeroMeanUnitVarianceDescr",
            "scale_linear": "#/$defs/ScaleLinearDescr",
            "scale_range": "#/$defs/ScaleRangeDescr",
            "sigmoid": "#/$defs/SigmoidDescr",
            "softmax": "#/$defs/SoftmaxDescr",
            "zero_mean_unit_variance": "#/$defs/ZeroMeanUnitVarianceDescr"
          },
          "propertyName": "id"
        },
        "oneOf": [
          {
            "$ref": "#/$defs/BinarizeDescr"
          },
          {
            "$ref": "#/$defs/ClipDescr"
          },
          {
            "$ref": "#/$defs/EnsureDtypeDescr"
          },
          {
            "$ref": "#/$defs/FixedZeroMeanUnitVarianceDescr"
          },
          {
            "$ref": "#/$defs/ScaleLinearDescr"
          },
          {
            "$ref": "#/$defs/ScaleRangeDescr"
          },
          {
            "$ref": "#/$defs/SigmoidDescr"
          },
          {
            "$ref": "#/$defs/SoftmaxDescr"
          },
          {
            "$ref": "#/$defs/ZeroMeanUnitVarianceDescr"
          }
        ]
      },
      "title": "Preprocessing",
      "type": "array"
    }
  },
  "required": [
    "axes"
  ],
  "title": "model.v0_5.InputTensorDescr",
  "type": "object"
}

Fields:

• description (Annotated[str, MaxLen(128)])
• axes (NotEmpty[Sequence[IO_AxisT]])
• test_tensor (FAIR[FileDescr_package | None])
• sample_tensor (FAIR[FileDescr_package | None])
• data (TensorDataDescr | NotEmpty[Sequence[TensorDataDescr]])
• id (TensorId)
• output_of (ModelId | None)
• optional (bool)
• pad (Padding | None)
• preprocessing (list[PreprocessingDescr])

Validators:

• _validate_axes → axes
• _validate_sample_tensor
• _check_data_type_across_channels → data
• _check_data_matches_channelaxis
• _validate_output_of
• _validate_preprocessing_kwargs

axes pydantic-field

axes: NotEmpty[Sequence[IO_AxisT]]

tensor axes

data pydantic-field

data: TensorDataDescr | NotEmpty[Sequence[TensorDataDescr]]

Description of the tensor's data values, optionally per channel. If specified per channel, the data type needs to match across channels.

description pydantic-field

description: Annotated[str, MaxLen(128)] = ''

free text description

dtype property

dtype: Literal[
    "float32",
    "float64",
    "uint8",
    "int8",
    "uint16",
    "int16",
    "uint32",
    "int32",
    "uint64",
    "int64",
    "bool",
]

dtype as specified under data.type or data[i].type

id pydantic-field

id: TensorId

Input tensor id. No duplicates are allowed across all inputs and outputs.

optional pydantic-field

optional: bool = False

indicates that this tensor may be None

output_of pydantic-field

output_of: ModelId | None = None

If this input tensor is the output of another model, specify the model id here. This model's input id must match the output id of the referenced model.

pad pydantic-field

pad: Padding | None = None

Explicitly specify how to pad this input tensor.

Use axes[i].pad to specify padding width.

Note

Non-blockwise sample prediction only applies padding for axes with a pad specification.

preprocessing pydantic-field

preprocessing: list[PreprocessingDescr]

Description of how this input should be preprocessed.

notes: - If preprocessing does not start with an 'ensure_dtype' entry, it is added to ensure an input tensor's data type matches the input tensor's data description. - If preprocessing does not end with an 'ensure_dtype' or 'binarize' entry, an 'ensure_dtype' step is added to ensure preprocessing steps are not unintentionally changing the data type.

sample_tensor pydantic-field

sample_tensor: FAIR[FileDescr_package | None] = None

A sample tensor to illustrate a possible input/output for the model, The sample image primarily serves to inform a human user about an example use case and is typically stored as .hdf5, .png or .tiff. It has to be readable by the imageio library (numpy's .npy format is not supported). The image dimensionality has to match the number of axes specified in this tensor description.

shape property

shape

test_tensor pydantic-field

test_tensor: FAIR[FileDescr_package | None] = None

An example tensor to use for testing. Using the model with the test input tensors is expected to yield the test output tensors. Each test tensor has be a an ndarray in the numpy.lib file format. The file extension must be '.npy'.

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get_axis_sizes_for_array

get_axis_sizes_for_array(
    array: NDArray[Any],
) -> dict[AxisId, int]

Source code in src/bioimageio/spec/model/v0_5.py

def get_axis_sizes_for_array(self, array: NDArray[Any]) -> dict[AxisId, int]:
    if len(array.shape) != len(self.axes):
        raise ValueError(
            f"Dimension mismatch: array shape {array.shape} (#{len(array.shape)})"
            + f" incompatible with {len(self.axes)} axes."
        )
    return {a.id: array.shape[i] for i, a in enumerate(self.axes)}

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

IntervalOrRatioDataDescr pydantic-model

Bases: Node

Show JSON schema:

{
  "additionalProperties": false,
  "properties": {
    "type": {
      "default": "float32",
      "enum": [
        "float32",
        "float64",
        "uint8",
        "int8",
        "uint16",
        "int16",
        "uint32",
        "int32",
        "uint64",
        "int64"
      ],
      "examples": [
        "float32",
        "float64",
        "uint8",
        "uint16"
      ],
      "title": "Type",
      "type": "string"
    },
    "range": {
      "default": [
        null,
        null
      ],
      "description": "Tuple `(minimum, maximum)` specifying the allowed range of the data in this tensor.\n`None` corresponds to min/max of what can be expressed by **type**.",
      "maxItems": 2,
      "minItems": 2,
      "prefixItems": [
        {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ]
        },
        {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ]
        }
      ],
      "title": "Range",
      "type": "array"
    },
    "unit": {
      "anyOf": [
        {
          "const": "arbitrary unit",
          "type": "string"
        },
        {
          "description": "An SI unit",
          "minLength": 1,
          "pattern": "^(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^[+-]?[1-9]\\d*)?((\u00b7(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^[+-]?[1-9]\\d*)?)|(/(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^+?[1-9]\\d*)?))*$",
          "title": "SiUnit",
          "type": "string"
        }
      ],
      "default": "arbitrary unit",
      "title": "Unit"
    },
    "scale": {
      "default": 1.0,
      "description": "Scale for data on an interval (or ratio) scale.",
      "title": "Scale",
      "type": "number"
    },
    "offset": {
      "anyOf": [
        {
          "type": "number"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Offset for data on a ratio scale.",
      "title": "Offset"
    }
  },
  "title": "model.v0_5.IntervalOrRatioDataDescr",
  "type": "object"
}

Fields:

• type (Annotated[IntervalOrRatioDType, Field(examples=['float32', 'float64', 'uint8', 'uint16'])])
• range (tuple[float | None, float | None])
• unit (Literal['arbitrary unit'] | SiUnit)
• scale (float)
• offset (float | None)

Validators:

• _replace_inf

offset pydantic-field

offset: float | None = None

Offset for data on a ratio scale.

range pydantic-field

range: tuple[float | None, float | None] = (None, None)

Tuple (minimum, maximum) specifying the allowed range of the data in this tensor. None corresponds to min/max of what can be expressed by type.

scale pydantic-field

scale: float = 1.0

Scale for data on an interval (or ratio) scale.

type pydantic-field

type: Annotated[
    IntervalOrRatioDType,
    Field(
        examples=["float32", "float64", "uint8", "uint16"]
    ),
] = "float32"

unit pydantic-field

unit: Literal['arbitrary unit'] | SiUnit = 'arbitrary unit'

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

KerasHdf5WeightsDescr pydantic-model

Bases: WeightsEntryDescrBase

Show JSON schema:

{
  "$defs": {
    "Author": {
      "additionalProperties": false,
      "properties": {
        "affiliation": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Affiliation",
          "title": "Affiliation"
        },
        "email": {
          "anyOf": [
            {
              "format": "email",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Email",
          "title": "Email"
        },
        "orcid": {
          "anyOf": [
            {
              "description": "An ORCID identifier, see https://orcid.org/",
              "title": "OrcidId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "An [ORCID iD](https://support.orcid.org/hc/en-us/sections/360001495313-What-is-ORCID\n) in hyphenated groups of 4 digits, (and [valid](\nhttps://support.orcid.org/hc/en-us/articles/360006897674-Structure-of-the-ORCID-Identifier\n) as per ISO 7064 11,2.)",
          "examples": [
            "0000-0001-2345-6789"
          ],
          "title": "Orcid"
        },
        "name": {
          "title": "Name",
          "type": "string"
        },
        "github_user": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Github User"
        }
      },
      "required": [
        "name"
      ],
      "title": "generic.v0_3.Author",
      "type": "object"
    },
    "RelativeFilePath": {
      "description": "A path relative to the `rdf.yaml` file (also if the RDF source is a URL).",
      "format": "path",
      "title": "RelativeFilePath",
      "type": "string"
    },
    "Version": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        }
      ],
      "description": "wraps a packaging.version.Version instance for validation in pydantic models",
      "title": "Version"
    }
  },
  "additionalProperties": false,
  "properties": {
    "source": {
      "anyOf": [
        {
          "description": "A URL with the HTTP or HTTPS scheme.",
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "title": "HttpUrl",
          "type": "string"
        },
        {
          "$ref": "#/$defs/RelativeFilePath"
        },
        {
          "format": "file-path",
          "title": "FilePath",
          "type": "string"
        }
      ],
      "description": "Source of the weights file.",
      "title": "FileSource"
    },
    "sha256": {
      "anyOf": [
        {
          "description": "A SHA-256 hash value",
          "maxLength": 64,
          "minLength": 64,
          "title": "Sha256",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "SHA256 hash value of the **source** file.",
      "title": "Sha256"
    },
    "authors": {
      "anyOf": [
        {
          "items": {
            "$ref": "#/$defs/Author"
          },
          "type": "array"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
      "title": "Authors"
    },
    "parent": {
      "anyOf": [
        {
          "enum": [
            "keras_hdf5",
            "keras_v3",
            "onnx",
            "pytorch_state_dict",
            "tensorflow_js",
            "tensorflow_saved_model_bundle",
            "torchscript"
          ],
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
      "examples": [
        "pytorch_state_dict"
      ],
      "title": "Parent"
    },
    "comment": {
      "default": "",
      "description": "A comment about this weights entry, for example how these weights were created.",
      "title": "Comment",
      "type": "string"
    },
    "tensorflow_version": {
      "$ref": "#/$defs/Version",
      "description": "TensorFlow version used to create these weights."
    }
  },
  "required": [
    "source",
    "tensorflow_version"
  ],
  "title": "model.v0_5.KerasHdf5WeightsDescr",
  "type": "object"
}

Fields:

• source (Annotated[FileSource, AfterValidator(wo_special_file_name)])
• sha256 (Sha256 | None)
• authors (list[Author] | None)
• parent (Annotated[WeightsFormat | None, Field(examples=['pytorch_state_dict'])])
• comment (str)
• tensorflow_version (Version)

Validators:

• _validate

authors pydantic-field

authors: list[Author] | None = None

Authors Either the person(s) that have trained this model resulting in the original weights file. (If this is the initial weights entry, i.e. it does not have a parent) Or the person(s) who have converted the weights to this weights format. (If this is a child weight, i.e. it has a parent field)

comment pydantic-field

comment: str = ''

A comment about this weights entry, for example how these weights were created.

parent pydantic-field

parent: Annotated[
    WeightsFormat | None,
    Field(examples=["pytorch_state_dict"]),
] = None

The source weights these weights were converted from. For example, if a model's weights were converted from the pytorch_state_dict format to torchscript, The pytorch_state_dict weights entry has no parent and is the parent of the torchscript weights. All weight entries except one (the initial set of weights resulting from training the model), need to have this field.

sha256 pydantic-field

sha256: Sha256 | None = None

SHA256 hash value of the source file.

source pydantic-field

source: Annotated[
    FileSource, AfterValidator(wo_special_file_name)
]

Source of the weights file.

suffix property

suffix: str

tensorflow_version pydantic-field

tensorflow_version: Version

TensorFlow version used to create these weights.

type class-attribute

type: WeightsFormat = 'keras_hdf5'

weights_format_name class-attribute

weights_format_name: str = 'Keras HDF5'

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

download

download(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

alias for .get_reader

Source code in src/bioimageio/spec/_internal/io.py

def download(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """alias for `.get_reader`"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

get_reader

get_reader(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

open the file source (download if needed)

Source code in src/bioimageio/spec/_internal/io.py

def get_reader(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """open the file source (download if needed)"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

validate_sha256

validate_sha256(force_recompute: bool = False) -> None

validate the sha256 hash value of the source file

Source code in src/bioimageio/spec/_internal/io.py

def validate_sha256(self, force_recompute: bool = False) -> None:
    """validate the sha256 hash value of the **source** file"""
    context = get_validation_context()
    src_str = str(self.source)
    if force_recompute:
        actual_sha = None
    else:
        actual_sha = context.known_files.get(src_str)

    if actual_sha is None:
        if context.perform_io_checks or force_recompute:
            reader = get_reader(self.source, sha256=self.sha256)
            if force_recompute:
                actual_sha = get_sha256(reader)
            else:
                actual_sha = reader.sha256

            context.known_files[src_str] = actual_sha
        elif context.known_files and src_str not in context.known_files:
            # perform_io_checks is False, but known files were given,
            # so we expect all file references to be in there
            raise ValueError(f"File {src_str} not found in `known_files`.")

    if actual_sha is None or self.sha256 == actual_sha:
        return
    elif self.sha256 is None or context.update_hashes:
        self.sha256 = actual_sha
    elif self.sha256 != actual_sha:
        raise ValueError(
            f"Sha256 mismatch for {self.source}. Expected {self.sha256}, got "
            + f"{actual_sha}. Update expected `sha256` or point to the matching "
            + "file."
        )

KerasV3WeightsDescr pydantic-model

Bases: WeightsEntryDescrBase

Show JSON schema:

{
  "$defs": {
    "Author": {
      "additionalProperties": false,
      "properties": {
        "affiliation": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Affiliation",
          "title": "Affiliation"
        },
        "email": {
          "anyOf": [
            {
              "format": "email",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Email",
          "title": "Email"
        },
        "orcid": {
          "anyOf": [
            {
              "description": "An ORCID identifier, see https://orcid.org/",
              "title": "OrcidId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "An [ORCID iD](https://support.orcid.org/hc/en-us/sections/360001495313-What-is-ORCID\n) in hyphenated groups of 4 digits, (and [valid](\nhttps://support.orcid.org/hc/en-us/articles/360006897674-Structure-of-the-ORCID-Identifier\n) as per ISO 7064 11,2.)",
          "examples": [
            "0000-0001-2345-6789"
          ],
          "title": "Orcid"
        },
        "name": {
          "title": "Name",
          "type": "string"
        },
        "github_user": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Github User"
        }
      },
      "required": [
        "name"
      ],
      "title": "generic.v0_3.Author",
      "type": "object"
    },
    "RelativeFilePath": {
      "description": "A path relative to the `rdf.yaml` file (also if the RDF source is a URL).",
      "format": "path",
      "title": "RelativeFilePath",
      "type": "string"
    },
    "Version": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        }
      ],
      "description": "wraps a packaging.version.Version instance for validation in pydantic models",
      "title": "Version"
    }
  },
  "additionalProperties": false,
  "properties": {
    "source": {
      "anyOf": [
        {
          "description": "A URL with the HTTP or HTTPS scheme.",
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "title": "HttpUrl",
          "type": "string"
        },
        {
          "$ref": "#/$defs/RelativeFilePath"
        },
        {
          "format": "file-path",
          "title": "FilePath",
          "type": "string"
        }
      ],
      "description": "Source of the .keras weights file.",
      "title": "FileSource"
    },
    "sha256": {
      "anyOf": [
        {
          "description": "A SHA-256 hash value",
          "maxLength": 64,
          "minLength": 64,
          "title": "Sha256",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "SHA256 hash value of the **source** file.",
      "title": "Sha256"
    },
    "authors": {
      "anyOf": [
        {
          "items": {
            "$ref": "#/$defs/Author"
          },
          "type": "array"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
      "title": "Authors"
    },
    "parent": {
      "anyOf": [
        {
          "enum": [
            "keras_hdf5",
            "keras_v3",
            "onnx",
            "pytorch_state_dict",
            "tensorflow_js",
            "tensorflow_saved_model_bundle",
            "torchscript"
          ],
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
      "examples": [
        "pytorch_state_dict"
      ],
      "title": "Parent"
    },
    "comment": {
      "default": "",
      "description": "A comment about this weights entry, for example how these weights were created.",
      "title": "Comment",
      "type": "string"
    },
    "keras_version": {
      "$ref": "#/$defs/Version",
      "description": "Keras version used to create these weights.",
      "ge": 3
    },
    "backend": {
      "description": "Keras backend used to create these weights.",
      "maxItems": 2,
      "minItems": 2,
      "prefixItems": [
        {
          "enum": [
            "tensorflow",
            "jax",
            "torch"
          ],
          "type": "string"
        },
        {
          "$ref": "#/$defs/Version"
        }
      ],
      "title": "Backend",
      "type": "array"
    }
  },
  "required": [
    "source",
    "keras_version",
    "backend"
  ],
  "title": "model.v0_5.KerasV3WeightsDescr",
  "type": "object"
}

Fields:

• sha256 (Sha256 | None)
• authors (list[Author] | None)
• parent (Annotated[WeightsFormat | None, Field(examples=['pytorch_state_dict'])])
• comment (str)
• keras_version (Annotated[Version, Ge(Version(3))])
• backend (tuple[Literal['tensorflow', 'jax', 'torch'], Version])
• source (Annotated[FileSource, AfterValidator(wo_special_file_name), WithSuffix('.keras', case_sensitive=True)])

Validators:

• _validate

authors pydantic-field

authors: list[Author] | None = None

Authors Either the person(s) that have trained this model resulting in the original weights file. (If this is the initial weights entry, i.e. it does not have a parent) Or the person(s) who have converted the weights to this weights format. (If this is a child weight, i.e. it has a parent field)

backend pydantic-field

backend: tuple[
    Literal["tensorflow", "jax", "torch"], Version
]

Keras backend used to create these weights.

comment pydantic-field

comment: str = ''

A comment about this weights entry, for example how these weights were created.

keras_version pydantic-field

keras_version: Annotated[Version, Ge(Version(3))]

Keras version used to create these weights.

parent pydantic-field

parent: Annotated[
    WeightsFormat | None,
    Field(examples=["pytorch_state_dict"]),
] = None

The source weights these weights were converted from. For example, if a model's weights were converted from the pytorch_state_dict format to torchscript, The pytorch_state_dict weights entry has no parent and is the parent of the torchscript weights. All weight entries except one (the initial set of weights resulting from training the model), need to have this field.

sha256 pydantic-field

sha256: Sha256 | None = None

SHA256 hash value of the source file.

source pydantic-field

source: Annotated[
    FileSource,
    AfterValidator(wo_special_file_name),
    WithSuffix(".keras", case_sensitive=True),
]

Source of the .keras weights file.

suffix property

suffix: str

type class-attribute

type: WeightsFormat = 'keras_v3'

weights_format_name class-attribute

weights_format_name: str = 'Keras v3'

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

download

download(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

alias for .get_reader

Source code in src/bioimageio/spec/_internal/io.py

def download(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """alias for `.get_reader`"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

get_reader

get_reader(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

open the file source (download if needed)

Source code in src/bioimageio/spec/_internal/io.py

def get_reader(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """open the file source (download if needed)"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

validate_sha256

validate_sha256(force_recompute: bool = False) -> None

validate the sha256 hash value of the source file

Source code in src/bioimageio/spec/_internal/io.py

def validate_sha256(self, force_recompute: bool = False) -> None:
    """validate the sha256 hash value of the **source** file"""
    context = get_validation_context()
    src_str = str(self.source)
    if force_recompute:
        actual_sha = None
    else:
        actual_sha = context.known_files.get(src_str)

    if actual_sha is None:
        if context.perform_io_checks or force_recompute:
            reader = get_reader(self.source, sha256=self.sha256)
            if force_recompute:
                actual_sha = get_sha256(reader)
            else:
                actual_sha = reader.sha256

            context.known_files[src_str] = actual_sha
        elif context.known_files and src_str not in context.known_files:
            # perform_io_checks is False, but known files were given,
            # so we expect all file references to be in there
            raise ValueError(f"File {src_str} not found in `known_files`.")

    if actual_sha is None or self.sha256 == actual_sha:
        return
    elif self.sha256 is None or context.update_hashes:
        self.sha256 = actual_sha
    elif self.sha256 != actual_sha:
        raise ValueError(
            f"Sha256 mismatch for {self.source}. Expected {self.sha256}, got "
            + f"{actual_sha}. Update expected `sha256` or point to the matching "
            + "file."
        )

LicenseId

Bases: ValidatedString

              flowchart TD
              bioimageio.spec.model.v0_5.LicenseId[LicenseId]
              bioimageio.spec._internal.validated_string.ValidatedString[ValidatedString]

                              bioimageio.spec._internal.validated_string.ValidatedString --> bioimageio.spec.model.v0_5.LicenseId
                


              click bioimageio.spec.model.v0_5.LicenseId href "" "bioimageio.spec.model.v0_5.LicenseId"
              click bioimageio.spec._internal.validated_string.ValidatedString href "" "bioimageio.spec._internal.validated_string.ValidatedString"
            

Methods:

Name	Description
__get_pydantic_core_schema__
__get_pydantic_json_schema__
__new__

Attributes:

Name	Type	Description
root_model	type[RootModel[Any]]	the pydantic root model to validate the string

root_model class-attribute

root_model: type[RootModel[Any]] = RootModel[
    LicenseIdLiteral
]

the pydantic root model to validate the string

__get_pydantic_core_schema__ classmethod

__get_pydantic_core_schema__(
    source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_core_schema__(
    cls, source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema:
    return no_info_after_validator_function(cls, handler(str))

__get_pydantic_json_schema__ classmethod

__get_pydantic_json_schema__(
    core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_json_schema__(
    cls, core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue:
    json_schema = cls.root_model.model_json_schema(mode=handler.mode)
    json_schema["title"] = cls.__name__.strip("_")
    if cls.__doc__:
        json_schema["description"] = cls.__doc__

    return json_schema

__new__

__new__(object: object)

Source code in src/bioimageio/spec/_internal/validated_string.py

def __new__(cls, object: object):
    _validated = cls.root_model.model_validate(object).root
    self = super().__new__(cls, _validated)
    self._validated = _validated
    return self._after_validator()

LinkedDataset pydantic-model

Bases: LinkedResourceBase

Reference to a bioimage.io dataset.

Show JSON schema:

{
  "$defs": {
    "Version": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        }
      ],
      "description": "wraps a packaging.version.Version instance for validation in pydantic models",
      "title": "Version"
    }
  },
  "additionalProperties": false,
  "description": "Reference to a bioimage.io dataset.",
  "properties": {
    "version": {
      "anyOf": [
        {
          "$ref": "#/$defs/Version"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The version of the linked resource following SemVer 2.0."
    },
    "id": {
      "description": "A valid dataset `id` from the bioimage.io collection.",
      "minLength": 1,
      "title": "DatasetId",
      "type": "string"
    }
  },
  "required": [
    "id"
  ],
  "title": "dataset.v0_3.LinkedDataset",
  "type": "object"
}

Fields:

• version (Version | None)
• id (DatasetId)

id pydantic-field

id: DatasetId

A valid dataset id from the bioimage.io collection.

version pydantic-field

version: Version | None = None

The version of the linked resource following SemVer 2.0.

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

LinkedModel pydantic-model

Bases: LinkedResourceBase

Reference to a bioimage.io model.

Show JSON schema:

{
  "$defs": {
    "Version": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        }
      ],
      "description": "wraps a packaging.version.Version instance for validation in pydantic models",
      "title": "Version"
    }
  },
  "additionalProperties": false,
  "description": "Reference to a bioimage.io model.",
  "properties": {
    "version": {
      "anyOf": [
        {
          "$ref": "#/$defs/Version"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The version of the linked resource following SemVer 2.0."
    },
    "id": {
      "description": "A valid model `id` from the bioimage.io collection.",
      "minLength": 1,
      "title": "ModelId",
      "type": "string"
    }
  },
  "required": [
    "id"
  ],
  "title": "model.v0_5.LinkedModel",
  "type": "object"
}

Fields:

• version (Version | None)
• id (ModelId)

id pydantic-field

id: ModelId

A valid model id from the bioimage.io collection.

version pydantic-field

version: Version | None = None

The version of the linked resource following SemVer 2.0.

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

LinkedResource pydantic-model

Bases: LinkedResourceBase

Reference to a bioimage.io resource

Show JSON schema:

{
  "$defs": {
    "Version": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        }
      ],
      "description": "wraps a packaging.version.Version instance for validation in pydantic models",
      "title": "Version"
    }
  },
  "additionalProperties": false,
  "description": "Reference to a bioimage.io resource",
  "properties": {
    "version": {
      "anyOf": [
        {
          "$ref": "#/$defs/Version"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The version of the linked resource following SemVer 2.0."
    },
    "id": {
      "description": "A valid resource `id` from the official bioimage.io collection.",
      "minLength": 1,
      "title": "ResourceId",
      "type": "string"
    }
  },
  "required": [
    "id"
  ],
  "title": "generic.v0_3.LinkedResource",
  "type": "object"
}

Fields:

• version (Version | None)
• id (ResourceId)

id pydantic-field

id: ResourceId

A valid resource id from the official bioimage.io collection.

version pydantic-field

version: Version | None = None

The version of the linked resource following SemVer 2.0.

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

Maintainer pydantic-model

Bases: _Maintainer_v0_2

Show JSON schema:

{
  "additionalProperties": false,
  "properties": {
    "affiliation": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Affiliation",
      "title": "Affiliation"
    },
    "email": {
      "anyOf": [
        {
          "format": "email",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Email",
      "title": "Email"
    },
    "orcid": {
      "anyOf": [
        {
          "description": "An ORCID identifier, see https://orcid.org/",
          "title": "OrcidId",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "An [ORCID iD](https://support.orcid.org/hc/en-us/sections/360001495313-What-is-ORCID\n) in hyphenated groups of 4 digits, (and [valid](\nhttps://support.orcid.org/hc/en-us/articles/360006897674-Structure-of-the-ORCID-Identifier\n) as per ISO 7064 11,2.)",
      "examples": [
        "0000-0001-2345-6789"
      ],
      "title": "Orcid"
    },
    "name": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Name"
    },
    "github_user": {
      "title": "Github User",
      "type": "string"
    }
  },
  "required": [
    "github_user"
  ],
  "title": "generic.v0_3.Maintainer",
  "type": "object"
}

Fields:

• affiliation (str | None)
• email (EmailStr | None)
• orcid (Annotated[OrcidId | None, Field(examples=['0000-0001-2345-6789'])])
• name (Annotated[str, Predicate(_has_no_slash)] | None)
• github_user (str)

Validators:

• validate_github_user → github_user

affiliation pydantic-field

affiliation: str | None = None

Affiliation

email pydantic-field

email: EmailStr | None = None

Email

github_user pydantic-field

github_user: str

name pydantic-field

name: Annotated[str, Predicate(_has_no_slash)] | None = None

orcid pydantic-field

orcid: Annotated[
    OrcidId | None, Field(examples=["0000-0001-2345-6789"])
] = None

An ORCID iD in hyphenated groups of 4 digits, (and valid as per ISO 7064 11,2.)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

validate_github_user pydantic-validator

validate_github_user(value: str)

Source code in src/bioimageio/spec/generic/v0_3.py

@field_validator("github_user", mode="after")
def validate_github_user(cls, value: str):
    return validate_github_user(value)

ModelDescr pydantic-model

Bases: GenericModelDescrBase

Specification of the fields used in a bioimage.io-compliant RDF to describe AI models with pretrained weights. These fields are typically stored in a YAML file which we call a model resource description file (model RDF).

Show JSON schema:

{
  "$defs": {
    "ArchitectureFromFileDescr": {
      "additionalProperties": false,
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "Architecture source file",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        },
        "callable": {
          "description": "Identifier of the callable that returns a torch.nn.Module instance.",
          "examples": [
            "MyNetworkClass",
            "get_my_model"
          ],
          "minLength": 1,
          "title": "Identifier",
          "type": "string"
        },
        "kwargs": {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "description": "key word arguments for the `callable`",
          "title": "Kwargs",
          "type": "object"
        }
      },
      "required": [
        "source",
        "callable"
      ],
      "title": "model.v0_5.ArchitectureFromFileDescr",
      "type": "object"
    },
    "ArchitectureFromLibraryDescr": {
      "additionalProperties": false,
      "properties": {
        "callable": {
          "description": "Identifier of the callable that returns a torch.nn.Module instance.",
          "examples": [
            "MyNetworkClass",
            "get_my_model"
          ],
          "minLength": 1,
          "title": "Identifier",
          "type": "string"
        },
        "kwargs": {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "description": "key word arguments for the `callable`",
          "title": "Kwargs",
          "type": "object"
        },
        "import_from": {
          "description": "Where to import the callable from, i.e. `from <import_from> import <callable>`",
          "title": "Import From",
          "type": "string"
        }
      },
      "required": [
        "callable",
        "import_from"
      ],
      "title": "model.v0_5.ArchitectureFromLibraryDescr",
      "type": "object"
    },
    "AttachmentsDescr": {
      "additionalProperties": true,
      "properties": {
        "files": {
          "description": "File attachments",
          "items": {
            "anyOf": [
              {
                "description": "A URL with the HTTP or HTTPS scheme.",
                "format": "uri",
                "maxLength": 2083,
                "minLength": 1,
                "title": "HttpUrl",
                "type": "string"
              },
              {
                "$ref": "#/$defs/RelativeFilePath"
              },
              {
                "format": "file-path",
                "title": "FilePath",
                "type": "string"
              }
            ],
            "title": "FileSource"
          },
          "title": "Files",
          "type": "array"
        }
      },
      "title": "generic.v0_2.AttachmentsDescr",
      "type": "object"
    },
    "BadgeDescr": {
      "additionalProperties": false,
      "description": "A custom badge",
      "properties": {
        "label": {
          "description": "badge label to display on hover",
          "examples": [
            "Open in Colab"
          ],
          "title": "Label",
          "type": "string"
        },
        "icon": {
          "anyOf": [
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "badge icon (included in bioimage.io package if not a URL)",
          "examples": [
            "https://colab.research.google.com/assets/colab-badge.svg"
          ],
          "title": "Icon"
        },
        "url": {
          "description": "target URL",
          "examples": [
            "https://colab.research.google.com/github/HenriquesLab/ZeroCostDL4Mic/blob/master/Colab_notebooks/U-net_2D_ZeroCostDL4Mic.ipynb"
          ],
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "title": "HttpUrl",
          "type": "string"
        }
      },
      "required": [
        "label",
        "url"
      ],
      "title": "generic.v0_2.BadgeDescr",
      "type": "object"
    },
    "BatchAxis": {
      "additionalProperties": false,
      "properties": {
        "id": {
          "default": "batch",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "batch",
          "title": "Type",
          "type": "string"
        },
        "size": {
          "anyOf": [
            {
              "const": 1,
              "type": "integer"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The batch size may be fixed to 1,\notherwise (the default) it may be chosen arbitrarily depending on available memory",
          "title": "Size"
        }
      },
      "required": [
        "type"
      ],
      "title": "model.v0_5.BatchAxis",
      "type": "object"
    },
    "BiasRisksLimitations": {
      "additionalProperties": true,
      "description": "Known biases, risks, technical limitations, and recommendations for model use.",
      "properties": {
        "known_biases": {
          "default": "In general bioimage models may suffer from biases caused by:\n\n- Imaging protocol dependencies\n- Use of a specific cell type\n- Species-specific training data limitations\n\n",
          "description": "Biases in training data or model behavior.",
          "title": "Known Biases",
          "type": "string"
        },
        "risks": {
          "default": "Common risks in bioimage analysis include:\n\n- Erroneously assuming generalization to unseen experimental conditions\n- Trusting (overconfident) model outputs without validation\n- Misinterpretation of results\n\n",
          "description": "Potential risks in the context of bioimage analysis.",
          "title": "Risks",
          "type": "string"
        },
        "limitations": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Technical limitations and failure modes.",
          "title": "Limitations"
        },
        "recommendations": {
          "default": "Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model.",
          "description": "Mitigation strategies regarding `known_biases`, `risks`, and `limitations`, as well as applicable best practices.\n\nConsider:\n- How to use a validation dataset?\n- How to manually validate?\n- Feasibility of domain adaptation for different experimental setups?",
          "title": "Recommendations",
          "type": "string"
        }
      },
      "title": "model.v0_5.BiasRisksLimitations",
      "type": "object"
    },
    "BinarizeAlongAxisKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [BinarizeDescr][]",
      "properties": {
        "threshold": {
          "description": "The fixed threshold values along `axis`",
          "items": {
            "type": "number"
          },
          "minItems": 1,
          "title": "Threshold",
          "type": "array"
        },
        "axis": {
          "description": "The `threshold` axis",
          "examples": [
            "channel"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        }
      },
      "required": [
        "threshold",
        "axis"
      ],
      "title": "model.v0_5.BinarizeAlongAxisKwargs",
      "type": "object"
    },
    "BinarizeDescr": {
      "additionalProperties": false,
      "description": "Binarize the tensor with a fixed threshold.\n\nValues above [BinarizeKwargs.threshold][]/[BinarizeAlongAxisKwargs.threshold][]\nwill be set to one, values below the threshold to zero.\n\nExamples:\n- in YAML\n    ```yaml\n    postprocessing:\n      - id: binarize\n        kwargs:\n          axis: 'channel'\n          threshold: [0.25, 0.5, 0.75]\n    ```\n- in Python:\n\n    >>> postprocessing = [BinarizeDescr(\n    ...   kwargs=BinarizeAlongAxisKwargs(\n    ...       axis=AxisId('channel'),\n    ...       threshold=[0.25, 0.5, 0.75],\n    ...   )\n    ... )]",
      "properties": {
        "id": {
          "const": "binarize",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "anyOf": [
            {
              "$ref": "#/$defs/BinarizeKwargs"
            },
            {
              "$ref": "#/$defs/BinarizeAlongAxisKwargs"
            }
          ],
          "title": "Kwargs"
        }
      },
      "required": [
        "id",
        "kwargs"
      ],
      "title": "model.v0_5.BinarizeDescr",
      "type": "object"
    },
    "BinarizeKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [BinarizeDescr][]",
      "properties": {
        "threshold": {
          "description": "The fixed threshold",
          "title": "Threshold",
          "type": "number"
        }
      },
      "required": [
        "threshold"
      ],
      "title": "model.v0_5.BinarizeKwargs",
      "type": "object"
    },
    "CellposeFlowDynamicsDescr": {
      "additionalProperties": false,
      "description": "Cellpose flow dynamics postprocessing as described in:\n- Carsen Stringer and Marius Pachitariu. [*Cellpose: a generalist algorithm for cellular segmentation*](https://www.nature.com/articles/s41592-020-01018-x). Nature Methods, 2021.\n\nNote: Only available if the `cellpose` package is installed.",
      "properties": {
        "id": {
          "const": "cellpose_flow_dynamics",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "$ref": "#/$defs/CellposeFlowDynamicsKwargs"
        }
      },
      "required": [
        "id",
        "kwargs"
      ],
      "title": "model.v0_5.CellposeFlowDynamicsDescr",
      "type": "object"
    },
    "CellposeFlowDynamicsKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [CellposeFlowDynamicsDescr][]",
      "properties": {
        "cellprob_threshold": {
          "title": "Cellprob Threshold",
          "type": "number"
        },
        "flow_threshold": {
          "title": "Flow Threshold",
          "type": "number"
        },
        "do_3D": {
          "title": "Do 3D",
          "type": "boolean"
        },
        "min_size": {
          "default": 15,
          "description": "Minimum size of objects to keep, in pixels. Default is 15, which is the default in Cellpose. Set to 0 to disable filtering by size.",
          "title": "Min Size",
          "type": "integer"
        },
        "output_dtype": {
          "default": "uint16",
          "enum": [
            "uint16",
            "uint32"
          ],
          "title": "Output Dtype",
          "type": "string"
        }
      },
      "required": [
        "cellprob_threshold",
        "flow_threshold",
        "do_3D"
      ],
      "title": "model.v0_5.CellposeFlowDynamicsKwargs",
      "type": "object"
    },
    "ChannelAxis": {
      "additionalProperties": false,
      "properties": {
        "id": {
          "default": "channel",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "channel",
          "title": "Type",
          "type": "string"
        },
        "channel_names": {
          "description": "Name/label for each channel. The number of channels is given by `len(channel_names)`.",
          "items": {
            "type": "string"
          },
          "minItems": 1,
          "title": "Channel Names",
          "type": "array"
        }
      },
      "required": [
        "type",
        "channel_names"
      ],
      "title": "model.v0_5.ChannelAxis",
      "type": "object"
    },
    "ClipDescr": {
      "additionalProperties": false,
      "description": "Set tensor values below min to min and above max to max.\n\nSee `ScaleRangeDescr` for examples.",
      "properties": {
        "id": {
          "const": "clip",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "$ref": "#/$defs/ClipKwargs"
        }
      },
      "required": [
        "id",
        "kwargs"
      ],
      "title": "model.v0_5.ClipDescr",
      "type": "object"
    },
    "ClipKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [ClipDescr][]",
      "properties": {
        "min": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Minimum value for clipping.\n\nExclusive with [min_percentile][]",
          "title": "Min"
        },
        "min_percentile": {
          "anyOf": [
            {
              "exclusiveMaximum": 100,
              "minimum": 0,
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Minimum percentile for clipping.\n\nExclusive with [min][].\n\nIn range [0, 100).",
          "title": "Min Percentile"
        },
        "max": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Maximum value for clipping.\n\nExclusive with `max_percentile`.",
          "title": "Max"
        },
        "max_percentile": {
          "anyOf": [
            {
              "exclusiveMinimum": 1,
              "maximum": 100,
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Maximum percentile for clipping.\n\nExclusive with `max`.\n\nIn range (1, 100].",
          "title": "Max Percentile"
        },
        "axes": {
          "anyOf": [
            {
              "items": {
                "maxLength": 16,
                "minLength": 1,
                "title": "AxisId",
                "type": "string"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The subset of axes to determine percentiles jointly,\n\ni.e. axes to reduce to compute min/max from `min_percentile`/`max_percentile`.\nFor example to clip 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x')\nresulting in a tensor of equal shape with clipped values per channel, specify `axes=('batch', 'x', 'y')`.\nTo clip samples independently, leave out the 'batch' axis.\n\nOnly valid if `min_percentile` and/or `max_percentile` are set.\n\nDefault: Compute percentiles over all axes jointly.",
          "examples": [
            [
              "batch",
              "x",
              "y"
            ]
          ],
          "title": "Axes"
        }
      },
      "title": "model.v0_5.ClipKwargs",
      "type": "object"
    },
    "ConstantPadding": {
      "additionalProperties": false,
      "properties": {
        "mode": {
          "const": "constant",
          "default": "constant",
          "title": "Mode",
          "type": "string"
        },
        "value": {
          "anyOf": [
            {
              "type": "integer"
            },
            {
              "type": "number"
            }
          ],
          "default": 0,
          "title": "Value"
        }
      },
      "title": "model.v0_5.ConstantPadding",
      "type": "object"
    },
    "ContinuousReproducibilityTolerance": {
      "additionalProperties": true,
      "properties": {
        "mismatched_elements_per_million": {
          "default": 100,
          "description": "Maximum number of mismatched elements/pixels per million to tolerate.",
          "maximum": 30000,
          "minimum": 0,
          "title": "Mismatched Elements Per Million",
          "type": "integer"
        },
        "weights_formats": {
          "default": [],
          "description": "Limits the weights formats these details apply to.",
          "items": {
            "enum": [
              "keras_hdf5",
              "keras_v3",
              "onnx",
              "pytorch_state_dict",
              "tensorflow_js",
              "tensorflow_saved_model_bundle",
              "torchscript"
            ],
            "type": "string"
          },
          "title": "Weights Formats",
          "type": "array"
        },
        "relative_tolerance": {
          "default": 0.001,
          "description": "Maximum relative tolerance of reproduced test tensor.\nNeeds to be 0 for discrete (integer and boolean) outputs.",
          "maximum": 0.01,
          "minimum": 0,
          "title": "Relative Tolerance",
          "type": "number"
        },
        "absolute_tolerance": {
          "default": 0.001,
          "description": "Maximum absolute tolerance of reproduced test tensor.\n\nThe `absolute_tolerance` may not be greater than 1% of the maximum absolute value of the test tensor\n(this serves to prevent false positives in the presence of very small numbers).\nNeeds to be 0 for discrete (integer and boolean) outputs.",
          "minimum": 0,
          "title": "Absolute Tolerance",
          "type": "number"
        },
        "output_ids": {
          "const": "any_continuous_output",
          "default": "any_continuous_output",
          "description": "Limits the output tensor IDs to all outputs of float data type.",
          "title": "Output Ids",
          "type": "string"
        }
      },
      "title": "model.v0_5.ContinuousReproducibilityTolerance",
      "type": "object"
    },
    "CustomProcessingDescr": {
      "additionalProperties": false,
      "description": "Custom (post)processing op \u2014 source file shipped inline with the model.\n\nSupports (post)processing that cannot be expressed by the built-in named\noperations (watershed, connected components, etc.)\nusing a simple Python callable interface.\n\nThe op is implemented in a ``.py`` file packaged alongside the model weights.\nTwo styles are supported:\n\n*Callable class* \u2014 kwargs go to ``__init__``, tensors arrive in ``__call__``:\n\n.. code-block:: python\n\n    # my_postprocess.py\n    import numpy as np\n\n    class my_postprocess:\n        def __init__(self, threshold: float = 0.5) -> None:\n            self.threshold = threshold\n        def __call__(self, *arrays: np.ndarray) -> np.ndarray:\n            # arrays = model output tensors in rdf.yaml declaration order\n            return (arrays[0] > self.threshold).astype(np.uint8)\n\n*Factory function* \u2014 alternative closure style, identical runtime behaviour:\n\n.. code-block:: python\n\n    # my_postprocess.py\n    import numpy as np\n\n    def my_postprocess(threshold: float = 0.5):\n        def run(*arrays: np.ndarray) -> np.ndarray:\n            return (arrays[0] > threshold).astype(np.uint8)\n        return run\n\nReference it in ``rdf.yaml`` with the source file included in the package:\n\n.. code-block:: yaml\n\n    postprocessing:\n      - id: custom\n        callable: my_postprocess   # class or function name in source\n        source: my_postprocess.py  # packaged alongside weights\n        sha256: <hash>             # sha256 of the source file\n        kwargs:                    # forwarded to __init__ / factory\n          threshold: 0.5\n\n**Security:** source files are SHA-256 verified before execution.\nExecution requires explicit opt-in in bioimageio.core and curator\nreview before Zoo publication.",
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "Python source file (included when packaging the model).",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        },
        "id": {
          "const": "custom",
          "title": "Id",
          "type": "string"
        },
        "callable": {
          "description": "Name of the callable class or factory function defined in ``source``.\n\nAt runtime: ``op = callable(**kwargs)``, then ``result = op(*output_tensors)``\nper image.  Both a class with ``__call__`` and a factory function returning\na callable satisfy this protocol.",
          "examples": [
            "my_postprocess_factory",
            "MyPostprocessClass"
          ],
          "title": "Callable",
          "type": "string"
        },
        "kwargs": {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "description": "Keyword arguments forwarded to the callable (``__init__`` or factory).",
          "title": "Kwargs",
          "type": "object"
        }
      },
      "required": [
        "source",
        "id",
        "callable"
      ],
      "title": "model.v0_5.CustomProcessingDescr",
      "type": "object"
    },
    "DataDependentSize": {
      "additionalProperties": false,
      "properties": {
        "min": {
          "default": 1,
          "exclusiveMinimum": 0,
          "title": "Min",
          "type": "integer"
        },
        "max": {
          "anyOf": [
            {
              "exclusiveMinimum": 1,
              "type": "integer"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Max"
        }
      },
      "title": "model.v0_5.DataDependentSize",
      "type": "object"
    },
    "Datetime": {
      "description": "Timestamp in [ISO 8601](#https://en.wikipedia.org/wiki/ISO_8601) format\nwith a few restrictions listed [here](https://docs.python.org/3/library/datetime.html#datetime.datetime.fromisoformat).",
      "format": "date-time",
      "title": "Datetime",
      "type": "string"
    },
    "DiscreteReproducibilityTolerance": {
      "additionalProperties": true,
      "properties": {
        "mismatched_elements_per_million": {
          "default": 1000,
          "description": "Maximum number of mismatched elements/pixels per million to tolerate.\n\nNote: Increased default mismatched elements per million compared to\n[`ReproducibilityTolerance`][ReproducibilityTolerance]/[`ContinuousReproducibilityTolerance`][ContinuousReproducibilityTolerance],\nsince a single pixel mismatch can be more significant than for continuous outputs with numeric tolerance.",
          "maximum": 30000,
          "minimum": 0,
          "title": "Mismatched Elements Per Million",
          "type": "integer"
        },
        "weights_formats": {
          "default": [],
          "description": "Limits the weights formats these details apply to.",
          "items": {
            "enum": [
              "keras_hdf5",
              "keras_v3",
              "onnx",
              "pytorch_state_dict",
              "tensorflow_js",
              "tensorflow_saved_model_bundle",
              "torchscript"
            ],
            "type": "string"
          },
          "title": "Weights Formats",
          "type": "array"
        },
        "relative_tolerance": {
          "const": 0,
          "default": 0,
          "description": "Relative tolerance does not apply to discrete outputs",
          "title": "Relative Tolerance",
          "type": "integer"
        },
        "absolute_tolerance": {
          "const": 0,
          "default": 0,
          "description": "Absolute tolerance does not apply to discrete outputs",
          "title": "Absolute Tolerance",
          "type": "integer"
        },
        "output_ids": {
          "const": "any_discrete_output",
          "default": "any_discrete_output",
          "description": "Limits the output tensor IDs to all outputs of integer or boolean data type.",
          "title": "Output Ids",
          "type": "string"
        }
      },
      "title": "model.v0_5.DiscreteReproducibilityTolerance",
      "type": "object"
    },
    "EdgePadding": {
      "additionalProperties": false,
      "properties": {
        "mode": {
          "const": "edge",
          "default": "edge",
          "title": "Mode",
          "type": "string"
        }
      },
      "title": "model.v0_5.EdgePadding",
      "type": "object"
    },
    "EnsureDtypeDescr": {
      "additionalProperties": false,
      "description": "Cast the tensor data type to `EnsureDtypeKwargs.dtype` (if not matching).\n\nThis can for example be used to ensure the inner neural network model gets a\ndifferent input tensor data type than the fully described bioimage.io model does.\n\nExamples:\n    The described bioimage.io model (incl. preprocessing) accepts any\n    float32-compatible tensor, normalizes it with percentiles and clipping and then\n    casts it to uint8, which is what the neural network in this example expects.\n    - in YAML\n        ```yaml\n        inputs:\n        - data:\n            type: float32  # described bioimage.io model is compatible with any float32 input tensor\n          preprocessing:\n          - id: scale_range\n              kwargs:\n              axes: ['y', 'x']\n              max_percentile: 99.8\n              min_percentile: 5.0\n          - id: clip\n              kwargs:\n              min: 0.0\n              max: 1.0\n          - id: ensure_dtype  # the neural network of the model requires uint8\n              kwargs:\n              dtype: uint8\n        ```\n    - in Python:\n        >>> preprocessing = [\n        ...     ScaleRangeDescr(\n        ...         kwargs=ScaleRangeKwargs(\n        ...           axes= (AxisId('y'), AxisId('x')),\n        ...           max_percentile= 99.8,\n        ...           min_percentile= 5.0,\n        ...         )\n        ...     ),\n        ...     ClipDescr(kwargs=ClipKwargs(min=0.0, max=1.0)),\n        ...     EnsureDtypeDescr(kwargs=EnsureDtypeKwargs(dtype=\"uint8\")),\n        ... ]",
      "properties": {
        "id": {
          "const": "ensure_dtype",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "$ref": "#/$defs/EnsureDtypeKwargs"
        }
      },
      "required": [
        "id",
        "kwargs"
      ],
      "title": "model.v0_5.EnsureDtypeDescr",
      "type": "object"
    },
    "EnsureDtypeKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [EnsureDtypeDescr][]",
      "properties": {
        "dtype": {
          "enum": [
            "float32",
            "float64",
            "uint8",
            "int8",
            "uint16",
            "int16",
            "uint32",
            "int32",
            "uint64",
            "int64",
            "bool"
          ],
          "title": "Dtype",
          "type": "string"
        }
      },
      "required": [
        "dtype"
      ],
      "title": "model.v0_5.EnsureDtypeKwargs",
      "type": "object"
    },
    "EnvironmentalImpact": {
      "additionalProperties": true,
      "description": "Environmental considerations for model training and deployment.\n\nCarbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).",
      "properties": {
        "hardware_type": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "GPU/CPU specifications",
          "title": "Hardware Type"
        },
        "hours_used": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Total compute hours",
          "title": "Hours Used"
        },
        "cloud_provider": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "If applicable",
          "title": "Cloud Provider"
        },
        "compute_region": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Geographic location",
          "title": "Compute Region"
        },
        "co2_emitted": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "kg CO2 equivalent\n\nCarbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).",
          "title": "Co2 Emitted"
        }
      },
      "title": "model.v0_5.EnvironmentalImpact",
      "type": "object"
    },
    "Evaluation": {
      "additionalProperties": true,
      "properties": {
        "model_id": {
          "anyOf": [
            {
              "minLength": 1,
              "title": "ModelId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Model being evaluated.",
          "title": "Model Id"
        },
        "dataset_id": {
          "description": "Dataset used for evaluation.",
          "minLength": 1,
          "title": "DatasetId",
          "type": "string"
        },
        "dataset_source": {
          "description": "Source of the dataset.",
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "title": "HttpUrl",
          "type": "string"
        },
        "dataset_role": {
          "description": "Role of the dataset used for evaluation.\n\n- `train`: dataset was (part of) the training data\n- `validation`: dataset was (part of) the validation data used during training, e.g. used for model selection or hyperparameter tuning\n- `test`: dataset was (part of) the designated test data; not used during training or validation, but acquired from the same source/distribution as training data\n- `independent`: dataset is entirely independent test data; not used during training or validation, and acquired from a different source/distribution than training data\n- `unknown`: role of the dataset is unknown; choose this if you are not certain if (a subset) of the data was seen by the model during training.",
          "enum": [
            "train",
            "validation",
            "test",
            "independent",
            "unknown"
          ],
          "title": "Dataset Role",
          "type": "string"
        },
        "sample_count": {
          "description": "Number of evaluated samples.",
          "title": "Sample Count",
          "type": "integer"
        },
        "evaluation_factors": {
          "description": "(Abbreviations of) each evaluation factor.\n\nEvaluation factors are criteria along which model performance is evaluated, e.g. different image conditions\nlike 'low SNR', 'high cell density', or different biological conditions like 'cell type A', 'cell type B'.\nAn 'overall' factor may be included to summarize performance across all conditions.",
          "items": {
            "maxLength": 16,
            "type": "string"
          },
          "title": "Evaluation Factors",
          "type": "array"
        },
        "evaluation_factors_long": {
          "description": "Descriptions (long form) of each evaluation factor.",
          "items": {
            "type": "string"
          },
          "title": "Evaluation Factors Long",
          "type": "array"
        },
        "metrics": {
          "description": "(Abbreviations of) metrics used for evaluation.",
          "items": {
            "maxLength": 16,
            "type": "string"
          },
          "title": "Metrics",
          "type": "array"
        },
        "metrics_long": {
          "description": "Description of each metric used.",
          "items": {
            "type": "string"
          },
          "title": "Metrics Long",
          "type": "array"
        },
        "results": {
          "description": "Results for each metric (rows; outer list) and each evaluation factor (columns; inner list).",
          "items": {
            "items": {
              "anyOf": [
                {
                  "type": "string"
                },
                {
                  "type": "number"
                },
                {
                  "type": "integer"
                }
              ]
            },
            "type": "array"
          },
          "title": "Results",
          "type": "array"
        },
        "results_summary": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Interpretation of results for general audience.\n\nConsider:\n    - Overall model performance\n    - Comparison to existing methods\n    - Limitations and areas for improvement",
          "title": "Results Summary"
        }
      },
      "required": [
        "dataset_id",
        "dataset_source",
        "dataset_role",
        "sample_count",
        "evaluation_factors",
        "evaluation_factors_long",
        "metrics",
        "metrics_long",
        "results"
      ],
      "title": "model.v0_5.Evaluation",
      "type": "object"
    },
    "FileDescr": {
      "additionalProperties": false,
      "description": "A file description",
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "File source",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        }
      },
      "required": [
        "source"
      ],
      "title": "_internal.io.FileDescr",
      "type": "object"
    },
    "FixedZeroMeanUnitVarianceAlongAxisKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [FixedZeroMeanUnitVarianceDescr][]",
      "properties": {
        "mean": {
          "description": "The mean value(s) to normalize with.",
          "items": {
            "type": "number"
          },
          "minItems": 1,
          "title": "Mean",
          "type": "array"
        },
        "std": {
          "description": "The standard deviation value(s) to normalize with.\nSize must match `mean` values.",
          "items": {
            "minimum": 1e-06,
            "type": "number"
          },
          "minItems": 1,
          "title": "Std",
          "type": "array"
        },
        "axis": {
          "description": "The axis of the mean/std values to normalize each entry along that dimension\nseparately.",
          "examples": [
            "channel",
            "index"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        }
      },
      "required": [
        "mean",
        "std",
        "axis"
      ],
      "title": "model.v0_5.FixedZeroMeanUnitVarianceAlongAxisKwargs",
      "type": "object"
    },
    "FixedZeroMeanUnitVarianceDescr": {
      "additionalProperties": false,
      "description": "Subtract a given mean and divide by the standard deviation.\n\nNormalize with fixed, precomputed values for\n`FixedZeroMeanUnitVarianceKwargs.mean` and `FixedZeroMeanUnitVarianceKwargs.std`\nUse `FixedZeroMeanUnitVarianceAlongAxisKwargs` for independent scaling along given\naxes.\n\nExamples:\n1. scalar value for whole tensor\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: fixed_zero_mean_unit_variance\n        kwargs:\n          mean: 103.5\n          std: 13.7\n    ```\n    - in Python\n    >>> preprocessing = [FixedZeroMeanUnitVarianceDescr(\n    ...   kwargs=FixedZeroMeanUnitVarianceKwargs(mean=103.5, std=13.7)\n    ... )]\n\n2. independently along an axis\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: fixed_zero_mean_unit_variance\n        kwargs:\n          axis: channel\n          mean: [101.5, 102.5, 103.5]\n          std: [11.7, 12.7, 13.7]\n    ```\n    - in Python\n    >>> preprocessing = [FixedZeroMeanUnitVarianceDescr(\n    ...   kwargs=FixedZeroMeanUnitVarianceAlongAxisKwargs(\n    ...     axis=AxisId(\"channel\"),\n    ...     mean=[101.5, 102.5, 103.5],\n    ...     std=[11.7, 12.7, 13.7],\n    ...   )\n    ... )]",
      "properties": {
        "id": {
          "const": "fixed_zero_mean_unit_variance",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "anyOf": [
            {
              "$ref": "#/$defs/FixedZeroMeanUnitVarianceKwargs"
            },
            {
              "$ref": "#/$defs/FixedZeroMeanUnitVarianceAlongAxisKwargs"
            }
          ],
          "title": "Kwargs"
        }
      },
      "required": [
        "id",
        "kwargs"
      ],
      "title": "model.v0_5.FixedZeroMeanUnitVarianceDescr",
      "type": "object"
    },
    "FixedZeroMeanUnitVarianceKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [FixedZeroMeanUnitVarianceDescr][]",
      "properties": {
        "mean": {
          "description": "The mean value to normalize with.",
          "title": "Mean",
          "type": "number"
        },
        "std": {
          "description": "The standard deviation value to normalize with.",
          "minimum": 1e-06,
          "title": "Std",
          "type": "number"
        }
      },
      "required": [
        "mean",
        "std"
      ],
      "title": "model.v0_5.FixedZeroMeanUnitVarianceKwargs",
      "type": "object"
    },
    "IndexInputAxis": {
      "additionalProperties": false,
      "properties": {
        "size": {
          "anyOf": [
            {
              "exclusiveMinimum": 0,
              "type": "integer"
            },
            {
              "$ref": "#/$defs/ParameterizedSize"
            },
            {
              "$ref": "#/$defs/SizeReference"
            }
          ],
          "description": "The size/length of this axis can be specified as\n- fixed integer\n- parameterized series of valid sizes ([ParameterizedSize][])\n- reference to another axis with an optional offset ([SizeReference][])",
          "examples": [
            10,
            {
              "min": 32,
              "step": 16
            },
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ],
          "title": "Size"
        },
        "id": {
          "default": "index",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "index",
          "title": "Type",
          "type": "string"
        },
        "concatenable": {
          "default": false,
          "description": "If a model has a `concatenable` input axis, it can be processed blockwise,\nsplitting a longer sample axis into blocks matching its input tensor description.\nOutput axes are concatenable if they have a [SizeReference][] to a concatenable\ninput axis.",
          "title": "Concatenable",
          "type": "boolean"
        }
      },
      "required": [
        "size",
        "type"
      ],
      "title": "model.v0_5.IndexInputAxis",
      "type": "object"
    },
    "IndexOutputAxis": {
      "additionalProperties": false,
      "properties": {
        "id": {
          "default": "index",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "index",
          "title": "Type",
          "type": "string"
        },
        "size": {
          "anyOf": [
            {
              "exclusiveMinimum": 0,
              "type": "integer"
            },
            {
              "$ref": "#/$defs/SizeReference"
            },
            {
              "$ref": "#/$defs/DataDependentSize"
            }
          ],
          "description": "The size/length of this axis can be specified as\n- fixed integer\n- reference to another axis with an optional offset ([SizeReference][])\n- data dependent size using [DataDependentSize][] (size is only known after model inference)",
          "examples": [
            10,
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ],
          "title": "Size"
        }
      },
      "required": [
        "type",
        "size"
      ],
      "title": "model.v0_5.IndexOutputAxis",
      "type": "object"
    },
    "InputTensorDescr": {
      "additionalProperties": false,
      "properties": {
        "id": {
          "default": "input",
          "description": "Input tensor id.\nNo duplicates are allowed across all inputs and outputs.",
          "maxLength": 32,
          "minLength": 1,
          "title": "TensorId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "free text description",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "axes": {
          "description": "tensor axes",
          "items": {
            "discriminator": {
              "mapping": {
                "batch": "#/$defs/BatchAxis",
                "channel": "#/$defs/ChannelAxis",
                "index": "#/$defs/IndexInputAxis",
                "space": "#/$defs/SpaceInputAxis",
                "time": "#/$defs/TimeInputAxis"
              },
              "propertyName": "type"
            },
            "oneOf": [
              {
                "$ref": "#/$defs/BatchAxis"
              },
              {
                "$ref": "#/$defs/ChannelAxis"
              },
              {
                "$ref": "#/$defs/IndexInputAxis"
              },
              {
                "$ref": "#/$defs/TimeInputAxis"
              },
              {
                "$ref": "#/$defs/SpaceInputAxis"
              }
            ]
          },
          "minItems": 1,
          "title": "Axes",
          "type": "array"
        },
        "test_tensor": {
          "anyOf": [
            {
              "$ref": "#/$defs/FileDescr"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "An example tensor to use for testing.\nUsing the model with the test input tensors is expected to yield the test output tensors.\nEach test tensor has be a an ndarray in the\n[numpy.lib file format](https://numpy.org/doc/stable/reference/generated/numpy.lib.format.html#module-numpy.lib.format).\nThe file extension must be '.npy'."
        },
        "sample_tensor": {
          "anyOf": [
            {
              "$ref": "#/$defs/FileDescr"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "A sample tensor to illustrate a possible input/output for the model,\nThe sample image primarily serves to inform a human user about an example use case\nand is typically stored as .hdf5, .png or .tiff.\nIt has to be readable by the [imageio library](https://imageio.readthedocs.io/en/stable/formats/index.html#supported-formats)\n(numpy's `.npy` format is not supported).\nThe image dimensionality has to match the number of axes specified in this tensor description."
        },
        "data": {
          "anyOf": [
            {
              "$ref": "#/$defs/NominalOrOrdinalDataDescr"
            },
            {
              "$ref": "#/$defs/IntervalOrRatioDataDescr"
            },
            {
              "items": {
                "anyOf": [
                  {
                    "$ref": "#/$defs/NominalOrOrdinalDataDescr"
                  },
                  {
                    "$ref": "#/$defs/IntervalOrRatioDataDescr"
                  }
                ]
              },
              "minItems": 1,
              "type": "array"
            }
          ],
          "default": {
            "type": "float32",
            "range": [
              null,
              null
            ],
            "unit": "arbitrary unit",
            "scale": 1.0,
            "offset": null
          },
          "description": "Description of the tensor's data values, optionally per channel.\nIf specified per channel, the data `type` needs to match across channels.",
          "title": "Data"
        },
        "output_of": {
          "anyOf": [
            {
              "minLength": 1,
              "title": "ModelId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "If this input tensor is the output of another model, specify the model id here.\nThis model's input id must match the output id of the referenced model.",
          "title": "Output Of"
        },
        "optional": {
          "default": false,
          "description": "indicates that this tensor may be `None`",
          "title": "Optional",
          "type": "boolean"
        },
        "pad": {
          "anyOf": [
            {
              "$ref": "#/$defs/ConstantPadding"
            },
            {
              "$ref": "#/$defs/EdgePadding"
            },
            {
              "$ref": "#/$defs/ReflectPadding"
            },
            {
              "$ref": "#/$defs/SymmetricPadding"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Explicitly specify how to pad this input tensor.\n\nUse `axes[i].pad` to specify padding width.\n\nNote:\n  Non-blockwise sample prediction only applies padding for axes with a `pad` specification.",
          "title": "Pad"
        },
        "preprocessing": {
          "description": "Description of how this input should be preprocessed.\n\nnotes:\n- If preprocessing does not start with an 'ensure_dtype' entry, it is added\n  to ensure an input tensor's data type matches the input tensor's data description.\n- If preprocessing does not end with an 'ensure_dtype' or 'binarize' entry, an\n  'ensure_dtype' step is added to ensure preprocessing steps are not unintentionally\n  changing the data type.",
          "items": {
            "discriminator": {
              "mapping": {
                "binarize": "#/$defs/BinarizeDescr",
                "clip": "#/$defs/ClipDescr",
                "ensure_dtype": "#/$defs/EnsureDtypeDescr",
                "fixed_zero_mean_unit_variance": "#/$defs/FixedZeroMeanUnitVarianceDescr",
                "scale_linear": "#/$defs/ScaleLinearDescr",
                "scale_range": "#/$defs/ScaleRangeDescr",
                "sigmoid": "#/$defs/SigmoidDescr",
                "softmax": "#/$defs/SoftmaxDescr",
                "zero_mean_unit_variance": "#/$defs/ZeroMeanUnitVarianceDescr"
              },
              "propertyName": "id"
            },
            "oneOf": [
              {
                "$ref": "#/$defs/BinarizeDescr"
              },
              {
                "$ref": "#/$defs/ClipDescr"
              },
              {
                "$ref": "#/$defs/EnsureDtypeDescr"
              },
              {
                "$ref": "#/$defs/FixedZeroMeanUnitVarianceDescr"
              },
              {
                "$ref": "#/$defs/ScaleLinearDescr"
              },
              {
                "$ref": "#/$defs/ScaleRangeDescr"
              },
              {
                "$ref": "#/$defs/SigmoidDescr"
              },
              {
                "$ref": "#/$defs/SoftmaxDescr"
              },
              {
                "$ref": "#/$defs/ZeroMeanUnitVarianceDescr"
              }
            ]
          },
          "title": "Preprocessing",
          "type": "array"
        }
      },
      "required": [
        "axes"
      ],
      "title": "model.v0_5.InputTensorDescr",
      "type": "object"
    },
    "IntervalOrRatioDataDescr": {
      "additionalProperties": false,
      "properties": {
        "type": {
          "default": "float32",
          "enum": [
            "float32",
            "float64",
            "uint8",
            "int8",
            "uint16",
            "int16",
            "uint32",
            "int32",
            "uint64",
            "int64"
          ],
          "examples": [
            "float32",
            "float64",
            "uint8",
            "uint16"
          ],
          "title": "Type",
          "type": "string"
        },
        "range": {
          "default": [
            null,
            null
          ],
          "description": "Tuple `(minimum, maximum)` specifying the allowed range of the data in this tensor.\n`None` corresponds to min/max of what can be expressed by **type**.",
          "maxItems": 2,
          "minItems": 2,
          "prefixItems": [
            {
              "anyOf": [
                {
                  "type": "number"
                },
                {
                  "type": "null"
                }
              ]
            },
            {
              "anyOf": [
                {
                  "type": "number"
                },
                {
                  "type": "null"
                }
              ]
            }
          ],
          "title": "Range",
          "type": "array"
        },
        "unit": {
          "anyOf": [
            {
              "const": "arbitrary unit",
              "type": "string"
            },
            {
              "description": "An SI unit",
              "minLength": 1,
              "pattern": "^(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^[+-]?[1-9]\\d*)?((\u00b7(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^[+-]?[1-9]\\d*)?)|(/(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^+?[1-9]\\d*)?))*$",
              "title": "SiUnit",
              "type": "string"
            }
          ],
          "default": "arbitrary unit",
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "description": "Scale for data on an interval (or ratio) scale.",
          "title": "Scale",
          "type": "number"
        },
        "offset": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Offset for data on a ratio scale.",
          "title": "Offset"
        }
      },
      "title": "model.v0_5.IntervalOrRatioDataDescr",
      "type": "object"
    },
    "KerasHdf5WeightsDescr": {
      "additionalProperties": false,
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "Source of the weights file.",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        },
        "authors": {
          "anyOf": [
            {
              "items": {
                "$ref": "#/$defs/bioimageio__spec__generic__v0_3__Author"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
          "title": "Authors"
        },
        "parent": {
          "anyOf": [
            {
              "enum": [
                "keras_hdf5",
                "keras_v3",
                "onnx",
                "pytorch_state_dict",
                "tensorflow_js",
                "tensorflow_saved_model_bundle",
                "torchscript"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
          "examples": [
            "pytorch_state_dict"
          ],
          "title": "Parent"
        },
        "comment": {
          "default": "",
          "description": "A comment about this weights entry, for example how these weights were created.",
          "title": "Comment",
          "type": "string"
        },
        "tensorflow_version": {
          "$ref": "#/$defs/Version",
          "description": "TensorFlow version used to create these weights."
        }
      },
      "required": [
        "source",
        "tensorflow_version"
      ],
      "title": "model.v0_5.KerasHdf5WeightsDescr",
      "type": "object"
    },
    "KerasV3WeightsDescr": {
      "additionalProperties": false,
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "Source of the .keras weights file.",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        },
        "authors": {
          "anyOf": [
            {
              "items": {
                "$ref": "#/$defs/bioimageio__spec__generic__v0_3__Author"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
          "title": "Authors"
        },
        "parent": {
          "anyOf": [
            {
              "enum": [
                "keras_hdf5",
                "keras_v3",
                "onnx",
                "pytorch_state_dict",
                "tensorflow_js",
                "tensorflow_saved_model_bundle",
                "torchscript"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
          "examples": [
            "pytorch_state_dict"
          ],
          "title": "Parent"
        },
        "comment": {
          "default": "",
          "description": "A comment about this weights entry, for example how these weights were created.",
          "title": "Comment",
          "type": "string"
        },
        "keras_version": {
          "$ref": "#/$defs/Version",
          "description": "Keras version used to create these weights.",
          "ge": 3
        },
        "backend": {
          "description": "Keras backend used to create these weights.",
          "maxItems": 2,
          "minItems": 2,
          "prefixItems": [
            {
              "enum": [
                "tensorflow",
                "jax",
                "torch"
              ],
              "type": "string"
            },
            {
              "$ref": "#/$defs/Version"
            }
          ],
          "title": "Backend",
          "type": "array"
        }
      },
      "required": [
        "source",
        "keras_version",
        "backend"
      ],
      "title": "model.v0_5.KerasV3WeightsDescr",
      "type": "object"
    },
    "LinkedDataset": {
      "additionalProperties": false,
      "description": "Reference to a bioimage.io dataset.",
      "properties": {
        "version": {
          "anyOf": [
            {
              "$ref": "#/$defs/Version"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The version of the linked resource following SemVer 2.0."
        },
        "id": {
          "description": "A valid dataset `id` from the bioimage.io collection.",
          "minLength": 1,
          "title": "DatasetId",
          "type": "string"
        }
      },
      "required": [
        "id"
      ],
      "title": "dataset.v0_3.LinkedDataset",
      "type": "object"
    },
    "LinkedModel": {
      "additionalProperties": false,
      "description": "Reference to a bioimage.io model.",
      "properties": {
        "version": {
          "anyOf": [
            {
              "$ref": "#/$defs/Version"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The version of the linked resource following SemVer 2.0."
        },
        "id": {
          "description": "A valid model `id` from the bioimage.io collection.",
          "minLength": 1,
          "title": "ModelId",
          "type": "string"
        }
      },
      "required": [
        "id"
      ],
      "title": "model.v0_5.LinkedModel",
      "type": "object"
    },
    "NominalOrOrdinalDataDescr": {
      "additionalProperties": false,
      "properties": {
        "values": {
          "anyOf": [
            {
              "items": {
                "type": "integer"
              },
              "minItems": 1,
              "type": "array"
            },
            {
              "items": {
                "type": "number"
              },
              "minItems": 1,
              "type": "array"
            },
            {
              "items": {
                "type": "boolean"
              },
              "minItems": 1,
              "type": "array"
            },
            {
              "items": {
                "type": "string"
              },
              "minItems": 1,
              "type": "array"
            }
          ],
          "description": "A fixed set of nominal or an ascending sequence of ordinal values.\nIn this case `data.type` is required to be an unsigend integer type, e.g. 'uint8'.\nString `values` are interpreted as labels for tensor values 0, ..., N.\nNote: as YAML 1.2 does not natively support a \"set\" datatype,\nnominal values should be given as a sequence (aka list/array) as well.",
          "title": "Values"
        },
        "type": {
          "default": "uint8",
          "enum": [
            "float32",
            "float64",
            "uint8",
            "int8",
            "uint16",
            "int16",
            "uint32",
            "int32",
            "uint64",
            "int64",
            "bool"
          ],
          "examples": [
            "float32",
            "uint8",
            "uint16",
            "int64",
            "bool"
          ],
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "const": "arbitrary unit",
              "type": "string"
            },
            {
              "description": "An SI unit",
              "minLength": 1,
              "pattern": "^(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^[+-]?[1-9]\\d*)?((\u00b7(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^[+-]?[1-9]\\d*)?)|(/(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^+?[1-9]\\d*)?))*$",
              "title": "SiUnit",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        }
      },
      "required": [
        "values"
      ],
      "title": "model.v0_5.NominalOrOrdinalDataDescr",
      "type": "object"
    },
    "OnnxWeightsDescr": {
      "additionalProperties": false,
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "Source of the weights file.",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        },
        "authors": {
          "anyOf": [
            {
              "items": {
                "$ref": "#/$defs/bioimageio__spec__generic__v0_3__Author"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
          "title": "Authors"
        },
        "parent": {
          "anyOf": [
            {
              "enum": [
                "keras_hdf5",
                "keras_v3",
                "onnx",
                "pytorch_state_dict",
                "tensorflow_js",
                "tensorflow_saved_model_bundle",
                "torchscript"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
          "examples": [
            "pytorch_state_dict"
          ],
          "title": "Parent"
        },
        "comment": {
          "default": "",
          "description": "A comment about this weights entry, for example how these weights were created.",
          "title": "Comment",
          "type": "string"
        },
        "opset_version": {
          "description": "ONNX opset version",
          "minimum": 7,
          "title": "Opset Version",
          "type": "integer"
        },
        "external_data": {
          "anyOf": [
            {
              "$ref": "#/$defs/FileDescr",
              "examples": [
                {
                  "source": "weights.onnx.data"
                }
              ]
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Source of the external ONNX data file holding the weights.\n(If present **source** holds the ONNX architecture without weights)."
        }
      },
      "required": [
        "source",
        "opset_version"
      ],
      "title": "model.v0_5.OnnxWeightsDescr",
      "type": "object"
    },
    "OutputTensorDescr": {
      "additionalProperties": false,
      "properties": {
        "id": {
          "default": "output",
          "description": "Output tensor id.\nNo duplicates are allowed across all inputs and outputs.",
          "maxLength": 32,
          "minLength": 1,
          "title": "TensorId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "free text description",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "axes": {
          "description": "tensor axes",
          "items": {
            "discriminator": {
              "mapping": {
                "batch": "#/$defs/BatchAxis",
                "channel": "#/$defs/ChannelAxis",
                "index": "#/$defs/IndexOutputAxis",
                "space": {
                  "oneOf": [
                    {
                      "$ref": "#/$defs/SpaceOutputAxis"
                    },
                    {
                      "$ref": "#/$defs/SpaceOutputAxisWithHalo"
                    }
                  ]
                },
                "time": {
                  "oneOf": [
                    {
                      "$ref": "#/$defs/TimeOutputAxis"
                    },
                    {
                      "$ref": "#/$defs/TimeOutputAxisWithHalo"
                    }
                  ]
                }
              },
              "propertyName": "type"
            },
            "oneOf": [
              {
                "$ref": "#/$defs/BatchAxis"
              },
              {
                "$ref": "#/$defs/ChannelAxis"
              },
              {
                "$ref": "#/$defs/IndexOutputAxis"
              },
              {
                "oneOf": [
                  {
                    "$ref": "#/$defs/TimeOutputAxis"
                  },
                  {
                    "$ref": "#/$defs/TimeOutputAxisWithHalo"
                  }
                ]
              },
              {
                "oneOf": [
                  {
                    "$ref": "#/$defs/SpaceOutputAxis"
                  },
                  {
                    "$ref": "#/$defs/SpaceOutputAxisWithHalo"
                  }
                ]
              }
            ]
          },
          "minItems": 1,
          "title": "Axes",
          "type": "array"
        },
        "test_tensor": {
          "anyOf": [
            {
              "$ref": "#/$defs/FileDescr"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "An example tensor to use for testing.\nUsing the model with the test input tensors is expected to yield the test output tensors.\nEach test tensor has be a an ndarray in the\n[numpy.lib file format](https://numpy.org/doc/stable/reference/generated/numpy.lib.format.html#module-numpy.lib.format).\nThe file extension must be '.npy'."
        },
        "sample_tensor": {
          "anyOf": [
            {
              "$ref": "#/$defs/FileDescr"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "A sample tensor to illustrate a possible input/output for the model,\nThe sample image primarily serves to inform a human user about an example use case\nand is typically stored as .hdf5, .png or .tiff.\nIt has to be readable by the [imageio library](https://imageio.readthedocs.io/en/stable/formats/index.html#supported-formats)\n(numpy's `.npy` format is not supported).\nThe image dimensionality has to match the number of axes specified in this tensor description."
        },
        "data": {
          "anyOf": [
            {
              "$ref": "#/$defs/NominalOrOrdinalDataDescr"
            },
            {
              "$ref": "#/$defs/IntervalOrRatioDataDescr"
            },
            {
              "items": {
                "anyOf": [
                  {
                    "$ref": "#/$defs/NominalOrOrdinalDataDescr"
                  },
                  {
                    "$ref": "#/$defs/IntervalOrRatioDataDescr"
                  }
                ]
              },
              "minItems": 1,
              "type": "array"
            }
          ],
          "default": {
            "type": "float32",
            "range": [
              null,
              null
            ],
            "unit": "arbitrary unit",
            "scale": 1.0,
            "offset": null
          },
          "description": "Description of the tensor's data values, optionally per channel.\nIf specified per channel, the data `type` needs to match across channels.",
          "title": "Data"
        },
        "postprocessing": {
          "description": "Description of how this output should be postprocessed.\n\nnote: `postprocessing` always ends with an 'ensure_dtype' operation.\n      If not given this is added to cast to this tensor's `data.type`.",
          "items": {
            "discriminator": {
              "mapping": {
                "binarize": "#/$defs/BinarizeDescr",
                "cellpose_flow_dynamics": "#/$defs/CellposeFlowDynamicsDescr",
                "clip": "#/$defs/ClipDescr",
                "custom": "#/$defs/CustomProcessingDescr",
                "ensure_dtype": "#/$defs/EnsureDtypeDescr",
                "fixed_zero_mean_unit_variance": "#/$defs/FixedZeroMeanUnitVarianceDescr",
                "scale_linear": "#/$defs/ScaleLinearDescr",
                "scale_mean_variance": "#/$defs/ScaleMeanVarianceDescr",
                "scale_range": "#/$defs/ScaleRangeDescr",
                "sigmoid": "#/$defs/SigmoidDescr",
                "softmax": "#/$defs/SoftmaxDescr",
                "stardist_postprocessing": "#/$defs/StardistPostprocessingDescr",
                "zero_mean_unit_variance": "#/$defs/ZeroMeanUnitVarianceDescr"
              },
              "propertyName": "id"
            },
            "oneOf": [
              {
                "$ref": "#/$defs/BinarizeDescr"
              },
              {
                "$ref": "#/$defs/CellposeFlowDynamicsDescr"
              },
              {
                "$ref": "#/$defs/ClipDescr"
              },
              {
                "$ref": "#/$defs/CustomProcessingDescr"
              },
              {
                "$ref": "#/$defs/EnsureDtypeDescr"
              },
              {
                "$ref": "#/$defs/FixedZeroMeanUnitVarianceDescr"
              },
              {
                "$ref": "#/$defs/ScaleLinearDescr"
              },
              {
                "$ref": "#/$defs/ScaleMeanVarianceDescr"
              },
              {
                "$ref": "#/$defs/ScaleRangeDescr"
              },
              {
                "$ref": "#/$defs/SigmoidDescr"
              },
              {
                "$ref": "#/$defs/SoftmaxDescr"
              },
              {
                "$ref": "#/$defs/StardistPostprocessingDescr"
              },
              {
                "$ref": "#/$defs/ZeroMeanUnitVarianceDescr"
              }
            ]
          },
          "title": "Postprocessing",
          "type": "array"
        }
      },
      "required": [
        "axes"
      ],
      "title": "model.v0_5.OutputTensorDescr",
      "type": "object"
    },
    "ParameterizedSize": {
      "additionalProperties": false,
      "description": "Describes a range of valid tensor axis sizes as `size = min + n*step`.\n\n- **min** and **step** are given by the model description.\n- All blocksize paramters n = 0,1,2,... yield a valid `size`.\n- A greater blocksize paramter n = 0,1,2,... results in a greater **size**.\n  This allows to adjust the axis size more generically.",
      "properties": {
        "min": {
          "exclusiveMinimum": 0,
          "title": "Min",
          "type": "integer"
        },
        "step": {
          "exclusiveMinimum": 0,
          "title": "Step",
          "type": "integer"
        }
      },
      "required": [
        "min",
        "step"
      ],
      "title": "model.v0_5.ParameterizedSize",
      "type": "object"
    },
    "PytorchStateDictWeightsDescr": {
      "additionalProperties": false,
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "Source of the weights file.",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        },
        "authors": {
          "anyOf": [
            {
              "items": {
                "$ref": "#/$defs/bioimageio__spec__generic__v0_3__Author"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
          "title": "Authors"
        },
        "parent": {
          "anyOf": [
            {
              "enum": [
                "keras_hdf5",
                "keras_v3",
                "onnx",
                "pytorch_state_dict",
                "tensorflow_js",
                "tensorflow_saved_model_bundle",
                "torchscript"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
          "examples": [
            "pytorch_state_dict"
          ],
          "title": "Parent"
        },
        "comment": {
          "default": "",
          "description": "A comment about this weights entry, for example how these weights were created.",
          "title": "Comment",
          "type": "string"
        },
        "architecture": {
          "anyOf": [
            {
              "$ref": "#/$defs/ArchitectureFromFileDescr"
            },
            {
              "$ref": "#/$defs/ArchitectureFromLibraryDescr"
            }
          ],
          "title": "Architecture"
        },
        "pytorch_version": {
          "$ref": "#/$defs/Version",
          "description": "Version of the PyTorch library used.\nIf `architecture.depencencies` is specified it has to include pytorch and any version pinning has to be compatible."
        },
        "dependencies": {
          "anyOf": [
            {
              "$ref": "#/$defs/FileDescr",
              "examples": [
                {
                  "source": "environment.yaml"
                }
              ]
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Custom depencies beyond pytorch described in a Conda environment file.\nAllows to specify custom dependencies, see conda docs:\n- [Exporting an environment file across platforms](https://conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html#exporting-an-environment-file-across-platforms)\n- [Creating an environment file manually](https://conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html#creating-an-environment-file-manually)\n\nThe conda environment file should include pytorch and any version pinning has to be compatible with\n**pytorch_version**."
        },
        "strict": {
          "default": true,
          "description": "Whether to allow missing or unexpected keys or to be strict about the architecture matching the state dict weights.",
          "title": "Strict",
          "type": "boolean"
        }
      },
      "required": [
        "source",
        "architecture",
        "pytorch_version"
      ],
      "title": "model.v0_5.PytorchStateDictWeightsDescr",
      "type": "object"
    },
    "ReflectPadding": {
      "additionalProperties": false,
      "properties": {
        "mode": {
          "const": "reflect",
          "default": "reflect",
          "title": "Mode",
          "type": "string"
        }
      },
      "title": "model.v0_5.ReflectPadding",
      "type": "object"
    },
    "RelativeFilePath": {
      "description": "A path relative to the `rdf.yaml` file (also if the RDF source is a URL).",
      "format": "path",
      "title": "RelativeFilePath",
      "type": "string"
    },
    "ReproducibilityTolerance": {
      "additionalProperties": true,
      "properties": {
        "mismatched_elements_per_million": {
          "default": 100,
          "description": "Maximum number of mismatched elements/pixels per million to tolerate.",
          "maximum": 30000,
          "minimum": 0,
          "title": "Mismatched Elements Per Million",
          "type": "integer"
        },
        "weights_formats": {
          "default": [],
          "description": "Limits the weights formats these details apply to.",
          "items": {
            "enum": [
              "keras_hdf5",
              "keras_v3",
              "onnx",
              "pytorch_state_dict",
              "tensorflow_js",
              "tensorflow_saved_model_bundle",
              "torchscript"
            ],
            "type": "string"
          },
          "title": "Weights Formats",
          "type": "array"
        },
        "relative_tolerance": {
          "default": 0.001,
          "description": "Maximum relative tolerance of reproduced test tensor.\nNeeds to be 0 for discrete (integer and boolean) outputs.",
          "maximum": 0.01,
          "minimum": 0,
          "title": "Relative Tolerance",
          "type": "number"
        },
        "absolute_tolerance": {
          "default": 0.001,
          "description": "Maximum absolute tolerance of reproduced test tensor.\n\nThe `absolute_tolerance` may not be greater than 1% of the maximum absolute value of the test tensor\n(this serves to prevent false positives in the presence of very small numbers).\nNeeds to be 0 for discrete (integer and boolean) outputs.",
          "minimum": 0,
          "title": "Absolute Tolerance",
          "type": "number"
        },
        "output_ids": {
          "default": [],
          "description": "Limits the output tensor IDs these reproducibility details apply to.\n\nIf empty, the reproducibility details apply to all outputs.",
          "items": {
            "maxLength": 32,
            "minLength": 1,
            "title": "TensorId",
            "type": "string"
          },
          "title": "Output Ids",
          "type": "array"
        }
      },
      "title": "model.v0_5.ReproducibilityTolerance",
      "type": "object"
    },
    "RunMode": {
      "additionalProperties": false,
      "properties": {
        "name": {
          "anyOf": [
            {
              "const": "deepimagej",
              "type": "string"
            },
            {
              "type": "string"
            }
          ],
          "description": "Run mode name",
          "title": "Name"
        },
        "kwargs": {
          "additionalProperties": true,
          "description": "Run mode specific key word arguments",
          "title": "Kwargs",
          "type": "object"
        }
      },
      "required": [
        "name"
      ],
      "title": "model.v0_4.RunMode",
      "type": "object"
    },
    "ScaleLinearAlongAxisKwargs": {
      "additionalProperties": false,
      "description": "Key word arguments for [ScaleLinearDescr][]",
      "properties": {
        "axis": {
          "description": "The axis of gain and offset values.",
          "examples": [
            "channel"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "gain": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "items": {
                "type": "number"
              },
              "minItems": 1,
              "type": "array"
            }
          ],
          "default": 1.0,
          "description": "multiplicative factor",
          "title": "Gain"
        },
        "offset": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "items": {
                "type": "number"
              },
              "minItems": 1,
              "type": "array"
            }
          ],
          "default": 0.0,
          "description": "additive term",
          "title": "Offset"
        }
      },
      "required": [
        "axis"
      ],
      "title": "model.v0_5.ScaleLinearAlongAxisKwargs",
      "type": "object"
    },
    "ScaleLinearDescr": {
      "additionalProperties": false,
      "description": "Fixed linear scaling.\n\nExamples:\n  1. Scale with scalar gain and offset\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: scale_linear\n        kwargs:\n          gain: 2.0\n          offset: 3.0\n    ```\n    - in Python:\n\n    >>> preprocessing = [\n    ...     ScaleLinearDescr(kwargs=ScaleLinearKwargs(gain= 2.0, offset=3.0))\n    ... ]\n\n  2. Independent scaling along an axis\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: scale_linear\n        kwargs:\n          axis: 'channel'\n          gain: [1.0, 2.0, 3.0]\n    ```\n    - in Python:\n\n    >>> preprocessing = [\n    ...     ScaleLinearDescr(\n    ...         kwargs=ScaleLinearAlongAxisKwargs(\n    ...             axis=AxisId(\"channel\"),\n    ...             gain=[1.0, 2.0, 3.0],\n    ...         )\n    ...     )\n    ... ]",
      "properties": {
        "id": {
          "const": "scale_linear",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "anyOf": [
            {
              "$ref": "#/$defs/ScaleLinearKwargs"
            },
            {
              "$ref": "#/$defs/ScaleLinearAlongAxisKwargs"
            }
          ],
          "title": "Kwargs"
        }
      },
      "required": [
        "id",
        "kwargs"
      ],
      "title": "model.v0_5.ScaleLinearDescr",
      "type": "object"
    },
    "ScaleLinearKwargs": {
      "additionalProperties": false,
      "description": "Key word arguments for [ScaleLinearDescr][]",
      "properties": {
        "gain": {
          "default": 1.0,
          "description": "multiplicative factor",
          "title": "Gain",
          "type": "number"
        },
        "offset": {
          "default": 0.0,
          "description": "additive term",
          "title": "Offset",
          "type": "number"
        }
      },
      "title": "model.v0_5.ScaleLinearKwargs",
      "type": "object"
    },
    "ScaleMeanVarianceDescr": {
      "additionalProperties": false,
      "description": "Scale a tensor's data distribution to match another tensor's mean/std.\n`out  = (tensor - mean) / (std + eps) * (ref_std + eps) + ref_mean.`",
      "properties": {
        "id": {
          "const": "scale_mean_variance",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "$ref": "#/$defs/ScaleMeanVarianceKwargs"
        }
      },
      "required": [
        "id",
        "kwargs"
      ],
      "title": "model.v0_5.ScaleMeanVarianceDescr",
      "type": "object"
    },
    "ScaleMeanVarianceKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [ScaleMeanVarianceKwargs][]",
      "properties": {
        "reference_tensor": {
          "description": "ID of unprocessed input tensor to match.",
          "maxLength": 32,
          "minLength": 1,
          "title": "TensorId",
          "type": "string"
        },
        "axes": {
          "anyOf": [
            {
              "items": {
                "maxLength": 16,
                "minLength": 1,
                "title": "AxisId",
                "type": "string"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The subset of axes to normalize jointly, i.e. axes to reduce to compute mean/std.\nFor example to normalize 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x')\nresulting in a tensor of equal shape normalized per channel, specify `axes=('batch', 'x', 'y')`.\nTo normalize samples independently, leave out the 'batch' axis.\nDefault: Scale all axes jointly.",
          "examples": [
            [
              "batch",
              "x",
              "y"
            ]
          ],
          "title": "Axes"
        },
        "eps": {
          "default": 1e-06,
          "description": "Epsilon for numeric stability:\n`out  = (tensor - mean) / (std + eps) * (ref_std + eps) + ref_mean.`",
          "exclusiveMinimum": 0,
          "maximum": 0.1,
          "title": "Eps",
          "type": "number"
        }
      },
      "required": [
        "reference_tensor"
      ],
      "title": "model.v0_5.ScaleMeanVarianceKwargs",
      "type": "object"
    },
    "ScaleRangeDescr": {
      "additionalProperties": false,
      "description": "Scale with percentiles.\n\nExamples:\n1. Scale linearly to map 5th percentile to 0 and 99.8th percentile to 1.0\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: scale_range\n        kwargs:\n          axes: ['y', 'x']\n          max_percentile: 99.8\n          min_percentile: 5.0\n    ```\n    - in Python\n\n    >>> preprocessing = [\n    ...     ScaleRangeDescr(\n    ...         kwargs=ScaleRangeKwargs(\n    ...           axes= (AxisId('y'), AxisId('x')),\n    ...           max_percentile= 99.8,\n    ...           min_percentile= 5.0,\n    ...         )\n    ...     )\n    ... ]\n\n  2. Combine the above scaling with additional clipping to clip values outside the range given by the percentiles.\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: scale_range\n        kwargs:\n          axes: ['y', 'x']\n          max_percentile: 99.8\n          min_percentile: 5.0\n       - id: clip\n         kwargs:\n          min: 0.0\n          max: 1.0\n    ```\n    - in Python\n\n    >>> preprocessing = [\n    ...   ScaleRangeDescr(\n    ...     kwargs=ScaleRangeKwargs(\n    ...       axes= (AxisId('y'), AxisId('x')),\n    ...       max_percentile= 99.8,\n    ...       min_percentile= 5.0,\n    ...     )\n    ...   ),\n    ...   ClipDescr(\n    ...     kwargs=ClipKwargs(\n    ...       min=0.0,\n    ...       max=1.0,\n    ...     )\n    ...   ),\n    ... ]",
      "properties": {
        "id": {
          "const": "scale_range",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "$ref": "#/$defs/ScaleRangeKwargs"
        }
      },
      "required": [
        "id"
      ],
      "title": "model.v0_5.ScaleRangeDescr",
      "type": "object"
    },
    "ScaleRangeKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [ScaleRangeDescr][]\n\nFor `min_percentile`=0.0 (the default) and `max_percentile`=100 (the default)\nthis processing step normalizes data to the [0, 1] intervall.\nFor other percentiles the normalized values will partially be outside the [0, 1]\nintervall. Use `ScaleRange` followed by `ClipDescr` if you want to limit the\nnormalized values to a range.",
      "properties": {
        "axes": {
          "anyOf": [
            {
              "items": {
                "maxLength": 16,
                "minLength": 1,
                "title": "AxisId",
                "type": "string"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The subset of axes to normalize jointly, i.e. axes to reduce to compute the min/max percentile value.\nFor example to normalize 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x')\nresulting in a tensor of equal shape normalized per channel, specify `axes=('batch', 'x', 'y')`.\nTo normalize samples independently, leave out the \"batch\" axis.\nDefault: Scale all axes jointly.",
          "examples": [
            [
              "batch",
              "x",
              "y"
            ]
          ],
          "title": "Axes"
        },
        "min_percentile": {
          "default": 0.0,
          "description": "The lower percentile used to determine the value to align with zero.",
          "exclusiveMaximum": 100,
          "minimum": 0,
          "title": "Min Percentile",
          "type": "number"
        },
        "max_percentile": {
          "default": 100.0,
          "description": "The upper percentile used to determine the value to align with one.\nHas to be bigger than `min_percentile`.\nThe range is 1 to 100 instead of 0 to 100 to avoid mistakenly\naccepting percentiles specified in the range 0.0 to 1.0.",
          "exclusiveMinimum": 1,
          "maximum": 100,
          "title": "Max Percentile",
          "type": "number"
        },
        "eps": {
          "default": 1e-06,
          "description": "Epsilon for numeric stability.\n`out = (tensor - v_lower) / (v_upper - v_lower + eps)`;\nwith `v_lower,v_upper` values at the respective percentiles.",
          "exclusiveMinimum": 0,
          "maximum": 0.1,
          "title": "Eps",
          "type": "number"
        },
        "reference_tensor": {
          "anyOf": [
            {
              "maxLength": 32,
              "minLength": 1,
              "title": "TensorId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "ID of the unprocessed input tensor to compute the percentiles from.\nDefault: The tensor itself.",
          "title": "Reference Tensor"
        }
      },
      "title": "model.v0_5.ScaleRangeKwargs",
      "type": "object"
    },
    "SigmoidDescr": {
      "additionalProperties": false,
      "description": "The logistic sigmoid function, a.k.a. expit function.\n\nExamples:\n- in YAML\n    ```yaml\n    postprocessing:\n      - id: sigmoid\n    ```\n- in Python:\n\n    >>> postprocessing = [SigmoidDescr()]",
      "properties": {
        "id": {
          "const": "sigmoid",
          "title": "Id",
          "type": "string"
        }
      },
      "required": [
        "id"
      ],
      "title": "model.v0_5.SigmoidDescr",
      "type": "object"
    },
    "SizeReference": {
      "additionalProperties": false,
      "description": "A tensor axis size (extent in pixels/frames) defined in relation to a reference axis.\n\n`axis.size = reference.size * reference.scale / axis.scale + offset`\n\nNote:\n1. The axis and the referenced axis need to have the same unit (or no unit).\n2. Batch axes may not be referenced.\n3. Fractions are rounded down.\n4. If the reference axis is `concatenable` the referencing axis is assumed to be\n    `concatenable` as well with the same block order.\n\nExample:\nAn unisotropic input image of w*h=100*49 pixels depicts a phsical space of 200*196mm\u00b2.\nLet's assume that we want to express the image height h in relation to its width w\ninstead of only accepting input images of exactly 100*49 pixels\n(for example to express a range of valid image shapes by parametrizing w, see `ParameterizedSize`).\n\n>>> w = SpaceInputAxis(id=AxisId(\"w\"), size=100, unit=\"millimeter\", scale=2)\n>>> h = SpaceInputAxis(\n...     id=AxisId(\"h\"),\n...     size=SizeReference(tensor_id=TensorId(\"input\"), axis_id=AxisId(\"w\"), offset=-1),\n...     unit=\"millimeter\",\n...     scale=4,\n... )\n>>> print(h.size.get_size(h, w))\n49\n\n\u21d2 h = w * w.scale / h.scale + offset = 100 * 2mm / 4mm - 1 = 49",
      "properties": {
        "tensor_id": {
          "description": "tensor id of the reference axis",
          "maxLength": 32,
          "minLength": 1,
          "title": "TensorId",
          "type": "string"
        },
        "axis_id": {
          "description": "axis id of the reference axis",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "offset": {
          "default": 0,
          "title": "Offset",
          "type": "integer"
        }
      },
      "required": [
        "tensor_id",
        "axis_id"
      ],
      "title": "model.v0_5.SizeReference",
      "type": "object"
    },
    "SoftmaxDescr": {
      "additionalProperties": false,
      "description": "The softmax function.\n\nExamples:\n- in YAML\n    ```yaml\n    postprocessing:\n      - id: softmax\n        kwargs:\n          axis: channel\n    ```\n- in Python:\n\n    >>> postprocessing = [SoftmaxDescr(kwargs=SoftmaxKwargs(axis=AxisId(\"channel\")))]",
      "properties": {
        "id": {
          "const": "softmax",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "$ref": "#/$defs/SoftmaxKwargs"
        }
      },
      "required": [
        "id"
      ],
      "title": "model.v0_5.SoftmaxDescr",
      "type": "object"
    },
    "SoftmaxKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [SoftmaxDescr][]",
      "properties": {
        "axis": {
          "default": "channel",
          "description": "The axis to apply the softmax function along.\nNote:\n    Defaults to 'channel' axis\n    (which may not exist, in which case\n    a different axis id has to be specified).",
          "examples": [
            "channel"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        }
      },
      "title": "model.v0_5.SoftmaxKwargs",
      "type": "object"
    },
    "SpaceInputAxis": {
      "additionalProperties": false,
      "properties": {
        "size": {
          "anyOf": [
            {
              "exclusiveMinimum": 0,
              "type": "integer"
            },
            {
              "$ref": "#/$defs/ParameterizedSize"
            },
            {
              "$ref": "#/$defs/SizeReference"
            }
          ],
          "description": "The size/length of this axis can be specified as\n- fixed integer\n- parameterized series of valid sizes ([ParameterizedSize][])\n- reference to another axis with an optional offset ([SizeReference][])",
          "examples": [
            10,
            {
              "min": 32,
              "step": 16
            },
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ],
          "title": "Size"
        },
        "id": {
          "default": "x",
          "examples": [
            "x",
            "y",
            "z"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "space",
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "enum": [
                "attometer",
                "angstrom",
                "centimeter",
                "decimeter",
                "exameter",
                "femtometer",
                "foot",
                "gigameter",
                "hectometer",
                "inch",
                "kilometer",
                "megameter",
                "meter",
                "micrometer",
                "mile",
                "millimeter",
                "nanometer",
                "parsec",
                "petameter",
                "picometer",
                "terameter",
                "yard",
                "yoctometer",
                "yottameter",
                "zeptometer",
                "zettameter"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "exclusiveMinimum": 0,
          "title": "Scale",
          "type": "number"
        },
        "concatenable": {
          "default": false,
          "description": "If a model has a `concatenable` input axis, it can be processed blockwise,\nsplitting a longer sample axis into blocks matching its input tensor description.\nOutput axes are concatenable if they have a [SizeReference][] to a concatenable\ninput axis.",
          "title": "Concatenable",
          "type": "boolean"
        }
      },
      "required": [
        "size",
        "type"
      ],
      "title": "model.v0_5.SpaceInputAxis",
      "type": "object"
    },
    "SpaceOutputAxis": {
      "additionalProperties": false,
      "properties": {
        "size": {
          "anyOf": [
            {
              "exclusiveMinimum": 0,
              "type": "integer"
            },
            {
              "$ref": "#/$defs/SizeReference"
            }
          ],
          "description": "The size/length of this axis can be specified as\n- fixed integer\n- reference to another axis with an optional offset (see [SizeReference][])",
          "examples": [
            10,
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ],
          "title": "Size"
        },
        "id": {
          "default": "x",
          "examples": [
            "x",
            "y",
            "z"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "space",
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "enum": [
                "attometer",
                "angstrom",
                "centimeter",
                "decimeter",
                "exameter",
                "femtometer",
                "foot",
                "gigameter",
                "hectometer",
                "inch",
                "kilometer",
                "megameter",
                "meter",
                "micrometer",
                "mile",
                "millimeter",
                "nanometer",
                "parsec",
                "petameter",
                "picometer",
                "terameter",
                "yard",
                "yoctometer",
                "yottameter",
                "zeptometer",
                "zettameter"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "exclusiveMinimum": 0,
          "title": "Scale",
          "type": "number"
        }
      },
      "required": [
        "size",
        "type"
      ],
      "title": "model.v0_5.SpaceOutputAxis",
      "type": "object"
    },
    "SpaceOutputAxisWithHalo": {
      "additionalProperties": false,
      "properties": {
        "halo": {
          "description": "The halo should be cropped from the output tensor to avoid boundary effects.\nIt is to be cropped from both sides, i.e. `size_after_crop = size - 2 * halo`.\nTo document a halo that is already cropped by the model use `size.offset` instead.",
          "minimum": 1,
          "title": "Halo",
          "type": "integer"
        },
        "size": {
          "$ref": "#/$defs/SizeReference",
          "description": "reference to another axis with an optional offset (see [SizeReference][])",
          "examples": [
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ]
        },
        "id": {
          "default": "x",
          "examples": [
            "x",
            "y",
            "z"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "space",
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "enum": [
                "attometer",
                "angstrom",
                "centimeter",
                "decimeter",
                "exameter",
                "femtometer",
                "foot",
                "gigameter",
                "hectometer",
                "inch",
                "kilometer",
                "megameter",
                "meter",
                "micrometer",
                "mile",
                "millimeter",
                "nanometer",
                "parsec",
                "petameter",
                "picometer",
                "terameter",
                "yard",
                "yoctometer",
                "yottameter",
                "zeptometer",
                "zettameter"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "exclusiveMinimum": 0,
          "title": "Scale",
          "type": "number"
        }
      },
      "required": [
        "halo",
        "size",
        "type"
      ],
      "title": "model.v0_5.SpaceOutputAxisWithHalo",
      "type": "object"
    },
    "StardistPostprocessingDescr": {
      "additionalProperties": false,
      "description": "Stardist postprocessing including non-maximum suppression and converting polygon representations to instance labels\n\nas described in:\n- Uwe Schmidt, Martin Weigert, Coleman Broaddus, and Gene Myers.\n[*Cell Detection with Star-convex Polygons*](https://arxiv.org/abs/1806.03535).\nInternational Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI), Granada, Spain, September 2018.\n- Martin Weigert, Uwe Schmidt, Robert Haase, Ko Sugawara, and Gene Myers.\n[*Star-convex Polyhedra for 3D Object Detection and Segmentation in Microscopy*](http://openaccess.thecvf.com/content_WACV_2020/papers/Weigert_Star-convex_Polyhedra_for_3D_Object_Detection_and_Segmentation_in_Microscopy_WACV_2020_paper.pdf).\nThe IEEE Winter Conference on Applications of Computer Vision (WACV), Snowmass Village, Colorado, March 2020.\n\nNote: Only available if the `stardist` package is installed.",
      "properties": {
        "id": {
          "const": "stardist_postprocessing",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "anyOf": [
            {
              "$ref": "#/$defs/StardistPostprocessingKwargs2D"
            },
            {
              "$ref": "#/$defs/StardistPostprocessingKwargs3D"
            }
          ],
          "title": "Kwargs"
        }
      },
      "required": [
        "id",
        "kwargs"
      ],
      "title": "model.v0_5.StardistPostprocessingDescr",
      "type": "object"
    },
    "StardistPostprocessingKwargs2D": {
      "additionalProperties": false,
      "properties": {
        "prob_threshold": {
          "description": "The probability threshold for object candidate selection.",
          "title": "Prob Threshold",
          "type": "number"
        },
        "nms_threshold": {
          "description": "The IoU threshold for non-maximum suppression.",
          "title": "Nms Threshold",
          "type": "number"
        },
        "n_rays": {
          "description": "Number of radial lines (rays) cast from the center of an object to its boundary.",
          "title": "N Rays",
          "type": "integer"
        },
        "grid": {
          "description": "Grid size of network predictions.",
          "maxItems": 2,
          "minItems": 2,
          "prefixItems": [
            {
              "type": "integer"
            },
            {
              "type": "integer"
            }
          ],
          "title": "Grid",
          "type": "array"
        },
        "b": {
          "anyOf": [
            {
              "type": "integer"
            },
            {
              "maxItems": 2,
              "minItems": 2,
              "prefixItems": [
                {
                  "maxItems": 2,
                  "minItems": 2,
                  "prefixItems": [
                    {
                      "type": "integer"
                    },
                    {
                      "type": "integer"
                    }
                  ],
                  "type": "array"
                },
                {
                  "maxItems": 2,
                  "minItems": 2,
                  "prefixItems": [
                    {
                      "type": "integer"
                    },
                    {
                      "type": "integer"
                    }
                  ],
                  "type": "array"
                }
              ],
              "type": "array"
            }
          ],
          "description": "Border region in which object probability is set to zero.",
          "title": "B"
        }
      },
      "required": [
        "prob_threshold",
        "nms_threshold",
        "n_rays",
        "grid",
        "b"
      ],
      "title": "model.v0_5.StardistPostprocessingKwargs2D",
      "type": "object"
    },
    "StardistPostprocessingKwargs3D": {
      "additionalProperties": false,
      "properties": {
        "prob_threshold": {
          "description": "The probability threshold for object candidate selection.",
          "title": "Prob Threshold",
          "type": "number"
        },
        "nms_threshold": {
          "description": "The IoU threshold for non-maximum suppression.",
          "title": "Nms Threshold",
          "type": "number"
        },
        "n_rays": {
          "description": "Number of radial lines (rays) cast from the center of an object to its boundary.",
          "title": "N Rays",
          "type": "integer"
        },
        "grid": {
          "description": "Grid size of network predictions.",
          "maxItems": 3,
          "minItems": 3,
          "prefixItems": [
            {
              "type": "integer"
            },
            {
              "type": "integer"
            },
            {
              "type": "integer"
            }
          ],
          "title": "Grid",
          "type": "array"
        },
        "b": {
          "anyOf": [
            {
              "type": "integer"
            },
            {
              "maxItems": 3,
              "minItems": 3,
              "prefixItems": [
                {
                  "maxItems": 2,
                  "minItems": 2,
                  "prefixItems": [
                    {
                      "type": "integer"
                    },
                    {
                      "type": "integer"
                    }
                  ],
                  "type": "array"
                },
                {
                  "maxItems": 2,
                  "minItems": 2,
                  "prefixItems": [
                    {
                      "type": "integer"
                    },
                    {
                      "type": "integer"
                    }
                  ],
                  "type": "array"
                },
                {
                  "maxItems": 2,
                  "minItems": 2,
                  "prefixItems": [
                    {
                      "type": "integer"
                    },
                    {
                      "type": "integer"
                    }
                  ],
                  "type": "array"
                }
              ],
              "type": "array"
            }
          ],
          "description": "Border region in which object probability is set to zero.",
          "title": "B"
        },
        "anisotropy": {
          "description": "Anisotropy factors for 3D star-convex polyhedra, i.e. the physical pixel size along each spatial axis.",
          "maxItems": 3,
          "minItems": 3,
          "prefixItems": [
            {
              "type": "number"
            },
            {
              "type": "number"
            },
            {
              "type": "number"
            }
          ],
          "title": "Anisotropy",
          "type": "array"
        },
        "overlap_label": {
          "anyOf": [
            {
              "type": "integer"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Optional label to apply to any area of overlapping predicted objects.",
          "title": "Overlap Label"
        }
      },
      "required": [
        "prob_threshold",
        "nms_threshold",
        "n_rays",
        "grid",
        "b",
        "anisotropy"
      ],
      "title": "model.v0_5.StardistPostprocessingKwargs3D",
      "type": "object"
    },
    "SymmetricPadding": {
      "additionalProperties": false,
      "properties": {
        "mode": {
          "const": "symmetric",
          "default": "symmetric",
          "title": "Mode",
          "type": "string"
        }
      },
      "title": "model.v0_5.SymmetricPadding",
      "type": "object"
    },
    "TensorflowJsWeightsDescr": {
      "additionalProperties": false,
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "The multi-file weights.\nAll required files/folders should be a zip archive.",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        },
        "authors": {
          "anyOf": [
            {
              "items": {
                "$ref": "#/$defs/bioimageio__spec__generic__v0_3__Author"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
          "title": "Authors"
        },
        "parent": {
          "anyOf": [
            {
              "enum": [
                "keras_hdf5",
                "keras_v3",
                "onnx",
                "pytorch_state_dict",
                "tensorflow_js",
                "tensorflow_saved_model_bundle",
                "torchscript"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
          "examples": [
            "pytorch_state_dict"
          ],
          "title": "Parent"
        },
        "comment": {
          "default": "",
          "description": "A comment about this weights entry, for example how these weights were created.",
          "title": "Comment",
          "type": "string"
        },
        "tensorflow_version": {
          "$ref": "#/$defs/Version",
          "description": "Version of the TensorFlow library used."
        }
      },
      "required": [
        "source",
        "tensorflow_version"
      ],
      "title": "model.v0_5.TensorflowJsWeightsDescr",
      "type": "object"
    },
    "TensorflowSavedModelBundleWeightsDescr": {
      "additionalProperties": false,
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "The multi-file weights.\nAll required files/folders should be a zip archive.",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        },
        "authors": {
          "anyOf": [
            {
              "items": {
                "$ref": "#/$defs/bioimageio__spec__generic__v0_3__Author"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
          "title": "Authors"
        },
        "parent": {
          "anyOf": [
            {
              "enum": [
                "keras_hdf5",
                "keras_v3",
                "onnx",
                "pytorch_state_dict",
                "tensorflow_js",
                "tensorflow_saved_model_bundle",
                "torchscript"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
          "examples": [
            "pytorch_state_dict"
          ],
          "title": "Parent"
        },
        "comment": {
          "default": "",
          "description": "A comment about this weights entry, for example how these weights were created.",
          "title": "Comment",
          "type": "string"
        },
        "tensorflow_version": {
          "$ref": "#/$defs/Version",
          "description": "Version of the TensorFlow library used."
        },
        "dependencies": {
          "anyOf": [
            {
              "$ref": "#/$defs/FileDescr",
              "examples": [
                {
                  "source": "environment.yaml"
                }
              ]
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Custom dependencies beyond tensorflow.\nShould include tensorflow and any version pinning has to be compatible with **tensorflow_version**."
        }
      },
      "required": [
        "source",
        "tensorflow_version"
      ],
      "title": "model.v0_5.TensorflowSavedModelBundleWeightsDescr",
      "type": "object"
    },
    "TimeInputAxis": {
      "additionalProperties": false,
      "properties": {
        "size": {
          "anyOf": [
            {
              "exclusiveMinimum": 0,
              "type": "integer"
            },
            {
              "$ref": "#/$defs/ParameterizedSize"
            },
            {
              "$ref": "#/$defs/SizeReference"
            }
          ],
          "description": "The size/length of this axis can be specified as\n- fixed integer\n- parameterized series of valid sizes ([ParameterizedSize][])\n- reference to another axis with an optional offset ([SizeReference][])",
          "examples": [
            10,
            {
              "min": 32,
              "step": 16
            },
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ],
          "title": "Size"
        },
        "id": {
          "default": "time",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "time",
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "enum": [
                "attosecond",
                "centisecond",
                "day",
                "decisecond",
                "exasecond",
                "femtosecond",
                "gigasecond",
                "hectosecond",
                "hour",
                "kilosecond",
                "megasecond",
                "microsecond",
                "millisecond",
                "minute",
                "nanosecond",
                "petasecond",
                "picosecond",
                "second",
                "terasecond",
                "yoctosecond",
                "yottasecond",
                "zeptosecond",
                "zettasecond"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "exclusiveMinimum": 0,
          "title": "Scale",
          "type": "number"
        },
        "concatenable": {
          "default": false,
          "description": "If a model has a `concatenable` input axis, it can be processed blockwise,\nsplitting a longer sample axis into blocks matching its input tensor description.\nOutput axes are concatenable if they have a [SizeReference][] to a concatenable\ninput axis.",
          "title": "Concatenable",
          "type": "boolean"
        }
      },
      "required": [
        "size",
        "type"
      ],
      "title": "model.v0_5.TimeInputAxis",
      "type": "object"
    },
    "TimeOutputAxis": {
      "additionalProperties": false,
      "properties": {
        "size": {
          "anyOf": [
            {
              "exclusiveMinimum": 0,
              "type": "integer"
            },
            {
              "$ref": "#/$defs/SizeReference"
            }
          ],
          "description": "The size/length of this axis can be specified as\n- fixed integer\n- reference to another axis with an optional offset (see [SizeReference][])",
          "examples": [
            10,
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ],
          "title": "Size"
        },
        "id": {
          "default": "time",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "time",
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "enum": [
                "attosecond",
                "centisecond",
                "day",
                "decisecond",
                "exasecond",
                "femtosecond",
                "gigasecond",
                "hectosecond",
                "hour",
                "kilosecond",
                "megasecond",
                "microsecond",
                "millisecond",
                "minute",
                "nanosecond",
                "petasecond",
                "picosecond",
                "second",
                "terasecond",
                "yoctosecond",
                "yottasecond",
                "zeptosecond",
                "zettasecond"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "exclusiveMinimum": 0,
          "title": "Scale",
          "type": "number"
        }
      },
      "required": [
        "size",
        "type"
      ],
      "title": "model.v0_5.TimeOutputAxis",
      "type": "object"
    },
    "TimeOutputAxisWithHalo": {
      "additionalProperties": false,
      "properties": {
        "halo": {
          "description": "The halo should be cropped from the output tensor to avoid boundary effects.\nIt is to be cropped from both sides, i.e. `size_after_crop = size - 2 * halo`.\nTo document a halo that is already cropped by the model use `size.offset` instead.",
          "minimum": 1,
          "title": "Halo",
          "type": "integer"
        },
        "size": {
          "$ref": "#/$defs/SizeReference",
          "description": "reference to another axis with an optional offset (see [SizeReference][])",
          "examples": [
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ]
        },
        "id": {
          "default": "time",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "time",
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "enum": [
                "attosecond",
                "centisecond",
                "day",
                "decisecond",
                "exasecond",
                "femtosecond",
                "gigasecond",
                "hectosecond",
                "hour",
                "kilosecond",
                "megasecond",
                "microsecond",
                "millisecond",
                "minute",
                "nanosecond",
                "petasecond",
                "picosecond",
                "second",
                "terasecond",
                "yoctosecond",
                "yottasecond",
                "zeptosecond",
                "zettasecond"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "exclusiveMinimum": 0,
          "title": "Scale",
          "type": "number"
        }
      },
      "required": [
        "halo",
        "size",
        "type"
      ],
      "title": "model.v0_5.TimeOutputAxisWithHalo",
      "type": "object"
    },
    "TorchscriptWeightsDescr": {
      "additionalProperties": false,
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "Source of the weights file.",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        },
        "authors": {
          "anyOf": [
            {
              "items": {
                "$ref": "#/$defs/bioimageio__spec__generic__v0_3__Author"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
          "title": "Authors"
        },
        "parent": {
          "anyOf": [
            {
              "enum": [
                "keras_hdf5",
                "keras_v3",
                "onnx",
                "pytorch_state_dict",
                "tensorflow_js",
                "tensorflow_saved_model_bundle",
                "torchscript"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
          "examples": [
            "pytorch_state_dict"
          ],
          "title": "Parent"
        },
        "comment": {
          "default": "",
          "description": "A comment about this weights entry, for example how these weights were created.",
          "title": "Comment",
          "type": "string"
        },
        "pytorch_version": {
          "$ref": "#/$defs/Version",
          "description": "Version of the PyTorch library used."
        }
      },
      "required": [
        "source",
        "pytorch_version"
      ],
      "title": "model.v0_5.TorchscriptWeightsDescr",
      "type": "object"
    },
    "TrainingDetails": {
      "additionalProperties": true,
      "properties": {
        "training_preprocessing": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Detailed image preprocessing steps during model training:\n\nMention:\n- *Normalization methods*\n- *Augmentation strategies*\n- *Resizing/resampling procedures*\n- *Artifact handling*",
          "title": "Training Preprocessing"
        },
        "training_epochs": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Number of training epochs.",
          "title": "Training Epochs"
        },
        "training_batch_size": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Batch size used in training.",
          "title": "Training Batch Size"
        },
        "initial_learning_rate": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Initial learning rate used in training.",
          "title": "Initial Learning Rate"
        },
        "learning_rate_schedule": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Learning rate schedule used in training.",
          "title": "Learning Rate Schedule"
        },
        "loss_function": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Loss function used in training, e.g. nn.MSELoss.",
          "title": "Loss Function"
        },
        "loss_function_kwargs": {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "description": "key word arguments for the `loss_function`",
          "title": "Loss Function Kwargs",
          "type": "object"
        },
        "optimizer": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "optimizer, e.g. torch.optim.Adam",
          "title": "Optimizer"
        },
        "optimizer_kwargs": {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "description": "key word arguments for the `optimizer`",
          "title": "Optimizer Kwargs",
          "type": "object"
        },
        "regularization": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Regularization techniques used during training, e.g. drop-out or weight decay.",
          "title": "Regularization"
        },
        "training_duration": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Total training duration in hours.",
          "title": "Training Duration"
        }
      },
      "title": "model.v0_5.TrainingDetails",
      "type": "object"
    },
    "Uploader": {
      "additionalProperties": false,
      "properties": {
        "email": {
          "description": "Email",
          "format": "email",
          "title": "Email",
          "type": "string"
        },
        "name": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "name",
          "title": "Name"
        }
      },
      "required": [
        "email"
      ],
      "title": "generic.v0_2.Uploader",
      "type": "object"
    },
    "Version": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        }
      ],
      "description": "wraps a packaging.version.Version instance for validation in pydantic models",
      "title": "Version"
    },
    "WeightsDescr": {
      "additionalProperties": false,
      "properties": {
        "keras_hdf5": {
          "anyOf": [
            {
              "$ref": "#/$defs/KerasHdf5WeightsDescr"
            },
            {
              "type": "null"
            }
          ],
          "default": null
        },
        "keras_v3": {
          "anyOf": [
            {
              "$ref": "#/$defs/KerasV3WeightsDescr"
            },
            {
              "type": "null"
            }
          ],
          "default": null
        },
        "onnx": {
          "anyOf": [
            {
              "$ref": "#/$defs/OnnxWeightsDescr"
            },
            {
              "type": "null"
            }
          ],
          "default": null
        },
        "pytorch_state_dict": {
          "anyOf": [
            {
              "$ref": "#/$defs/PytorchStateDictWeightsDescr"
            },
            {
              "type": "null"
            }
          ],
          "default": null
        },
        "tensorflow_js": {
          "anyOf": [
            {
              "$ref": "#/$defs/TensorflowJsWeightsDescr"
            },
            {
              "type": "null"
            }
          ],
          "default": null
        },
        "tensorflow_saved_model_bundle": {
          "anyOf": [
            {
              "$ref": "#/$defs/TensorflowSavedModelBundleWeightsDescr"
            },
            {
              "type": "null"
            }
          ],
          "default": null
        },
        "torchscript": {
          "anyOf": [
            {
              "$ref": "#/$defs/TorchscriptWeightsDescr"
            },
            {
              "type": "null"
            }
          ],
          "default": null
        }
      },
      "title": "model.v0_5.WeightsDescr",
      "type": "object"
    },
    "YamlValue": {
      "anyOf": [
        {
          "type": "boolean"
        },
        {
          "format": "date",
          "type": "string"
        },
        {
          "format": "date-time",
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        },
        {
          "type": "string"
        },
        {
          "items": {
            "$ref": "#/$defs/YamlValue"
          },
          "type": "array"
        },
        {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "type": "object"
        },
        {
          "type": "null"
        }
      ]
    },
    "ZeroMeanUnitVarianceDescr": {
      "additionalProperties": false,
      "description": "Subtract mean and divide by variance.\n\nExamples:\n    Subtract tensor mean and variance\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: zero_mean_unit_variance\n    ```\n    - in Python\n    >>> preprocessing = [ZeroMeanUnitVarianceDescr()]",
      "properties": {
        "id": {
          "const": "zero_mean_unit_variance",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "$ref": "#/$defs/ZeroMeanUnitVarianceKwargs"
        }
      },
      "required": [
        "id"
      ],
      "title": "model.v0_5.ZeroMeanUnitVarianceDescr",
      "type": "object"
    },
    "ZeroMeanUnitVarianceKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [ZeroMeanUnitVarianceDescr][]",
      "properties": {
        "axes": {
          "anyOf": [
            {
              "items": {
                "maxLength": 16,
                "minLength": 1,
                "title": "AxisId",
                "type": "string"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The subset of axes to normalize jointly, i.e. axes to reduce to compute mean/std.\nFor example to normalize 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x')\nresulting in a tensor of equal shape normalized per channel, specify `axes=('batch', 'x', 'y')`.\nTo normalize each sample independently leave out the 'batch' axis.\nDefault: Scale all axes jointly.",
          "examples": [
            [
              "batch",
              "x",
              "y"
            ]
          ],
          "title": "Axes"
        },
        "eps": {
          "default": 1e-06,
          "description": "epsilon for numeric stability: `out = (tensor - mean) / (std + eps)`.",
          "exclusiveMinimum": 0,
          "maximum": 0.1,
          "title": "Eps",
          "type": "number"
        }
      },
      "title": "model.v0_5.ZeroMeanUnitVarianceKwargs",
      "type": "object"
    },
    "bioimageio__spec__dataset__v0_2__DatasetDescr": {
      "additionalProperties": false,
      "description": "A bioimage.io dataset resource description file (dataset RDF) describes a dataset relevant to bioimage\nprocessing.",
      "properties": {
        "name": {
          "description": "A human-friendly name of the resource description",
          "minLength": 1,
          "title": "Name",
          "type": "string"
        },
        "description": {
          "title": "Description",
          "type": "string"
        },
        "covers": {
          "description": "Cover images. Please use an image smaller than 500KB and an aspect ratio width to height of 2:1.\nThe supported image formats are: ('.gif', '.jpeg', '.jpg', '.png', '.svg', '.tif', '.tiff')",
          "examples": [
            [
              "cover.png"
            ]
          ],
          "items": {
            "anyOf": [
              {
                "description": "A URL with the HTTP or HTTPS scheme.",
                "format": "uri",
                "maxLength": 2083,
                "minLength": 1,
                "title": "HttpUrl",
                "type": "string"
              },
              {
                "$ref": "#/$defs/RelativeFilePath"
              },
              {
                "format": "file-path",
                "title": "FilePath",
                "type": "string"
              }
            ],
            "title": "FileSource"
          },
          "title": "Covers",
          "type": "array"
        },
        "id_emoji": {
          "anyOf": [
            {
              "examples": [
                "\ud83e\udd88",
                "\ud83e\udda5"
              ],
              "maxLength": 1,
              "minLength": 1,
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "UTF-8 emoji for display alongside the `id`.",
          "title": "Id Emoji"
        },
        "authors": {
          "description": "The authors are the creators of the RDF and the primary points of contact.",
          "items": {
            "$ref": "#/$defs/bioimageio__spec__generic__v0_2__Author"
          },
          "title": "Authors",
          "type": "array"
        },
        "attachments": {
          "anyOf": [
            {
              "$ref": "#/$defs/AttachmentsDescr"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "file and other attachments"
        },
        "cite": {
          "description": "citations",
          "items": {
            "$ref": "#/$defs/bioimageio__spec__generic__v0_2__CiteEntry"
          },
          "title": "Cite",
          "type": "array"
        },
        "config": {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "description": "A field for custom configuration that can contain any keys not present in the RDF spec.\nThis means you should not store, for example, a github repo URL in `config` since we already have the\n`git_repo` field defined in the spec.\nKeys in `config` may be very specific to a tool or consumer software. To avoid conflicting definitions,\nit is recommended to wrap added configuration into a sub-field named with the specific domain or tool name,\nfor example:\n```yaml\nconfig:\n    bioimageio:  # here is the domain name\n        my_custom_key: 3837283\n        another_key:\n            nested: value\n    imagej:       # config specific to ImageJ\n        macro_dir: path/to/macro/file\n```\nIf possible, please use [`snake_case`](https://en.wikipedia.org/wiki/Snake_case) for keys in `config`.\nYou may want to list linked files additionally under `attachments` to include them when packaging a resource\n(packaging a resource means downloading/copying important linked files and creating a ZIP archive that contains\nan altered rdf.yaml file with local references to the downloaded files)",
          "examples": [
            {
              "bioimageio": {
                "another_key": {
                  "nested": "value"
                },
                "my_custom_key": 3837283
              },
              "imagej": {
                "macro_dir": "path/to/macro/file"
              }
            }
          ],
          "title": "Config",
          "type": "object"
        },
        "download_url": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "URL to download the resource from (deprecated)",
          "title": "Download Url"
        },
        "git_repo": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "A URL to the Git repository where the resource is being developed.",
          "examples": [
            "https://github.com/bioimage-io/spec-bioimage-io/tree/main/example_descriptions/models/unet2d_nuclei_broad"
          ],
          "title": "Git Repo"
        },
        "icon": {
          "anyOf": [
            {
              "maxLength": 2,
              "minLength": 1,
              "type": "string"
            },
            {
              "anyOf": [
                {
                  "description": "A URL with the HTTP or HTTPS scheme.",
                  "format": "uri",
                  "maxLength": 2083,
                  "minLength": 1,
                  "title": "HttpUrl",
                  "type": "string"
                },
                {
                  "$ref": "#/$defs/RelativeFilePath"
                },
                {
                  "format": "file-path",
                  "title": "FilePath",
                  "type": "string"
                }
              ],
              "title": "FileSource"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "An icon for illustration",
          "title": "Icon"
        },
        "links": {
          "description": "IDs of other bioimage.io resources",
          "examples": [
            [
              "ilastik/ilastik",
              "deepimagej/deepimagej",
              "zero/notebook_u-net_3d_zerocostdl4mic"
            ]
          ],
          "items": {
            "type": "string"
          },
          "title": "Links",
          "type": "array"
        },
        "uploader": {
          "anyOf": [
            {
              "$ref": "#/$defs/Uploader"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The person who uploaded the model (e.g. to bioimage.io)"
        },
        "maintainers": {
          "description": "Maintainers of this resource.\nIf not specified `authors` are maintainers and at least some of them should specify their `github_user` name",
          "items": {
            "$ref": "#/$defs/bioimageio__spec__generic__v0_2__Maintainer"
          },
          "title": "Maintainers",
          "type": "array"
        },
        "rdf_source": {
          "anyOf": [
            {
              "anyOf": [
                {
                  "description": "A URL with the HTTP or HTTPS scheme.",
                  "format": "uri",
                  "maxLength": 2083,
                  "minLength": 1,
                  "title": "HttpUrl",
                  "type": "string"
                },
                {
                  "$ref": "#/$defs/RelativeFilePath"
                },
                {
                  "format": "file-path",
                  "title": "FilePath",
                  "type": "string"
                }
              ],
              "title": "FileSource"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Resource description file (RDF) source; used to keep track of where an rdf.yaml was loaded from.\nDo not set this field in a YAML file.",
          "title": "Rdf Source"
        },
        "tags": {
          "description": "Associated tags",
          "examples": [
            [
              "unet2d",
              "pytorch",
              "nucleus",
              "segmentation",
              "dsb2018"
            ]
          ],
          "items": {
            "type": "string"
          },
          "title": "Tags",
          "type": "array"
        },
        "version": {
          "anyOf": [
            {
              "$ref": "#/$defs/Version"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The version of the resource following SemVer 2.0."
        },
        "version_number": {
          "anyOf": [
            {
              "type": "integer"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "version number (n-th published version, not the semantic version)",
          "title": "Version Number"
        },
        "format_version": {
          "const": "0.2.4",
          "description": "The format version of this resource specification\n(not the `version` of the resource description)\nWhen creating a new resource always use the latest micro/patch version described here.\nThe `format_version` is important for any consumer software to understand how to parse the fields.",
          "title": "Format Version",
          "type": "string"
        },
        "badges": {
          "description": "badges associated with this resource",
          "items": {
            "$ref": "#/$defs/BadgeDescr"
          },
          "title": "Badges",
          "type": "array"
        },
        "documentation": {
          "anyOf": [
            {
              "anyOf": [
                {
                  "description": "A URL with the HTTP or HTTPS scheme.",
                  "format": "uri",
                  "maxLength": 2083,
                  "minLength": 1,
                  "title": "HttpUrl",
                  "type": "string"
                },
                {
                  "$ref": "#/$defs/RelativeFilePath"
                },
                {
                  "format": "file-path",
                  "title": "FilePath",
                  "type": "string"
                }
              ],
              "title": "FileSource"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "URL or relative path to a markdown file with additional documentation.\nThe recommended documentation file name is `README.md`. An `.md` suffix is mandatory.",
          "examples": [
            "https://raw.githubusercontent.com/bioimage-io/spec-bioimage-io/main/example_descriptions/models/unet2d_nuclei_broad/README.md",
            "README.md"
          ],
          "title": "Documentation"
        },
        "license": {
          "anyOf": [
            {
              "enum": [
                "0BSD",
                "3D-Slicer-1.0",
                "AAL",
                "Abstyles",
                "AdaCore-doc",
                "Adobe-2006",
                "Adobe-Display-PostScript",
                "Adobe-Glyph",
                "Adobe-Utopia",
                "ADSL",
                "AFL-1.1",
                "AFL-1.2",
                "AFL-2.0",
                "AFL-2.1",
                "AFL-3.0",
                "Afmparse",
                "AGPL-1.0-only",
                "AGPL-1.0-or-later",
                "AGPL-3.0-only",
                "AGPL-3.0-or-later",
                "Aladdin",
                "AMD-newlib",
                "AMDPLPA",
                "AML",
                "AML-glslang",
                "AMPAS",
                "ANTLR-PD",
                "ANTLR-PD-fallback",
                "any-OSI",
                "any-OSI-perl-modules",
                "Apache-1.0",
                "Apache-1.1",
                "Apache-2.0",
                "APAFML",
                "APL-1.0",
                "App-s2p",
                "APSL-1.0",
                "APSL-1.1",
                "APSL-1.2",
                "APSL-2.0",
                "Arphic-1999",
                "Artistic-1.0",
                "Artistic-1.0-cl8",
                "Artistic-1.0-Perl",
                "Artistic-2.0",
                "Artistic-dist",
                "Aspell-RU",
                "ASWF-Digital-Assets-1.0",
                "ASWF-Digital-Assets-1.1",
                "Baekmuk",
                "Bahyph",
                "Barr",
                "bcrypt-Solar-Designer",
                "Beerware",
                "Bitstream-Charter",
                "Bitstream-Vera",
                "BitTorrent-1.0",
                "BitTorrent-1.1",
                "blessing",
                "BlueOak-1.0.0",
                "Boehm-GC",
                "Boehm-GC-without-fee",
                "Borceux",
                "Brian-Gladman-2-Clause",
                "Brian-Gladman-3-Clause",
                "BSD-1-Clause",
                "BSD-2-Clause",
                "BSD-2-Clause-Darwin",
                "BSD-2-Clause-first-lines",
                "BSD-2-Clause-Patent",
                "BSD-2-Clause-pkgconf-disclaimer",
                "BSD-2-Clause-Views",
                "BSD-3-Clause",
                "BSD-3-Clause-acpica",
                "BSD-3-Clause-Attribution",
                "BSD-3-Clause-Clear",
                "BSD-3-Clause-flex",
                "BSD-3-Clause-HP",
                "BSD-3-Clause-LBNL",
                "BSD-3-Clause-Modification",
                "BSD-3-Clause-No-Military-License",
                "BSD-3-Clause-No-Nuclear-License",
                "BSD-3-Clause-No-Nuclear-License-2014",
                "BSD-3-Clause-No-Nuclear-Warranty",
                "BSD-3-Clause-Open-MPI",
                "BSD-3-Clause-Sun",
                "BSD-4-Clause",
                "BSD-4-Clause-Shortened",
                "BSD-4-Clause-UC",
                "BSD-4.3RENO",
                "BSD-4.3TAHOE",
                "BSD-Advertising-Acknowledgement",
                "BSD-Attribution-HPND-disclaimer",
                "BSD-Inferno-Nettverk",
                "BSD-Protection",
                "BSD-Source-beginning-file",
                "BSD-Source-Code",
                "BSD-Systemics",
                "BSD-Systemics-W3Works",
                "BSL-1.0",
                "BUSL-1.1",
                "bzip2-1.0.6",
                "C-UDA-1.0",
                "CAL-1.0",
                "CAL-1.0-Combined-Work-Exception",
                "Caldera",
                "Caldera-no-preamble",
                "Catharon",
                "CATOSL-1.1",
                "CC-BY-1.0",
                "CC-BY-2.0",
                "CC-BY-2.5",
                "CC-BY-2.5-AU",
                "CC-BY-3.0",
                "CC-BY-3.0-AT",
                "CC-BY-3.0-AU",
                "CC-BY-3.0-DE",
                "CC-BY-3.0-IGO",
                "CC-BY-3.0-NL",
                "CC-BY-3.0-US",
                "CC-BY-4.0",
                "CC-BY-NC-1.0",
                "CC-BY-NC-2.0",
                "CC-BY-NC-2.5",
                "CC-BY-NC-3.0",
                "CC-BY-NC-3.0-DE",
                "CC-BY-NC-4.0",
                "CC-BY-NC-ND-1.0",
                "CC-BY-NC-ND-2.0",
                "CC-BY-NC-ND-2.5",
                "CC-BY-NC-ND-3.0",
                "CC-BY-NC-ND-3.0-DE",
                "CC-BY-NC-ND-3.0-IGO",
                "CC-BY-NC-ND-4.0",
                "CC-BY-NC-SA-1.0",
                "CC-BY-NC-SA-2.0",
                "CC-BY-NC-SA-2.0-DE",
                "CC-BY-NC-SA-2.0-FR",
                "CC-BY-NC-SA-2.0-UK",
                "CC-BY-NC-SA-2.5",
                "CC-BY-NC-SA-3.0",
                "CC-BY-NC-SA-3.0-DE",
                "CC-BY-NC-SA-3.0-IGO",
                "CC-BY-NC-SA-4.0",
                "CC-BY-ND-1.0",
                "CC-BY-ND-2.0",
                "CC-BY-ND-2.5",
                "CC-BY-ND-3.0",
                "CC-BY-ND-3.0-DE",
                "CC-BY-ND-4.0",
                "CC-BY-SA-1.0",
                "CC-BY-SA-2.0",
                "CC-BY-SA-2.0-UK",
                "CC-BY-SA-2.1-JP",
                "CC-BY-SA-2.5",
                "CC-BY-SA-3.0",
                "CC-BY-SA-3.0-AT",
                "CC-BY-SA-3.0-DE",
                "CC-BY-SA-3.0-IGO",
                "CC-BY-SA-4.0",
                "CC-PDDC",
                "CC-PDM-1.0",
                "CC-SA-1.0",
                "CC0-1.0",
                "CDDL-1.0",
                "CDDL-1.1",
                "CDL-1.0",
                "CDLA-Permissive-1.0",
                "CDLA-Permissive-2.0",
                "CDLA-Sharing-1.0",
                "CECILL-1.0",
                "CECILL-1.1",
                "CECILL-2.0",
                "CECILL-2.1",
                "CECILL-B",
                "CECILL-C",
                "CERN-OHL-1.1",
                "CERN-OHL-1.2",
                "CERN-OHL-P-2.0",
                "CERN-OHL-S-2.0",
                "CERN-OHL-W-2.0",
                "CFITSIO",
                "check-cvs",
                "checkmk",
                "ClArtistic",
                "Clips",
                "CMU-Mach",
                "CMU-Mach-nodoc",
                "CNRI-Jython",
                "CNRI-Python",
                "CNRI-Python-GPL-Compatible",
                "COIL-1.0",
                "Community-Spec-1.0",
                "Condor-1.1",
                "copyleft-next-0.3.0",
                "copyleft-next-0.3.1",
                "Cornell-Lossless-JPEG",
                "CPAL-1.0",
                "CPL-1.0",
                "CPOL-1.02",
                "Cronyx",
                "Crossword",
                "CryptoSwift",
                "CrystalStacker",
                "CUA-OPL-1.0",
                "Cube",
                "curl",
                "cve-tou",
                "D-FSL-1.0",
                "DEC-3-Clause",
                "diffmark",
                "DL-DE-BY-2.0",
                "DL-DE-ZERO-2.0",
                "DOC",
                "DocBook-DTD",
                "DocBook-Schema",
                "DocBook-Stylesheet",
                "DocBook-XML",
                "Dotseqn",
                "DRL-1.0",
                "DRL-1.1",
                "DSDP",
                "dtoa",
                "dvipdfm",
                "ECL-1.0",
                "ECL-2.0",
                "EFL-1.0",
                "EFL-2.0",
                "eGenix",
                "Elastic-2.0",
                "Entessa",
                "EPICS",
                "EPL-1.0",
                "EPL-2.0",
                "ErlPL-1.1",
                "etalab-2.0",
                "EUDatagrid",
                "EUPL-1.0",
                "EUPL-1.1",
                "EUPL-1.2",
                "Eurosym",
                "Fair",
                "FBM",
                "FDK-AAC",
                "Ferguson-Twofish",
                "Frameworx-1.0",
                "FreeBSD-DOC",
                "FreeImage",
                "FSFAP",
                "FSFAP-no-warranty-disclaimer",
                "FSFUL",
                "FSFULLR",
                "FSFULLRSD",
                "FSFULLRWD",
                "FSL-1.1-ALv2",
                "FSL-1.1-MIT",
                "FTL",
                "Furuseth",
                "fwlw",
                "Game-Programming-Gems",
                "GCR-docs",
                "GD",
                "generic-xts",
                "GFDL-1.1-invariants-only",
                "GFDL-1.1-invariants-or-later",
                "GFDL-1.1-no-invariants-only",
                "GFDL-1.1-no-invariants-or-later",
                "GFDL-1.1-only",
                "GFDL-1.1-or-later",
                "GFDL-1.2-invariants-only",
                "GFDL-1.2-invariants-or-later",
                "GFDL-1.2-no-invariants-only",
                "GFDL-1.2-no-invariants-or-later",
                "GFDL-1.2-only",
                "GFDL-1.2-or-later",
                "GFDL-1.3-invariants-only",
                "GFDL-1.3-invariants-or-later",
                "GFDL-1.3-no-invariants-only",
                "GFDL-1.3-no-invariants-or-later",
                "GFDL-1.3-only",
                "GFDL-1.3-or-later",
                "Giftware",
                "GL2PS",
                "Glide",
                "Glulxe",
                "GLWTPL",
                "gnuplot",
                "GPL-1.0-only",
                "GPL-1.0-or-later",
                "GPL-2.0-only",
                "GPL-2.0-or-later",
                "GPL-3.0-only",
                "GPL-3.0-or-later",
                "Graphics-Gems",
                "gSOAP-1.3b",
                "gtkbook",
                "Gutmann",
                "HaskellReport",
                "HDF5",
                "hdparm",
                "HIDAPI",
                "Hippocratic-2.1",
                "HP-1986",
                "HP-1989",
                "HPND",
                "HPND-DEC",
                "HPND-doc",
                "HPND-doc-sell",
                "HPND-export-US",
                "HPND-export-US-acknowledgement",
                "HPND-export-US-modify",
                "HPND-export2-US",
                "HPND-Fenneberg-Livingston",
                "HPND-INRIA-IMAG",
                "HPND-Intel",
                "HPND-Kevlin-Henney",
                "HPND-Markus-Kuhn",
                "HPND-merchantability-variant",
                "HPND-MIT-disclaimer",
                "HPND-Netrek",
                "HPND-Pbmplus",
                "HPND-sell-MIT-disclaimer-xserver",
                "HPND-sell-regexpr",
                "HPND-sell-variant",
                "HPND-sell-variant-MIT-disclaimer",
                "HPND-sell-variant-MIT-disclaimer-rev",
                "HPND-UC",
                "HPND-UC-export-US",
                "HTMLTIDY",
                "IBM-pibs",
                "ICU",
                "IEC-Code-Components-EULA",
                "IJG",
                "IJG-short",
                "ImageMagick",
                "iMatix",
                "Imlib2",
                "Info-ZIP",
                "Inner-Net-2.0",
                "InnoSetup",
                "Intel",
                "Intel-ACPI",
                "Interbase-1.0",
                "IPA",
                "IPL-1.0",
                "ISC",
                "ISC-Veillard",
                "Jam",
                "JasPer-2.0",
                "jove",
                "JPL-image",
                "JPNIC",
                "JSON",
                "Kastrup",
                "Kazlib",
                "Knuth-CTAN",
                "LAL-1.2",
                "LAL-1.3",
                "Latex2e",
                "Latex2e-translated-notice",
                "Leptonica",
                "LGPL-2.0-only",
                "LGPL-2.0-or-later",
                "LGPL-2.1-only",
                "LGPL-2.1-or-later",
                "LGPL-3.0-only",
                "LGPL-3.0-or-later",
                "LGPLLR",
                "Libpng",
                "libpng-1.6.35",
                "libpng-2.0",
                "libselinux-1.0",
                "libtiff",
                "libutil-David-Nugent",
                "LiLiQ-P-1.1",
                "LiLiQ-R-1.1",
                "LiLiQ-Rplus-1.1",
                "Linux-man-pages-1-para",
                "Linux-man-pages-copyleft",
                "Linux-man-pages-copyleft-2-para",
                "Linux-man-pages-copyleft-var",
                "Linux-OpenIB",
                "LOOP",
                "LPD-document",
                "LPL-1.0",
                "LPL-1.02",
                "LPPL-1.0",
                "LPPL-1.1",
                "LPPL-1.2",
                "LPPL-1.3a",
                "LPPL-1.3c",
                "lsof",
                "Lucida-Bitmap-Fonts",
                "LZMA-SDK-9.11-to-9.20",
                "LZMA-SDK-9.22",
                "Mackerras-3-Clause",
                "Mackerras-3-Clause-acknowledgment",
                "magaz",
                "mailprio",
                "MakeIndex",
                "man2html",
                "Martin-Birgmeier",
                "McPhee-slideshow",
                "metamail",
                "Minpack",
                "MIPS",
                "MirOS",
                "MIT",
                "MIT-0",
                "MIT-advertising",
                "MIT-Click",
                "MIT-CMU",
                "MIT-enna",
                "MIT-feh",
                "MIT-Festival",
                "MIT-Khronos-old",
                "MIT-Modern-Variant",
                "MIT-open-group",
                "MIT-testregex",
                "MIT-Wu",
                "MITNFA",
                "MMIXware",
                "Motosoto",
                "MPEG-SSG",
                "mpi-permissive",
                "mpich2",
                "MPL-1.0",
                "MPL-1.1",
                "MPL-2.0",
                "MPL-2.0-no-copyleft-exception",
                "mplus",
                "MS-LPL",
                "MS-PL",
                "MS-RL",
                "MTLL",
                "MulanPSL-1.0",
                "MulanPSL-2.0",
                "Multics",
                "Mup",
                "NAIST-2003",
                "NASA-1.3",
                "Naumen",
                "NBPL-1.0",
                "NCBI-PD",
                "NCGL-UK-2.0",
                "NCL",
                "NCSA",
                "NetCDF",
                "Newsletr",
                "NGPL",
                "ngrep",
                "NICTA-1.0",
                "NIST-PD",
                "NIST-PD-fallback",
                "NIST-Software",
                "NLOD-1.0",
                "NLOD-2.0",
                "NLPL",
                "Nokia",
                "NOSL",
                "Noweb",
                "NPL-1.0",
                "NPL-1.1",
                "NPOSL-3.0",
                "NRL",
                "NTIA-PD",
                "NTP",
                "NTP-0",
                "O-UDA-1.0",
                "OAR",
                "OCCT-PL",
                "OCLC-2.0",
                "ODbL-1.0",
                "ODC-By-1.0",
                "OFFIS",
                "OFL-1.0",
                "OFL-1.0-no-RFN",
                "OFL-1.0-RFN",
                "OFL-1.1",
                "OFL-1.1-no-RFN",
                "OFL-1.1-RFN",
                "OGC-1.0",
                "OGDL-Taiwan-1.0",
                "OGL-Canada-2.0",
                "OGL-UK-1.0",
                "OGL-UK-2.0",
                "OGL-UK-3.0",
                "OGTSL",
                "OLDAP-1.1",
                "OLDAP-1.2",
                "OLDAP-1.3",
                "OLDAP-1.4",
                "OLDAP-2.0",
                "OLDAP-2.0.1",
                "OLDAP-2.1",
                "OLDAP-2.2",
                "OLDAP-2.2.1",
                "OLDAP-2.2.2",
                "OLDAP-2.3",
                "OLDAP-2.4",
                "OLDAP-2.5",
                "OLDAP-2.6",
                "OLDAP-2.7",
                "OLDAP-2.8",
                "OLFL-1.3",
                "OML",
                "OpenPBS-2.3",
                "OpenSSL",
                "OpenSSL-standalone",
                "OpenVision",
                "OPL-1.0",
                "OPL-UK-3.0",
                "OPUBL-1.0",
                "OSET-PL-2.1",
                "OSL-1.0",
                "OSL-1.1",
                "OSL-2.0",
                "OSL-2.1",
                "OSL-3.0",
                "PADL",
                "Parity-6.0.0",
                "Parity-7.0.0",
                "PDDL-1.0",
                "PHP-3.0",
                "PHP-3.01",
                "Pixar",
                "pkgconf",
                "Plexus",
                "pnmstitch",
                "PolyForm-Noncommercial-1.0.0",
                "PolyForm-Small-Business-1.0.0",
                "PostgreSQL",
                "PPL",
                "PSF-2.0",
                "psfrag",
                "psutils",
                "Python-2.0",
                "Python-2.0.1",
                "python-ldap",
                "Qhull",
                "QPL-1.0",
                "QPL-1.0-INRIA-2004",
                "radvd",
                "Rdisc",
                "RHeCos-1.1",
                "RPL-1.1",
                "RPL-1.5",
                "RPSL-1.0",
                "RSA-MD",
                "RSCPL",
                "Ruby",
                "Ruby-pty",
                "SAX-PD",
                "SAX-PD-2.0",
                "Saxpath",
                "SCEA",
                "SchemeReport",
                "Sendmail",
                "Sendmail-8.23",
                "Sendmail-Open-Source-1.1",
                "SGI-B-1.0",
                "SGI-B-1.1",
                "SGI-B-2.0",
                "SGI-OpenGL",
                "SGP4",
                "SHL-0.5",
                "SHL-0.51",
                "SimPL-2.0",
                "SISSL",
                "SISSL-1.2",
                "SL",
                "Sleepycat",
                "SMAIL-GPL",
                "SMLNJ",
                "SMPPL",
                "SNIA",
                "snprintf",
                "SOFA",
                "softSurfer",
                "Soundex",
                "Spencer-86",
                "Spencer-94",
                "Spencer-99",
                "SPL-1.0",
                "ssh-keyscan",
                "SSH-OpenSSH",
                "SSH-short",
                "SSLeay-standalone",
                "SSPL-1.0",
                "SugarCRM-1.1.3",
                "SUL-1.0",
                "Sun-PPP",
                "Sun-PPP-2000",
                "SunPro",
                "SWL",
                "swrule",
                "Symlinks",
                "TAPR-OHL-1.0",
                "TCL",
                "TCP-wrappers",
                "TermReadKey",
                "TGPPL-1.0",
                "ThirdEye",
                "threeparttable",
                "TMate",
                "TORQUE-1.1",
                "TOSL",
                "TPDL",
                "TPL-1.0",
                "TrustedQSL",
                "TTWL",
                "TTYP0",
                "TU-Berlin-1.0",
                "TU-Berlin-2.0",
                "Ubuntu-font-1.0",
                "UCAR",
                "UCL-1.0",
                "ulem",
                "UMich-Merit",
                "Unicode-3.0",
                "Unicode-DFS-2015",
                "Unicode-DFS-2016",
                "Unicode-TOU",
                "UnixCrypt",
                "Unlicense",
                "Unlicense-libtelnet",
                "Unlicense-libwhirlpool",
                "UPL-1.0",
                "URT-RLE",
                "Vim",
                "VOSTROM",
                "VSL-1.0",
                "W3C",
                "W3C-19980720",
                "W3C-20150513",
                "w3m",
                "Watcom-1.0",
                "Widget-Workshop",
                "Wsuipa",
                "WTFPL",
                "wwl",
                "X11",
                "X11-distribute-modifications-variant",
                "X11-swapped",
                "Xdebug-1.03",
                "Xerox",
                "Xfig",
                "XFree86-1.1",
                "xinetd",
                "xkeyboard-config-Zinoviev",
                "xlock",
                "Xnet",
                "xpp",
                "XSkat",
                "xzoom",
                "YPL-1.0",
                "YPL-1.1",
                "Zed",
                "Zeeff",
                "Zend-2.0",
                "Zimbra-1.3",
                "Zimbra-1.4",
                "Zlib",
                "zlib-acknowledgement",
                "ZPL-1.1",
                "ZPL-2.0",
                "ZPL-2.1"
              ],
              "title": "LicenseId",
              "type": "string"
            },
            {
              "enum": [
                "AGPL-1.0",
                "AGPL-3.0",
                "BSD-2-Clause-FreeBSD",
                "BSD-2-Clause-NetBSD",
                "bzip2-1.0.5",
                "eCos-2.0",
                "GFDL-1.1",
                "GFDL-1.2",
                "GFDL-1.3",
                "GPL-1.0",
                "GPL-1.0+",
                "GPL-2.0",
                "GPL-2.0+",
                "GPL-2.0-with-autoconf-exception",
                "GPL-2.0-with-bison-exception",
                "GPL-2.0-with-classpath-exception",
                "GPL-2.0-with-font-exception",
                "GPL-2.0-with-GCC-exception",
                "GPL-3.0",
                "GPL-3.0+",
                "GPL-3.0-with-autoconf-exception",
                "GPL-3.0-with-GCC-exception",
                "LGPL-2.0",
                "LGPL-2.0+",
                "LGPL-2.1",
                "LGPL-2.1+",
                "LGPL-3.0",
                "LGPL-3.0+",
                "Net-SNMP",
                "Nunit",
                "StandardML-NJ",
                "wxWindows"
              ],
              "title": "DeprecatedLicenseId",
              "type": "string"
            },
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "A [SPDX license identifier](https://spdx.org/licenses/).\nWe do not support custom license beyond the SPDX license list, if you need that please\n[open a GitHub issue](https://github.com/bioimage-io/spec-bioimage-io/issues/new/choose\n) to discuss your intentions with the community.",
          "examples": [
            "CC0-1.0",
            "MIT",
            "BSD-2-Clause"
          ],
          "title": "License"
        },
        "type": {
          "const": "dataset",
          "title": "Type",
          "type": "string"
        },
        "id": {
          "anyOf": [
            {
              "minLength": 1,
              "title": "DatasetId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "bioimage.io-wide unique resource identifier\nassigned by bioimage.io; version **un**specific.",
          "title": "Id"
        },
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "\"URL to the source of the dataset.",
          "title": "Source"
        }
      },
      "required": [
        "name",
        "description",
        "format_version",
        "type"
      ],
      "title": "dataset 0.2.4",
      "type": "object"
    },
    "bioimageio__spec__dataset__v0_3__DatasetDescr": {
      "additionalProperties": false,
      "description": "A bioimage.io dataset resource description file (dataset RDF) describes a dataset relevant to bioimage\nprocessing.",
      "properties": {
        "name": {
          "description": "A human-friendly name of the resource description.\nMay only contains letters, digits, underscore, minus, parentheses and spaces.",
          "maxLength": 128,
          "minLength": 5,
          "title": "Name",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A string containing a brief description.",
          "maxLength": 1024,
          "title": "Description",
          "type": "string"
        },
        "covers": {
          "description": "Cover images. Please use an image smaller than 500KB and an aspect ratio width to height of 2:1 or 1:1.\nThe supported image formats are: ('.gif', '.jpeg', '.jpg', '.png', '.svg')",
          "examples": [
            [
              {
                "source": "cover.png"
              }
            ]
          ],
          "items": {
            "$ref": "#/$defs/FileDescr"
          },
          "title": "Covers",
          "type": "array"
        },
        "documentation": {
          "anyOf": [
            {
              "$ref": "#/$defs/FileDescr",
              "examples": [
                {
                  "source": "README.md"
                }
              ]
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Additional model documentation.\nThe recommended documentation source file name is `README.md`. An `.md` suffix is mandatory."
        },
        "id_emoji": {
          "anyOf": [
            {
              "examples": [
                "\ud83e\udd88",
                "\ud83e\udda5"
              ],
              "maxLength": 2,
              "minLength": 1,
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "UTF-8 emoji for display alongside the `id`.",
          "title": "Id Emoji"
        },
        "authors": {
          "description": "The authors are the creators of this resource description and the primary points of contact.",
          "items": {
            "$ref": "#/$defs/bioimageio__spec__generic__v0_3__Author"
          },
          "title": "Authors",
          "type": "array"
        },
        "attachments": {
          "description": "file attachments",
          "items": {
            "$ref": "#/$defs/FileDescr"
          },
          "title": "Attachments",
          "type": "array"
        },
        "cite": {
          "description": "citations",
          "items": {
            "$ref": "#/$defs/bioimageio__spec__generic__v0_3__CiteEntry"
          },
          "title": "Cite",
          "type": "array"
        },
        "license": {
          "anyOf": [
            {
              "enum": [
                "0BSD",
                "3D-Slicer-1.0",
                "AAL",
                "Abstyles",
                "AdaCore-doc",
                "Adobe-2006",
                "Adobe-Display-PostScript",
                "Adobe-Glyph",
                "Adobe-Utopia",
                "ADSL",
                "AFL-1.1",
                "AFL-1.2",
                "AFL-2.0",
                "AFL-2.1",
                "AFL-3.0",
                "Afmparse",
                "AGPL-1.0-only",
                "AGPL-1.0-or-later",
                "AGPL-3.0-only",
                "AGPL-3.0-or-later",
                "Aladdin",
                "AMD-newlib",
                "AMDPLPA",
                "AML",
                "AML-glslang",
                "AMPAS",
                "ANTLR-PD",
                "ANTLR-PD-fallback",
                "any-OSI",
                "any-OSI-perl-modules",
                "Apache-1.0",
                "Apache-1.1",
                "Apache-2.0",
                "APAFML",
                "APL-1.0",
                "App-s2p",
                "APSL-1.0",
                "APSL-1.1",
                "APSL-1.2",
                "APSL-2.0",
                "Arphic-1999",
                "Artistic-1.0",
                "Artistic-1.0-cl8",
                "Artistic-1.0-Perl",
                "Artistic-2.0",
                "Artistic-dist",
                "Aspell-RU",
                "ASWF-Digital-Assets-1.0",
                "ASWF-Digital-Assets-1.1",
                "Baekmuk",
                "Bahyph",
                "Barr",
                "bcrypt-Solar-Designer",
                "Beerware",
                "Bitstream-Charter",
                "Bitstream-Vera",
                "BitTorrent-1.0",
                "BitTorrent-1.1",
                "blessing",
                "BlueOak-1.0.0",
                "Boehm-GC",
                "Boehm-GC-without-fee",
                "Borceux",
                "Brian-Gladman-2-Clause",
                "Brian-Gladman-3-Clause",
                "BSD-1-Clause",
                "BSD-2-Clause",
                "BSD-2-Clause-Darwin",
                "BSD-2-Clause-first-lines",
                "BSD-2-Clause-Patent",
                "BSD-2-Clause-pkgconf-disclaimer",
                "BSD-2-Clause-Views",
                "BSD-3-Clause",
                "BSD-3-Clause-acpica",
                "BSD-3-Clause-Attribution",
                "BSD-3-Clause-Clear",
                "BSD-3-Clause-flex",
                "BSD-3-Clause-HP",
                "BSD-3-Clause-LBNL",
                "BSD-3-Clause-Modification",
                "BSD-3-Clause-No-Military-License",
                "BSD-3-Clause-No-Nuclear-License",
                "BSD-3-Clause-No-Nuclear-License-2014",
                "BSD-3-Clause-No-Nuclear-Warranty",
                "BSD-3-Clause-Open-MPI",
                "BSD-3-Clause-Sun",
                "BSD-4-Clause",
                "BSD-4-Clause-Shortened",
                "BSD-4-Clause-UC",
                "BSD-4.3RENO",
                "BSD-4.3TAHOE",
                "BSD-Advertising-Acknowledgement",
                "BSD-Attribution-HPND-disclaimer",
                "BSD-Inferno-Nettverk",
                "BSD-Protection",
                "BSD-Source-beginning-file",
                "BSD-Source-Code",
                "BSD-Systemics",
                "BSD-Systemics-W3Works",
                "BSL-1.0",
                "BUSL-1.1",
                "bzip2-1.0.6",
                "C-UDA-1.0",
                "CAL-1.0",
                "CAL-1.0-Combined-Work-Exception",
                "Caldera",
                "Caldera-no-preamble",
                "Catharon",
                "CATOSL-1.1",
                "CC-BY-1.0",
                "CC-BY-2.0",
                "CC-BY-2.5",
                "CC-BY-2.5-AU",
                "CC-BY-3.0",
                "CC-BY-3.0-AT",
                "CC-BY-3.0-AU",
                "CC-BY-3.0-DE",
                "CC-BY-3.0-IGO",
                "CC-BY-3.0-NL",
                "CC-BY-3.0-US",
                "CC-BY-4.0",
                "CC-BY-NC-1.0",
                "CC-BY-NC-2.0",
                "CC-BY-NC-2.5",
                "CC-BY-NC-3.0",
                "CC-BY-NC-3.0-DE",
                "CC-BY-NC-4.0",
                "CC-BY-NC-ND-1.0",
                "CC-BY-NC-ND-2.0",
                "CC-BY-NC-ND-2.5",
                "CC-BY-NC-ND-3.0",
                "CC-BY-NC-ND-3.0-DE",
                "CC-BY-NC-ND-3.0-IGO",
                "CC-BY-NC-ND-4.0",
                "CC-BY-NC-SA-1.0",
                "CC-BY-NC-SA-2.0",
                "CC-BY-NC-SA-2.0-DE",
                "CC-BY-NC-SA-2.0-FR",
                "CC-BY-NC-SA-2.0-UK",
                "CC-BY-NC-SA-2.5",
                "CC-BY-NC-SA-3.0",
                "CC-BY-NC-SA-3.0-DE",
                "CC-BY-NC-SA-3.0-IGO",
                "CC-BY-NC-SA-4.0",
                "CC-BY-ND-1.0",
                "CC-BY-ND-2.0",
                "CC-BY-ND-2.5",
                "CC-BY-ND-3.0",
                "CC-BY-ND-3.0-DE",
                "CC-BY-ND-4.0",
                "CC-BY-SA-1.0",
                "CC-BY-SA-2.0",
                "CC-BY-SA-2.0-UK",
                "CC-BY-SA-2.1-JP",
                "CC-BY-SA-2.5",
                "CC-BY-SA-3.0",
                "CC-BY-SA-3.0-AT",
                "CC-BY-SA-3.0-DE",
                "CC-BY-SA-3.0-IGO",
                "CC-BY-SA-4.0",
                "CC-PDDC",
                "CC-PDM-1.0",
                "CC-SA-1.0",
                "CC0-1.0",
                "CDDL-1.0",
                "CDDL-1.1",
                "CDL-1.0",
                "CDLA-Permissive-1.0",
                "CDLA-Permissive-2.0",
                "CDLA-Sharing-1.0",
                "CECILL-1.0",
                "CECILL-1.1",
                "CECILL-2.0",
                "CECILL-2.1",
                "CECILL-B",
                "CECILL-C",
                "CERN-OHL-1.1",
                "CERN-OHL-1.2",
                "CERN-OHL-P-2.0",
                "CERN-OHL-S-2.0",
                "CERN-OHL-W-2.0",
                "CFITSIO",
                "check-cvs",
                "checkmk",
                "ClArtistic",
                "Clips",
                "CMU-Mach",
                "CMU-Mach-nodoc",
                "CNRI-Jython",
                "CNRI-Python",
                "CNRI-Python-GPL-Compatible",
                "COIL-1.0",
                "Community-Spec-1.0",
                "Condor-1.1",
                "copyleft-next-0.3.0",
                "copyleft-next-0.3.1",
                "Cornell-Lossless-JPEG",
                "CPAL-1.0",
                "CPL-1.0",
                "CPOL-1.02",
                "Cronyx",
                "Crossword",
                "CryptoSwift",
                "CrystalStacker",
                "CUA-OPL-1.0",
                "Cube",
                "curl",
                "cve-tou",
                "D-FSL-1.0",
                "DEC-3-Clause",
                "diffmark",
                "DL-DE-BY-2.0",
                "DL-DE-ZERO-2.0",
                "DOC",
                "DocBook-DTD",
                "DocBook-Schema",
                "DocBook-Stylesheet",
                "DocBook-XML",
                "Dotseqn",
                "DRL-1.0",
                "DRL-1.1",
                "DSDP",
                "dtoa",
                "dvipdfm",
                "ECL-1.0",
                "ECL-2.0",
                "EFL-1.0",
                "EFL-2.0",
                "eGenix",
                "Elastic-2.0",
                "Entessa",
                "EPICS",
                "EPL-1.0",
                "EPL-2.0",
                "ErlPL-1.1",
                "etalab-2.0",
                "EUDatagrid",
                "EUPL-1.0",
                "EUPL-1.1",
                "EUPL-1.2",
                "Eurosym",
                "Fair",
                "FBM",
                "FDK-AAC",
                "Ferguson-Twofish",
                "Frameworx-1.0",
                "FreeBSD-DOC",
                "FreeImage",
                "FSFAP",
                "FSFAP-no-warranty-disclaimer",
                "FSFUL",
                "FSFULLR",
                "FSFULLRSD",
                "FSFULLRWD",
                "FSL-1.1-ALv2",
                "FSL-1.1-MIT",
                "FTL",
                "Furuseth",
                "fwlw",
                "Game-Programming-Gems",
                "GCR-docs",
                "GD",
                "generic-xts",
                "GFDL-1.1-invariants-only",
                "GFDL-1.1-invariants-or-later",
                "GFDL-1.1-no-invariants-only",
                "GFDL-1.1-no-invariants-or-later",
                "GFDL-1.1-only",
                "GFDL-1.1-or-later",
                "GFDL-1.2-invariants-only",
                "GFDL-1.2-invariants-or-later",
                "GFDL-1.2-no-invariants-only",
                "GFDL-1.2-no-invariants-or-later",
                "GFDL-1.2-only",
                "GFDL-1.2-or-later",
                "GFDL-1.3-invariants-only",
                "GFDL-1.3-invariants-or-later",
                "GFDL-1.3-no-invariants-only",
                "GFDL-1.3-no-invariants-or-later",
                "GFDL-1.3-only",
                "GFDL-1.3-or-later",
                "Giftware",
                "GL2PS",
                "Glide",
                "Glulxe",
                "GLWTPL",
                "gnuplot",
                "GPL-1.0-only",
                "GPL-1.0-or-later",
                "GPL-2.0-only",
                "GPL-2.0-or-later",
                "GPL-3.0-only",
                "GPL-3.0-or-later",
                "Graphics-Gems",
                "gSOAP-1.3b",
                "gtkbook",
                "Gutmann",
                "HaskellReport",
                "HDF5",
                "hdparm",
                "HIDAPI",
                "Hippocratic-2.1",
                "HP-1986",
                "HP-1989",
                "HPND",
                "HPND-DEC",
                "HPND-doc",
                "HPND-doc-sell",
                "HPND-export-US",
                "HPND-export-US-acknowledgement",
                "HPND-export-US-modify",
                "HPND-export2-US",
                "HPND-Fenneberg-Livingston",
                "HPND-INRIA-IMAG",
                "HPND-Intel",
                "HPND-Kevlin-Henney",
                "HPND-Markus-Kuhn",
                "HPND-merchantability-variant",
                "HPND-MIT-disclaimer",
                "HPND-Netrek",
                "HPND-Pbmplus",
                "HPND-sell-MIT-disclaimer-xserver",
                "HPND-sell-regexpr",
                "HPND-sell-variant",
                "HPND-sell-variant-MIT-disclaimer",
                "HPND-sell-variant-MIT-disclaimer-rev",
                "HPND-UC",
                "HPND-UC-export-US",
                "HTMLTIDY",
                "IBM-pibs",
                "ICU",
                "IEC-Code-Components-EULA",
                "IJG",
                "IJG-short",
                "ImageMagick",
                "iMatix",
                "Imlib2",
                "Info-ZIP",
                "Inner-Net-2.0",
                "InnoSetup",
                "Intel",
                "Intel-ACPI",
                "Interbase-1.0",
                "IPA",
                "IPL-1.0",
                "ISC",
                "ISC-Veillard",
                "Jam",
                "JasPer-2.0",
                "jove",
                "JPL-image",
                "JPNIC",
                "JSON",
                "Kastrup",
                "Kazlib",
                "Knuth-CTAN",
                "LAL-1.2",
                "LAL-1.3",
                "Latex2e",
                "Latex2e-translated-notice",
                "Leptonica",
                "LGPL-2.0-only",
                "LGPL-2.0-or-later",
                "LGPL-2.1-only",
                "LGPL-2.1-or-later",
                "LGPL-3.0-only",
                "LGPL-3.0-or-later",
                "LGPLLR",
                "Libpng",
                "libpng-1.6.35",
                "libpng-2.0",
                "libselinux-1.0",
                "libtiff",
                "libutil-David-Nugent",
                "LiLiQ-P-1.1",
                "LiLiQ-R-1.1",
                "LiLiQ-Rplus-1.1",
                "Linux-man-pages-1-para",
                "Linux-man-pages-copyleft",
                "Linux-man-pages-copyleft-2-para",
                "Linux-man-pages-copyleft-var",
                "Linux-OpenIB",
                "LOOP",
                "LPD-document",
                "LPL-1.0",
                "LPL-1.02",
                "LPPL-1.0",
                "LPPL-1.1",
                "LPPL-1.2",
                "LPPL-1.3a",
                "LPPL-1.3c",
                "lsof",
                "Lucida-Bitmap-Fonts",
                "LZMA-SDK-9.11-to-9.20",
                "LZMA-SDK-9.22",
                "Mackerras-3-Clause",
                "Mackerras-3-Clause-acknowledgment",
                "magaz",
                "mailprio",
                "MakeIndex",
                "man2html",
                "Martin-Birgmeier",
                "McPhee-slideshow",
                "metamail",
                "Minpack",
                "MIPS",
                "MirOS",
                "MIT",
                "MIT-0",
                "MIT-advertising",
                "MIT-Click",
                "MIT-CMU",
                "MIT-enna",
                "MIT-feh",
                "MIT-Festival",
                "MIT-Khronos-old",
                "MIT-Modern-Variant",
                "MIT-open-group",
                "MIT-testregex",
                "MIT-Wu",
                "MITNFA",
                "MMIXware",
                "Motosoto",
                "MPEG-SSG",
                "mpi-permissive",
                "mpich2",
                "MPL-1.0",
                "MPL-1.1",
                "MPL-2.0",
                "MPL-2.0-no-copyleft-exception",
                "mplus",
                "MS-LPL",
                "MS-PL",
                "MS-RL",
                "MTLL",
                "MulanPSL-1.0",
                "MulanPSL-2.0",
                "Multics",
                "Mup",
                "NAIST-2003",
                "NASA-1.3",
                "Naumen",
                "NBPL-1.0",
                "NCBI-PD",
                "NCGL-UK-2.0",
                "NCL",
                "NCSA",
                "NetCDF",
                "Newsletr",
                "NGPL",
                "ngrep",
                "NICTA-1.0",
                "NIST-PD",
                "NIST-PD-fallback",
                "NIST-Software",
                "NLOD-1.0",
                "NLOD-2.0",
                "NLPL",
                "Nokia",
                "NOSL",
                "Noweb",
                "NPL-1.0",
                "NPL-1.1",
                "NPOSL-3.0",
                "NRL",
                "NTIA-PD",
                "NTP",
                "NTP-0",
                "O-UDA-1.0",
                "OAR",
                "OCCT-PL",
                "OCLC-2.0",
                "ODbL-1.0",
                "ODC-By-1.0",
                "OFFIS",
                "OFL-1.0",
                "OFL-1.0-no-RFN",
                "OFL-1.0-RFN",
                "OFL-1.1",
                "OFL-1.1-no-RFN",
                "OFL-1.1-RFN",
                "OGC-1.0",
                "OGDL-Taiwan-1.0",
                "OGL-Canada-2.0",
                "OGL-UK-1.0",
                "OGL-UK-2.0",
                "OGL-UK-3.0",
                "OGTSL",
                "OLDAP-1.1",
                "OLDAP-1.2",
                "OLDAP-1.3",
                "OLDAP-1.4",
                "OLDAP-2.0",
                "OLDAP-2.0.1",
                "OLDAP-2.1",
                "OLDAP-2.2",
                "OLDAP-2.2.1",
                "OLDAP-2.2.2",
                "OLDAP-2.3",
                "OLDAP-2.4",
                "OLDAP-2.5",
                "OLDAP-2.6",
                "OLDAP-2.7",
                "OLDAP-2.8",
                "OLFL-1.3",
                "OML",
                "OpenPBS-2.3",
                "OpenSSL",
                "OpenSSL-standalone",
                "OpenVision",
                "OPL-1.0",
                "OPL-UK-3.0",
                "OPUBL-1.0",
                "OSET-PL-2.1",
                "OSL-1.0",
                "OSL-1.1",
                "OSL-2.0",
                "OSL-2.1",
                "OSL-3.0",
                "PADL",
                "Parity-6.0.0",
                "Parity-7.0.0",
                "PDDL-1.0",
                "PHP-3.0",
                "PHP-3.01",
                "Pixar",
                "pkgconf",
                "Plexus",
                "pnmstitch",
                "PolyForm-Noncommercial-1.0.0",
                "PolyForm-Small-Business-1.0.0",
                "PostgreSQL",
                "PPL",
                "PSF-2.0",
                "psfrag",
                "psutils",
                "Python-2.0",
                "Python-2.0.1",
                "python-ldap",
                "Qhull",
                "QPL-1.0",
                "QPL-1.0-INRIA-2004",
                "radvd",
                "Rdisc",
                "RHeCos-1.1",
                "RPL-1.1",
                "RPL-1.5",
                "RPSL-1.0",
                "RSA-MD",
                "RSCPL",
                "Ruby",
                "Ruby-pty",
                "SAX-PD",
                "SAX-PD-2.0",
                "Saxpath",
                "SCEA",
                "SchemeReport",
                "Sendmail",
                "Sendmail-8.23",
                "Sendmail-Open-Source-1.1",
                "SGI-B-1.0",
                "SGI-B-1.1",
                "SGI-B-2.0",
                "SGI-OpenGL",
                "SGP4",
                "SHL-0.5",
                "SHL-0.51",
                "SimPL-2.0",
                "SISSL",
                "SISSL-1.2",
                "SL",
                "Sleepycat",
                "SMAIL-GPL",
                "SMLNJ",
                "SMPPL",
                "SNIA",
                "snprintf",
                "SOFA",
                "softSurfer",
                "Soundex",
                "Spencer-86",
                "Spencer-94",
                "Spencer-99",
                "SPL-1.0",
                "ssh-keyscan",
                "SSH-OpenSSH",
                "SSH-short",
                "SSLeay-standalone",
                "SSPL-1.0",
                "SugarCRM-1.1.3",
                "SUL-1.0",
                "Sun-PPP",
                "Sun-PPP-2000",
                "SunPro",
                "SWL",
                "swrule",
                "Symlinks",
                "TAPR-OHL-1.0",
                "TCL",
                "TCP-wrappers",
                "TermReadKey",
                "TGPPL-1.0",
                "ThirdEye",
                "threeparttable",
                "TMate",
                "TORQUE-1.1",
                "TOSL",
                "TPDL",
                "TPL-1.0",
                "TrustedQSL",
                "TTWL",
                "TTYP0",
                "TU-Berlin-1.0",
                "TU-Berlin-2.0",
                "Ubuntu-font-1.0",
                "UCAR",
                "UCL-1.0",
                "ulem",
                "UMich-Merit",
                "Unicode-3.0",
                "Unicode-DFS-2015",
                "Unicode-DFS-2016",
                "Unicode-TOU",
                "UnixCrypt",
                "Unlicense",
                "Unlicense-libtelnet",
                "Unlicense-libwhirlpool",
                "UPL-1.0",
                "URT-RLE",
                "Vim",
                "VOSTROM",
                "VSL-1.0",
                "W3C",
                "W3C-19980720",
                "W3C-20150513",
                "w3m",
                "Watcom-1.0",
                "Widget-Workshop",
                "Wsuipa",
                "WTFPL",
                "wwl",
                "X11",
                "X11-distribute-modifications-variant",
                "X11-swapped",
                "Xdebug-1.03",
                "Xerox",
                "Xfig",
                "XFree86-1.1",
                "xinetd",
                "xkeyboard-config-Zinoviev",
                "xlock",
                "Xnet",
                "xpp",
                "XSkat",
                "xzoom",
                "YPL-1.0",
                "YPL-1.1",
                "Zed",
                "Zeeff",
                "Zend-2.0",
                "Zimbra-1.3",
                "Zimbra-1.4",
                "Zlib",
                "zlib-acknowledgement",
                "ZPL-1.1",
                "ZPL-2.0",
                "ZPL-2.1"
              ],
              "title": "LicenseId",
              "type": "string"
            },
            {
              "enum": [
                "AGPL-1.0",
                "AGPL-3.0",
                "BSD-2-Clause-FreeBSD",
                "BSD-2-Clause-NetBSD",
                "bzip2-1.0.5",
                "eCos-2.0",
                "GFDL-1.1",
                "GFDL-1.2",
                "GFDL-1.3",
                "GPL-1.0",
                "GPL-1.0+",
                "GPL-2.0",
                "GPL-2.0+",
                "GPL-2.0-with-autoconf-exception",
                "GPL-2.0-with-bison-exception",
                "GPL-2.0-with-classpath-exception",
                "GPL-2.0-with-font-exception",
                "GPL-2.0-with-GCC-exception",
                "GPL-3.0",
                "GPL-3.0+",
                "GPL-3.0-with-autoconf-exception",
                "GPL-3.0-with-GCC-exception",
                "LGPL-2.0",
                "LGPL-2.0+",
                "LGPL-2.1",
                "LGPL-2.1+",
                "LGPL-3.0",
                "LGPL-3.0+",
                "Net-SNMP",
                "Nunit",
                "StandardML-NJ",
                "wxWindows"
              ],
              "title": "DeprecatedLicenseId",
              "type": "string"
            },
            {
              "$ref": "#/$defs/FileDescr"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "A [SPDX license identifier](https://spdx.org/licenses/) or a custom license file.",
          "examples": [
            "CC0-1.0",
            "MIT",
            "BSD-2-Clause"
          ],
          "title": "License"
        },
        "git_repo": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "A URL to the Git repository where the resource is being developed.",
          "examples": [
            "https://github.com/bioimage-io/spec-bioimage-io/tree/main/example_descriptions/models/unet2d_nuclei_broad"
          ],
          "title": "Git Repo"
        },
        "icon": {
          "anyOf": [
            {
              "maxLength": 2,
              "minLength": 1,
              "type": "string"
            },
            {
              "$ref": "#/$defs/FileDescr"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "An icon for illustration, e.g. on bioimage.io",
          "title": "Icon"
        },
        "links": {
          "description": "IDs of other bioimage.io resources",
          "examples": [
            [
              "ilastik/ilastik",
              "deepimagej/deepimagej",
              "zero/notebook_u-net_3d_zerocostdl4mic"
            ]
          ],
          "items": {
            "type": "string"
          },
          "title": "Links",
          "type": "array"
        },
        "uploader": {
          "anyOf": [
            {
              "$ref": "#/$defs/Uploader"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The person who uploaded the model (e.g. to bioimage.io)"
        },
        "maintainers": {
          "description": "Maintainers of this resource.\nIf not specified, `authors` are maintainers and at least some of them has to specify their `github_user` name",
          "items": {
            "$ref": "#/$defs/bioimageio__spec__generic__v0_3__Maintainer"
          },
          "title": "Maintainers",
          "type": "array"
        },
        "tags": {
          "description": "Associated tags",
          "examples": [
            [
              "unet2d",
              "pytorch",
              "nucleus",
              "segmentation",
              "dsb2018"
            ]
          ],
          "items": {
            "type": "string"
          },
          "title": "Tags",
          "type": "array"
        },
        "version": {
          "anyOf": [
            {
              "$ref": "#/$defs/Version"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The version of the resource following SemVer 2.0."
        },
        "version_comment": {
          "anyOf": [
            {
              "maxLength": 512,
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "A comment on the version of the resource.",
          "title": "Version Comment"
        },
        "format_version": {
          "const": "0.3.4",
          "description": "The **format** version of this resource specification",
          "title": "Format Version",
          "type": "string"
        },
        "badges": {
          "description": "badges associated with this resource",
          "items": {
            "$ref": "#/$defs/BadgeDescr"
          },
          "title": "Badges",
          "type": "array"
        },
        "config": {
          "$ref": "#/$defs/bioimageio__spec__generic__v0_3__Config",
          "description": "A field for custom configuration that can contain any keys not present in the RDF spec.\nThis means you should not store, for example, a GitHub repo URL in `config` since there is a `git_repo` field.\nKeys in `config` may be very specific to a tool or consumer software. To avoid conflicting definitions,\nit is recommended to wrap added configuration into a sub-field named with the specific domain or tool name,\nfor example:\n```yaml\nconfig:\n    giraffe_neckometer:  # here is the domain name\n        length: 3837283\n        address:\n            home: zoo\n    imagej:              # config specific to ImageJ\n        macro_dir: path/to/macro/file\n```\nIf possible, please use [`snake_case`](https://en.wikipedia.org/wiki/Snake_case) for keys in `config`.\nYou may want to list linked files additionally under `attachments` to include them when packaging a resource.\n(Packaging a resource means downloading/copying important linked files and creating a ZIP archive that contains\nan altered rdf.yaml file with local references to the downloaded files.)"
        },
        "type": {
          "const": "dataset",
          "title": "Type",
          "type": "string"
        },
        "id": {
          "anyOf": [
            {
              "minLength": 1,
              "title": "DatasetId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "bioimage.io-wide unique resource identifier\nassigned by bioimage.io; version **un**specific.",
          "title": "Id"
        },
        "parent": {
          "anyOf": [
            {
              "minLength": 1,
              "title": "DatasetId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The description from which this one is derived",
          "title": "Parent"
        },
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "\"URL to the source of the dataset.",
          "title": "Source"
        }
      },
      "required": [
        "name",
        "format_version",
        "type"
      ],
      "title": "dataset 0.3.4",
      "type": "object"
    },
    "bioimageio__spec__generic__v0_2__Author": {
      "additionalProperties": false,
      "properties": {
        "affiliation": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Affiliation",
          "title": "Affiliation"
        },
        "email": {
          "anyOf": [
            {
              "format": "email",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Email",
          "title": "Email"
        },
        "orcid": {
          "anyOf": [
            {
              "description": "An ORCID identifier, see https://orcid.org/",
              "title": "OrcidId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "An [ORCID iD](https://support.orcid.org/hc/en-us/sections/360001495313-What-is-ORCID\n) in hyphenated groups of 4 digits, (and [valid](\nhttps://support.orcid.org/hc/en-us/articles/360006897674-Structure-of-the-ORCID-Identifier\n) as per ISO 7064 11,2.)",
          "examples": [
            "0000-0001-2345-6789"
          ],
          "title": "Orcid"
        },
        "name": {
          "title": "Name",
          "type": "string"
        },
        "github_user": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Github User"
        }
      },
      "required": [
        "name"
      ],
      "title": "generic.v0_2.Author",
      "type": "object"
    },
    "bioimageio__spec__generic__v0_2__CiteEntry": {
      "additionalProperties": false,
      "properties": {
        "text": {
          "description": "free text description",
          "title": "Text",
          "type": "string"
        },
        "doi": {
          "anyOf": [
            {
              "description": "A digital object identifier, see https://www.doi.org/",
              "pattern": "^10\\.[0-9]{4}.+$",
              "title": "Doi",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "A digital object identifier (DOI) is the prefered citation reference.\nSee https://www.doi.org/ for details. (alternatively specify `url`)",
          "title": "Doi"
        },
        "url": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "URL to cite (preferably specify a `doi` instead)",
          "title": "Url"
        }
      },
      "required": [
        "text"
      ],
      "title": "generic.v0_2.CiteEntry",
      "type": "object"
    },
    "bioimageio__spec__generic__v0_2__Maintainer": {
      "additionalProperties": false,
      "properties": {
        "affiliation": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Affiliation",
          "title": "Affiliation"
        },
        "email": {
          "anyOf": [
            {
              "format": "email",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Email",
          "title": "Email"
        },
        "orcid": {
          "anyOf": [
            {
              "description": "An ORCID identifier, see https://orcid.org/",
              "title": "OrcidId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "An [ORCID iD](https://support.orcid.org/hc/en-us/sections/360001495313-What-is-ORCID\n) in hyphenated groups of 4 digits, (and [valid](\nhttps://support.orcid.org/hc/en-us/articles/360006897674-Structure-of-the-ORCID-Identifier\n) as per ISO 7064 11,2.)",
          "examples": [
            "0000-0001-2345-6789"
          ],
          "title": "Orcid"
        },
        "name": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Name"
        },
        "github_user": {
          "title": "Github User",
          "type": "string"
        }
      },
      "required": [
        "github_user"
      ],
      "title": "generic.v0_2.Maintainer",
      "type": "object"
    },
    "bioimageio__spec__generic__v0_3__Author": {
      "additionalProperties": false,
      "properties": {
        "affiliation": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Affiliation",
          "title": "Affiliation"
        },
        "email": {
          "anyOf": [
            {
              "format": "email",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Email",
          "title": "Email"
        },
        "orcid": {
          "anyOf": [
            {
              "description": "An ORCID identifier, see https://orcid.org/",
              "title": "OrcidId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "An [ORCID iD](https://support.orcid.org/hc/en-us/sections/360001495313-What-is-ORCID\n) in hyphenated groups of 4 digits, (and [valid](\nhttps://support.orcid.org/hc/en-us/articles/360006897674-Structure-of-the-ORCID-Identifier\n) as per ISO 7064 11,2.)",
          "examples": [
            "0000-0001-2345-6789"
          ],
          "title": "Orcid"
        },
        "name": {
          "title": "Name",
          "type": "string"
        },
        "github_user": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Github User"
        }
      },
      "required": [
        "name"
      ],
      "title": "generic.v0_3.Author",
      "type": "object"
    },
    "bioimageio__spec__generic__v0_3__BioimageioConfig": {
      "additionalProperties": true,
      "description": "bioimage.io internal metadata.",
      "properties": {},
      "title": "generic.v0_3.BioimageioConfig",
      "type": "object"
    },
    "bioimageio__spec__generic__v0_3__CiteEntry": {
      "additionalProperties": false,
      "description": "A citation that should be referenced in work using this resource.",
      "properties": {
        "text": {
          "description": "free text description",
          "title": "Text",
          "type": "string"
        },
        "doi": {
          "anyOf": [
            {
              "description": "A digital object identifier, see https://www.doi.org/",
              "pattern": "^10\\.[0-9]{4}.+$",
              "title": "Doi",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "A digital object identifier (DOI) is the prefered citation reference.\nSee https://www.doi.org/ for details.\nNote:\n    Either **doi** or **url** have to be specified.",
          "title": "Doi"
        },
        "url": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "URL to cite (preferably specify a **doi** instead/also).\nNote:\n    Either **doi** or **url** have to be specified.",
          "title": "Url"
        }
      },
      "required": [
        "text"
      ],
      "title": "generic.v0_3.CiteEntry",
      "type": "object"
    },
    "bioimageio__spec__generic__v0_3__Config": {
      "additionalProperties": true,
      "description": "A place to store additional metadata (often tool specific).\n\nSuch additional metadata is typically set programmatically by the respective tool\nor by people with specific insights into the tool.\nIf you want to store additional metadata that does not match any of the other\nfields, think of a key unlikely to collide with anyone elses use-case/tool and save\nit here.\n\nPlease consider creating [an issue in the bioimageio.spec repository](https://github.com/bioimage-io/spec-bioimage-io/issues/new?template=Blank+issue)\nif you are not sure if an existing field could cover your use case\nor if you think such a field should exist.",
      "properties": {
        "bioimageio": {
          "$ref": "#/$defs/bioimageio__spec__generic__v0_3__BioimageioConfig"
        }
      },
      "title": "generic.v0_3.Config",
      "type": "object"
    },
    "bioimageio__spec__generic__v0_3__Maintainer": {
      "additionalProperties": false,
      "properties": {
        "affiliation": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Affiliation",
          "title": "Affiliation"
        },
        "email": {
          "anyOf": [
            {
              "format": "email",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Email",
          "title": "Email"
        },
        "orcid": {
          "anyOf": [
            {
              "description": "An ORCID identifier, see https://orcid.org/",
              "title": "OrcidId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "An [ORCID iD](https://support.orcid.org/hc/en-us/sections/360001495313-What-is-ORCID\n) in hyphenated groups of 4 digits, (and [valid](\nhttps://support.orcid.org/hc/en-us/articles/360006897674-Structure-of-the-ORCID-Identifier\n) as per ISO 7064 11,2.)",
          "examples": [
            "0000-0001-2345-6789"
          ],
          "title": "Orcid"
        },
        "name": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Name"
        },
        "github_user": {
          "title": "Github User",
          "type": "string"
        }
      },
      "required": [
        "github_user"
      ],
      "title": "generic.v0_3.Maintainer",
      "type": "object"
    },
    "bioimageio__spec__model__v0_5__BioimageioConfig": {
      "additionalProperties": true,
      "properties": {
        "reproducibility_tolerance": {
          "description": "Tolerances to allow when reproducing the model's test outputs\nfrom the model's test inputs.\nOnly the first entry matching tensor id and weights format is considered.\nFor defaults see [`ContinuousReproducibilityTolerance`][ContinuousReproducibilityTolerance]/[`DiscreteReproducibilityTolerance`][DiscreteReproducibilityTolerance]",
          "items": {
            "anyOf": [
              {
                "$ref": "#/$defs/ReproducibilityTolerance"
              },
              {
                "$ref": "#/$defs/ContinuousReproducibilityTolerance"
              },
              {
                "$ref": "#/$defs/DiscreteReproducibilityTolerance"
              }
            ]
          },
          "title": "Reproducibility Tolerance",
          "type": "array"
        },
        "funded_by": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Funding agency, grant number if applicable",
          "title": "Funded By"
        },
        "architecture_type": {
          "anyOf": [
            {
              "maxLength": 32,
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Model architecture type, e.g., 3D U-Net, ResNet, transformer",
          "title": "Architecture Type"
        },
        "architecture_description": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Text description of model architecture.",
          "title": "Architecture Description"
        },
        "modality": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Input modality, e.g., fluorescence microscopy, electron microscopy",
          "title": "Modality"
        },
        "target_structure": {
          "description": "Biological structure(s) the model is designed to analyze, e.g., nuclei, mitochondria, cells",
          "items": {
            "type": "string"
          },
          "title": "Target Structure",
          "type": "array"
        },
        "task": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Bioimage-specific task type, e.g., segmentation, classification, detection, denoising",
          "title": "Task"
        },
        "new_version": {
          "anyOf": [
            {
              "minLength": 1,
              "title": "ModelId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "A new version of this model exists with a different model id.",
          "title": "New Version"
        },
        "out_of_scope_use": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Describe how the model may be misused in bioimage analysis contexts and what users should **not** do with the model.",
          "title": "Out Of Scope Use"
        },
        "bias_risks_limitations": {
          "$ref": "#/$defs/BiasRisksLimitations",
          "description": "Description of known bias, risks, and technical limitations for in-scope model use."
        },
        "model_parameter_count": {
          "anyOf": [
            {
              "type": "integer"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Total number of model parameters.",
          "title": "Model Parameter Count"
        },
        "training": {
          "$ref": "#/$defs/TrainingDetails",
          "description": "Details on how the model was trained."
        },
        "inference_time": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Average inference time per image/tile. Specify hardware and image size. Multiple examples can be given.",
          "title": "Inference Time"
        },
        "memory_requirements_inference": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "GPU memory needed for inference. Multiple examples with different image size can be given.",
          "title": "Memory Requirements Inference"
        },
        "memory_requirements_training": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "GPU memory needed for training. Multiple examples with different image/batch sizes can be given.",
          "title": "Memory Requirements Training"
        },
        "evaluations": {
          "description": "Quantitative model evaluations.\n\nNote:\n    At the moment we recommend to include only a single test dataset\n    (with evaluation factors that may mark subsets of the dataset)\n    to avoid confusion and make the presentation of results cleaner.",
          "items": {
            "$ref": "#/$defs/Evaluation"
          },
          "title": "Evaluations",
          "type": "array"
        },
        "environmental_impact": {
          "$ref": "#/$defs/EnvironmentalImpact",
          "description": "Environmental considerations for model training and deployment"
        }
      },
      "title": "model.v0_5.BioimageioConfig",
      "type": "object"
    },
    "bioimageio__spec__model__v0_5__Config": {
      "additionalProperties": true,
      "properties": {
        "bioimageio": {
          "$ref": "#/$defs/bioimageio__spec__model__v0_5__BioimageioConfig"
        },
        "stardist": {
          "$ref": "#/$defs/YamlValue",
          "default": null
        }
      },
      "title": "model.v0_5.Config",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "description": "Specification of the fields used in a bioimage.io-compliant RDF to describe AI models with pretrained weights.\nThese fields are typically stored in a YAML file which we call a model resource description file (model RDF).",
  "properties": {
    "name": {
      "description": "A human-readable name of this model.\nIt should be no longer than 64 characters\nand may only contain letter, number, underscore, minus, parentheses and spaces.\nWe recommend to chose a name that refers to the model's task and image modality.",
      "maxLength": 128,
      "minLength": 5,
      "title": "Name",
      "type": "string"
    },
    "description": {
      "default": "",
      "description": "A string containing a brief description.",
      "maxLength": 1024,
      "title": "Description",
      "type": "string"
    },
    "covers": {
      "description": "Cover images. Please use an image smaller than 500KB and an aspect ratio width to height of 2:1 or 1:1.\nThe supported image formats are: ('.gif', '.jpeg', '.jpg', '.png', '.svg')",
      "examples": [
        [
          {
            "source": "cover.png"
          }
        ]
      ],
      "items": {
        "$ref": "#/$defs/FileDescr"
      },
      "title": "Covers",
      "type": "array"
    },
    "documentation": {
      "anyOf": [
        {
          "$ref": "#/$defs/FileDescr",
          "examples": [
            {
              "source": "README.md"
            }
          ]
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Additional model documentation.\nThe recommended documentation source file name is `README.md`. An `.md` suffix is mandatory.\nThe documentation should include a '#[#] Validation' (sub)section\nwith details on how to quantitatively validate the model on unseen data."
    },
    "id_emoji": {
      "anyOf": [
        {
          "examples": [
            "\ud83e\udd88",
            "\ud83e\udda5"
          ],
          "maxLength": 2,
          "minLength": 1,
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "UTF-8 emoji for display alongside the `id`.",
      "title": "Id Emoji"
    },
    "authors": {
      "description": "The authors are the creators of the model RDF and the primary points of contact.",
      "items": {
        "$ref": "#/$defs/bioimageio__spec__generic__v0_3__Author"
      },
      "title": "Authors",
      "type": "array"
    },
    "attachments": {
      "description": "file attachments",
      "items": {
        "$ref": "#/$defs/FileDescr"
      },
      "title": "Attachments",
      "type": "array"
    },
    "cite": {
      "description": "citations",
      "items": {
        "$ref": "#/$defs/bioimageio__spec__generic__v0_3__CiteEntry"
      },
      "title": "Cite",
      "type": "array"
    },
    "license": {
      "anyOf": [
        {
          "enum": [
            "0BSD",
            "3D-Slicer-1.0",
            "AAL",
            "Abstyles",
            "AdaCore-doc",
            "Adobe-2006",
            "Adobe-Display-PostScript",
            "Adobe-Glyph",
            "Adobe-Utopia",
            "ADSL",
            "AFL-1.1",
            "AFL-1.2",
            "AFL-2.0",
            "AFL-2.1",
            "AFL-3.0",
            "Afmparse",
            "AGPL-1.0-only",
            "AGPL-1.0-or-later",
            "AGPL-3.0-only",
            "AGPL-3.0-or-later",
            "Aladdin",
            "AMD-newlib",
            "AMDPLPA",
            "AML",
            "AML-glslang",
            "AMPAS",
            "ANTLR-PD",
            "ANTLR-PD-fallback",
            "any-OSI",
            "any-OSI-perl-modules",
            "Apache-1.0",
            "Apache-1.1",
            "Apache-2.0",
            "APAFML",
            "APL-1.0",
            "App-s2p",
            "APSL-1.0",
            "APSL-1.1",
            "APSL-1.2",
            "APSL-2.0",
            "Arphic-1999",
            "Artistic-1.0",
            "Artistic-1.0-cl8",
            "Artistic-1.0-Perl",
            "Artistic-2.0",
            "Artistic-dist",
            "Aspell-RU",
            "ASWF-Digital-Assets-1.0",
            "ASWF-Digital-Assets-1.1",
            "Baekmuk",
            "Bahyph",
            "Barr",
            "bcrypt-Solar-Designer",
            "Beerware",
            "Bitstream-Charter",
            "Bitstream-Vera",
            "BitTorrent-1.0",
            "BitTorrent-1.1",
            "blessing",
            "BlueOak-1.0.0",
            "Boehm-GC",
            "Boehm-GC-without-fee",
            "Borceux",
            "Brian-Gladman-2-Clause",
            "Brian-Gladman-3-Clause",
            "BSD-1-Clause",
            "BSD-2-Clause",
            "BSD-2-Clause-Darwin",
            "BSD-2-Clause-first-lines",
            "BSD-2-Clause-Patent",
            "BSD-2-Clause-pkgconf-disclaimer",
            "BSD-2-Clause-Views",
            "BSD-3-Clause",
            "BSD-3-Clause-acpica",
            "BSD-3-Clause-Attribution",
            "BSD-3-Clause-Clear",
            "BSD-3-Clause-flex",
            "BSD-3-Clause-HP",
            "BSD-3-Clause-LBNL",
            "BSD-3-Clause-Modification",
            "BSD-3-Clause-No-Military-License",
            "BSD-3-Clause-No-Nuclear-License",
            "BSD-3-Clause-No-Nuclear-License-2014",
            "BSD-3-Clause-No-Nuclear-Warranty",
            "BSD-3-Clause-Open-MPI",
            "BSD-3-Clause-Sun",
            "BSD-4-Clause",
            "BSD-4-Clause-Shortened",
            "BSD-4-Clause-UC",
            "BSD-4.3RENO",
            "BSD-4.3TAHOE",
            "BSD-Advertising-Acknowledgement",
            "BSD-Attribution-HPND-disclaimer",
            "BSD-Inferno-Nettverk",
            "BSD-Protection",
            "BSD-Source-beginning-file",
            "BSD-Source-Code",
            "BSD-Systemics",
            "BSD-Systemics-W3Works",
            "BSL-1.0",
            "BUSL-1.1",
            "bzip2-1.0.6",
            "C-UDA-1.0",
            "CAL-1.0",
            "CAL-1.0-Combined-Work-Exception",
            "Caldera",
            "Caldera-no-preamble",
            "Catharon",
            "CATOSL-1.1",
            "CC-BY-1.0",
            "CC-BY-2.0",
            "CC-BY-2.5",
            "CC-BY-2.5-AU",
            "CC-BY-3.0",
            "CC-BY-3.0-AT",
            "CC-BY-3.0-AU",
            "CC-BY-3.0-DE",
            "CC-BY-3.0-IGO",
            "CC-BY-3.0-NL",
            "CC-BY-3.0-US",
            "CC-BY-4.0",
            "CC-BY-NC-1.0",
            "CC-BY-NC-2.0",
            "CC-BY-NC-2.5",
            "CC-BY-NC-3.0",
            "CC-BY-NC-3.0-DE",
            "CC-BY-NC-4.0",
            "CC-BY-NC-ND-1.0",
            "CC-BY-NC-ND-2.0",
            "CC-BY-NC-ND-2.5",
            "CC-BY-NC-ND-3.0",
            "CC-BY-NC-ND-3.0-DE",
            "CC-BY-NC-ND-3.0-IGO",
            "CC-BY-NC-ND-4.0",
            "CC-BY-NC-SA-1.0",
            "CC-BY-NC-SA-2.0",
            "CC-BY-NC-SA-2.0-DE",
            "CC-BY-NC-SA-2.0-FR",
            "CC-BY-NC-SA-2.0-UK",
            "CC-BY-NC-SA-2.5",
            "CC-BY-NC-SA-3.0",
            "CC-BY-NC-SA-3.0-DE",
            "CC-BY-NC-SA-3.0-IGO",
            "CC-BY-NC-SA-4.0",
            "CC-BY-ND-1.0",
            "CC-BY-ND-2.0",
            "CC-BY-ND-2.5",
            "CC-BY-ND-3.0",
            "CC-BY-ND-3.0-DE",
            "CC-BY-ND-4.0",
            "CC-BY-SA-1.0",
            "CC-BY-SA-2.0",
            "CC-BY-SA-2.0-UK",
            "CC-BY-SA-2.1-JP",
            "CC-BY-SA-2.5",
            "CC-BY-SA-3.0",
            "CC-BY-SA-3.0-AT",
            "CC-BY-SA-3.0-DE",
            "CC-BY-SA-3.0-IGO",
            "CC-BY-SA-4.0",
            "CC-PDDC",
            "CC-PDM-1.0",
            "CC-SA-1.0",
            "CC0-1.0",
            "CDDL-1.0",
            "CDDL-1.1",
            "CDL-1.0",
            "CDLA-Permissive-1.0",
            "CDLA-Permissive-2.0",
            "CDLA-Sharing-1.0",
            "CECILL-1.0",
            "CECILL-1.1",
            "CECILL-2.0",
            "CECILL-2.1",
            "CECILL-B",
            "CECILL-C",
            "CERN-OHL-1.1",
            "CERN-OHL-1.2",
            "CERN-OHL-P-2.0",
            "CERN-OHL-S-2.0",
            "CERN-OHL-W-2.0",
            "CFITSIO",
            "check-cvs",
            "checkmk",
            "ClArtistic",
            "Clips",
            "CMU-Mach",
            "CMU-Mach-nodoc",
            "CNRI-Jython",
            "CNRI-Python",
            "CNRI-Python-GPL-Compatible",
            "COIL-1.0",
            "Community-Spec-1.0",
            "Condor-1.1",
            "copyleft-next-0.3.0",
            "copyleft-next-0.3.1",
            "Cornell-Lossless-JPEG",
            "CPAL-1.0",
            "CPL-1.0",
            "CPOL-1.02",
            "Cronyx",
            "Crossword",
            "CryptoSwift",
            "CrystalStacker",
            "CUA-OPL-1.0",
            "Cube",
            "curl",
            "cve-tou",
            "D-FSL-1.0",
            "DEC-3-Clause",
            "diffmark",
            "DL-DE-BY-2.0",
            "DL-DE-ZERO-2.0",
            "DOC",
            "DocBook-DTD",
            "DocBook-Schema",
            "DocBook-Stylesheet",
            "DocBook-XML",
            "Dotseqn",
            "DRL-1.0",
            "DRL-1.1",
            "DSDP",
            "dtoa",
            "dvipdfm",
            "ECL-1.0",
            "ECL-2.0",
            "EFL-1.0",
            "EFL-2.0",
            "eGenix",
            "Elastic-2.0",
            "Entessa",
            "EPICS",
            "EPL-1.0",
            "EPL-2.0",
            "ErlPL-1.1",
            "etalab-2.0",
            "EUDatagrid",
            "EUPL-1.0",
            "EUPL-1.1",
            "EUPL-1.2",
            "Eurosym",
            "Fair",
            "FBM",
            "FDK-AAC",
            "Ferguson-Twofish",
            "Frameworx-1.0",
            "FreeBSD-DOC",
            "FreeImage",
            "FSFAP",
            "FSFAP-no-warranty-disclaimer",
            "FSFUL",
            "FSFULLR",
            "FSFULLRSD",
            "FSFULLRWD",
            "FSL-1.1-ALv2",
            "FSL-1.1-MIT",
            "FTL",
            "Furuseth",
            "fwlw",
            "Game-Programming-Gems",
            "GCR-docs",
            "GD",
            "generic-xts",
            "GFDL-1.1-invariants-only",
            "GFDL-1.1-invariants-or-later",
            "GFDL-1.1-no-invariants-only",
            "GFDL-1.1-no-invariants-or-later",
            "GFDL-1.1-only",
            "GFDL-1.1-or-later",
            "GFDL-1.2-invariants-only",
            "GFDL-1.2-invariants-or-later",
            "GFDL-1.2-no-invariants-only",
            "GFDL-1.2-no-invariants-or-later",
            "GFDL-1.2-only",
            "GFDL-1.2-or-later",
            "GFDL-1.3-invariants-only",
            "GFDL-1.3-invariants-or-later",
            "GFDL-1.3-no-invariants-only",
            "GFDL-1.3-no-invariants-or-later",
            "GFDL-1.3-only",
            "GFDL-1.3-or-later",
            "Giftware",
            "GL2PS",
            "Glide",
            "Glulxe",
            "GLWTPL",
            "gnuplot",
            "GPL-1.0-only",
            "GPL-1.0-or-later",
            "GPL-2.0-only",
            "GPL-2.0-or-later",
            "GPL-3.0-only",
            "GPL-3.0-or-later",
            "Graphics-Gems",
            "gSOAP-1.3b",
            "gtkbook",
            "Gutmann",
            "HaskellReport",
            "HDF5",
            "hdparm",
            "HIDAPI",
            "Hippocratic-2.1",
            "HP-1986",
            "HP-1989",
            "HPND",
            "HPND-DEC",
            "HPND-doc",
            "HPND-doc-sell",
            "HPND-export-US",
            "HPND-export-US-acknowledgement",
            "HPND-export-US-modify",
            "HPND-export2-US",
            "HPND-Fenneberg-Livingston",
            "HPND-INRIA-IMAG",
            "HPND-Intel",
            "HPND-Kevlin-Henney",
            "HPND-Markus-Kuhn",
            "HPND-merchantability-variant",
            "HPND-MIT-disclaimer",
            "HPND-Netrek",
            "HPND-Pbmplus",
            "HPND-sell-MIT-disclaimer-xserver",
            "HPND-sell-regexpr",
            "HPND-sell-variant",
            "HPND-sell-variant-MIT-disclaimer",
            "HPND-sell-variant-MIT-disclaimer-rev",
            "HPND-UC",
            "HPND-UC-export-US",
            "HTMLTIDY",
            "IBM-pibs",
            "ICU",
            "IEC-Code-Components-EULA",
            "IJG",
            "IJG-short",
            "ImageMagick",
            "iMatix",
            "Imlib2",
            "Info-ZIP",
            "Inner-Net-2.0",
            "InnoSetup",
            "Intel",
            "Intel-ACPI",
            "Interbase-1.0",
            "IPA",
            "IPL-1.0",
            "ISC",
            "ISC-Veillard",
            "Jam",
            "JasPer-2.0",
            "jove",
            "JPL-image",
            "JPNIC",
            "JSON",
            "Kastrup",
            "Kazlib",
            "Knuth-CTAN",
            "LAL-1.2",
            "LAL-1.3",
            "Latex2e",
            "Latex2e-translated-notice",
            "Leptonica",
            "LGPL-2.0-only",
            "LGPL-2.0-or-later",
            "LGPL-2.1-only",
            "LGPL-2.1-or-later",
            "LGPL-3.0-only",
            "LGPL-3.0-or-later",
            "LGPLLR",
            "Libpng",
            "libpng-1.6.35",
            "libpng-2.0",
            "libselinux-1.0",
            "libtiff",
            "libutil-David-Nugent",
            "LiLiQ-P-1.1",
            "LiLiQ-R-1.1",
            "LiLiQ-Rplus-1.1",
            "Linux-man-pages-1-para",
            "Linux-man-pages-copyleft",
            "Linux-man-pages-copyleft-2-para",
            "Linux-man-pages-copyleft-var",
            "Linux-OpenIB",
            "LOOP",
            "LPD-document",
            "LPL-1.0",
            "LPL-1.02",
            "LPPL-1.0",
            "LPPL-1.1",
            "LPPL-1.2",
            "LPPL-1.3a",
            "LPPL-1.3c",
            "lsof",
            "Lucida-Bitmap-Fonts",
            "LZMA-SDK-9.11-to-9.20",
            "LZMA-SDK-9.22",
            "Mackerras-3-Clause",
            "Mackerras-3-Clause-acknowledgment",
            "magaz",
            "mailprio",
            "MakeIndex",
            "man2html",
            "Martin-Birgmeier",
            "McPhee-slideshow",
            "metamail",
            "Minpack",
            "MIPS",
            "MirOS",
            "MIT",
            "MIT-0",
            "MIT-advertising",
            "MIT-Click",
            "MIT-CMU",
            "MIT-enna",
            "MIT-feh",
            "MIT-Festival",
            "MIT-Khronos-old",
            "MIT-Modern-Variant",
            "MIT-open-group",
            "MIT-testregex",
            "MIT-Wu",
            "MITNFA",
            "MMIXware",
            "Motosoto",
            "MPEG-SSG",
            "mpi-permissive",
            "mpich2",
            "MPL-1.0",
            "MPL-1.1",
            "MPL-2.0",
            "MPL-2.0-no-copyleft-exception",
            "mplus",
            "MS-LPL",
            "MS-PL",
            "MS-RL",
            "MTLL",
            "MulanPSL-1.0",
            "MulanPSL-2.0",
            "Multics",
            "Mup",
            "NAIST-2003",
            "NASA-1.3",
            "Naumen",
            "NBPL-1.0",
            "NCBI-PD",
            "NCGL-UK-2.0",
            "NCL",
            "NCSA",
            "NetCDF",
            "Newsletr",
            "NGPL",
            "ngrep",
            "NICTA-1.0",
            "NIST-PD",
            "NIST-PD-fallback",
            "NIST-Software",
            "NLOD-1.0",
            "NLOD-2.0",
            "NLPL",
            "Nokia",
            "NOSL",
            "Noweb",
            "NPL-1.0",
            "NPL-1.1",
            "NPOSL-3.0",
            "NRL",
            "NTIA-PD",
            "NTP",
            "NTP-0",
            "O-UDA-1.0",
            "OAR",
            "OCCT-PL",
            "OCLC-2.0",
            "ODbL-1.0",
            "ODC-By-1.0",
            "OFFIS",
            "OFL-1.0",
            "OFL-1.0-no-RFN",
            "OFL-1.0-RFN",
            "OFL-1.1",
            "OFL-1.1-no-RFN",
            "OFL-1.1-RFN",
            "OGC-1.0",
            "OGDL-Taiwan-1.0",
            "OGL-Canada-2.0",
            "OGL-UK-1.0",
            "OGL-UK-2.0",
            "OGL-UK-3.0",
            "OGTSL",
            "OLDAP-1.1",
            "OLDAP-1.2",
            "OLDAP-1.3",
            "OLDAP-1.4",
            "OLDAP-2.0",
            "OLDAP-2.0.1",
            "OLDAP-2.1",
            "OLDAP-2.2",
            "OLDAP-2.2.1",
            "OLDAP-2.2.2",
            "OLDAP-2.3",
            "OLDAP-2.4",
            "OLDAP-2.5",
            "OLDAP-2.6",
            "OLDAP-2.7",
            "OLDAP-2.8",
            "OLFL-1.3",
            "OML",
            "OpenPBS-2.3",
            "OpenSSL",
            "OpenSSL-standalone",
            "OpenVision",
            "OPL-1.0",
            "OPL-UK-3.0",
            "OPUBL-1.0",
            "OSET-PL-2.1",
            "OSL-1.0",
            "OSL-1.1",
            "OSL-2.0",
            "OSL-2.1",
            "OSL-3.0",
            "PADL",
            "Parity-6.0.0",
            "Parity-7.0.0",
            "PDDL-1.0",
            "PHP-3.0",
            "PHP-3.01",
            "Pixar",
            "pkgconf",
            "Plexus",
            "pnmstitch",
            "PolyForm-Noncommercial-1.0.0",
            "PolyForm-Small-Business-1.0.0",
            "PostgreSQL",
            "PPL",
            "PSF-2.0",
            "psfrag",
            "psutils",
            "Python-2.0",
            "Python-2.0.1",
            "python-ldap",
            "Qhull",
            "QPL-1.0",
            "QPL-1.0-INRIA-2004",
            "radvd",
            "Rdisc",
            "RHeCos-1.1",
            "RPL-1.1",
            "RPL-1.5",
            "RPSL-1.0",
            "RSA-MD",
            "RSCPL",
            "Ruby",
            "Ruby-pty",
            "SAX-PD",
            "SAX-PD-2.0",
            "Saxpath",
            "SCEA",
            "SchemeReport",
            "Sendmail",
            "Sendmail-8.23",
            "Sendmail-Open-Source-1.1",
            "SGI-B-1.0",
            "SGI-B-1.1",
            "SGI-B-2.0",
            "SGI-OpenGL",
            "SGP4",
            "SHL-0.5",
            "SHL-0.51",
            "SimPL-2.0",
            "SISSL",
            "SISSL-1.2",
            "SL",
            "Sleepycat",
            "SMAIL-GPL",
            "SMLNJ",
            "SMPPL",
            "SNIA",
            "snprintf",
            "SOFA",
            "softSurfer",
            "Soundex",
            "Spencer-86",
            "Spencer-94",
            "Spencer-99",
            "SPL-1.0",
            "ssh-keyscan",
            "SSH-OpenSSH",
            "SSH-short",
            "SSLeay-standalone",
            "SSPL-1.0",
            "SugarCRM-1.1.3",
            "SUL-1.0",
            "Sun-PPP",
            "Sun-PPP-2000",
            "SunPro",
            "SWL",
            "swrule",
            "Symlinks",
            "TAPR-OHL-1.0",
            "TCL",
            "TCP-wrappers",
            "TermReadKey",
            "TGPPL-1.0",
            "ThirdEye",
            "threeparttable",
            "TMate",
            "TORQUE-1.1",
            "TOSL",
            "TPDL",
            "TPL-1.0",
            "TrustedQSL",
            "TTWL",
            "TTYP0",
            "TU-Berlin-1.0",
            "TU-Berlin-2.0",
            "Ubuntu-font-1.0",
            "UCAR",
            "UCL-1.0",
            "ulem",
            "UMich-Merit",
            "Unicode-3.0",
            "Unicode-DFS-2015",
            "Unicode-DFS-2016",
            "Unicode-TOU",
            "UnixCrypt",
            "Unlicense",
            "Unlicense-libtelnet",
            "Unlicense-libwhirlpool",
            "UPL-1.0",
            "URT-RLE",
            "Vim",
            "VOSTROM",
            "VSL-1.0",
            "W3C",
            "W3C-19980720",
            "W3C-20150513",
            "w3m",
            "Watcom-1.0",
            "Widget-Workshop",
            "Wsuipa",
            "WTFPL",
            "wwl",
            "X11",
            "X11-distribute-modifications-variant",
            "X11-swapped",
            "Xdebug-1.03",
            "Xerox",
            "Xfig",
            "XFree86-1.1",
            "xinetd",
            "xkeyboard-config-Zinoviev",
            "xlock",
            "Xnet",
            "xpp",
            "XSkat",
            "xzoom",
            "YPL-1.0",
            "YPL-1.1",
            "Zed",
            "Zeeff",
            "Zend-2.0",
            "Zimbra-1.3",
            "Zimbra-1.4",
            "Zlib",
            "zlib-acknowledgement",
            "ZPL-1.1",
            "ZPL-2.0",
            "ZPL-2.1"
          ],
          "title": "LicenseId",
          "type": "string"
        },
        {
          "enum": [
            "AGPL-1.0",
            "AGPL-3.0",
            "BSD-2-Clause-FreeBSD",
            "BSD-2-Clause-NetBSD",
            "bzip2-1.0.5",
            "eCos-2.0",
            "GFDL-1.1",
            "GFDL-1.2",
            "GFDL-1.3",
            "GPL-1.0",
            "GPL-1.0+",
            "GPL-2.0",
            "GPL-2.0+",
            "GPL-2.0-with-autoconf-exception",
            "GPL-2.0-with-bison-exception",
            "GPL-2.0-with-classpath-exception",
            "GPL-2.0-with-font-exception",
            "GPL-2.0-with-GCC-exception",
            "GPL-3.0",
            "GPL-3.0+",
            "GPL-3.0-with-autoconf-exception",
            "GPL-3.0-with-GCC-exception",
            "LGPL-2.0",
            "LGPL-2.0+",
            "LGPL-2.1",
            "LGPL-2.1+",
            "LGPL-3.0",
            "LGPL-3.0+",
            "Net-SNMP",
            "Nunit",
            "StandardML-NJ",
            "wxWindows"
          ],
          "title": "DeprecatedLicenseId",
          "type": "string"
        },
        {
          "$ref": "#/$defs/FileDescr"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "A [SPDX license identifier](https://spdx.org/licenses/) or a custom license file.",
      "examples": [
        "CC0-1.0",
        "MIT",
        "BSD-2-Clause"
      ],
      "title": "License"
    },
    "git_repo": {
      "anyOf": [
        {
          "description": "A URL with the HTTP or HTTPS scheme.",
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "title": "HttpUrl",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "A URL to the Git repository where the resource is being developed.",
      "examples": [
        "https://github.com/bioimage-io/spec-bioimage-io/tree/main/example_descriptions/models/unet2d_nuclei_broad"
      ],
      "title": "Git Repo"
    },
    "icon": {
      "anyOf": [
        {
          "maxLength": 2,
          "minLength": 1,
          "type": "string"
        },
        {
          "$ref": "#/$defs/FileDescr"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "An icon for illustration, e.g. on bioimage.io",
      "title": "Icon"
    },
    "links": {
      "description": "IDs of other bioimage.io resources",
      "examples": [
        [
          "ilastik/ilastik",
          "deepimagej/deepimagej",
          "zero/notebook_u-net_3d_zerocostdl4mic"
        ]
      ],
      "items": {
        "type": "string"
      },
      "title": "Links",
      "type": "array"
    },
    "uploader": {
      "anyOf": [
        {
          "$ref": "#/$defs/Uploader"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The person who uploaded the model (e.g. to bioimage.io)"
    },
    "maintainers": {
      "description": "Maintainers of this resource.\nIf not specified, `authors` are maintainers and at least some of them has to specify their `github_user` name",
      "items": {
        "$ref": "#/$defs/bioimageio__spec__generic__v0_3__Maintainer"
      },
      "title": "Maintainers",
      "type": "array"
    },
    "tags": {
      "description": "Associated tags",
      "examples": [
        [
          "unet2d",
          "pytorch",
          "nucleus",
          "segmentation",
          "dsb2018"
        ]
      ],
      "items": {
        "type": "string"
      },
      "title": "Tags",
      "type": "array"
    },
    "version": {
      "anyOf": [
        {
          "$ref": "#/$defs/Version"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The version of the resource following SemVer 2.0."
    },
    "version_comment": {
      "anyOf": [
        {
          "maxLength": 512,
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "A comment on the version of the resource.",
      "title": "Version Comment"
    },
    "format_version": {
      "const": "0.5.14",
      "description": "Version of the bioimage.io model description specification used.\nWhen creating a new model always use the latest micro/patch version described here.\nThe `format_version` is important for any consumer software to understand how to parse the fields.",
      "title": "Format Version",
      "type": "string"
    },
    "type": {
      "const": "model",
      "description": "Specialized resource type 'model'",
      "title": "Type",
      "type": "string"
    },
    "id": {
      "anyOf": [
        {
          "minLength": 1,
          "title": "ModelId",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "bioimage.io-wide unique resource identifier\nassigned by bioimage.io; version **un**specific.",
      "title": "Id"
    },
    "inputs": {
      "description": "Describes the input tensors expected by this model.",
      "items": {
        "$ref": "#/$defs/InputTensorDescr"
      },
      "minItems": 1,
      "title": "Inputs",
      "type": "array"
    },
    "outputs": {
      "description": "Describes the output tensors.",
      "items": {
        "$ref": "#/$defs/OutputTensorDescr"
      },
      "minItems": 1,
      "title": "Outputs",
      "type": "array"
    },
    "packaged_by": {
      "description": "The persons that have packaged and uploaded this model.\nOnly required if those persons differ from the `authors`.",
      "items": {
        "$ref": "#/$defs/bioimageio__spec__generic__v0_3__Author"
      },
      "title": "Packaged By",
      "type": "array"
    },
    "parent": {
      "anyOf": [
        {
          "$ref": "#/$defs/LinkedModel"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The model from which this model is derived, e.g. by fine-tuning the weights."
    },
    "run_mode": {
      "anyOf": [
        {
          "$ref": "#/$defs/RunMode"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Custom run mode for this model: for more complex prediction procedures like test time\ndata augmentation that currently cannot be expressed in the specification.\nNo standard run modes are defined yet."
    },
    "timestamp": {
      "$ref": "#/$defs/Datetime",
      "description": "Timestamp in [ISO 8601](#https://en.wikipedia.org/wiki/ISO_8601) format\nwith a few restrictions listed [here](https://docs.python.org/3/library/datetime.html#datetime.datetime.fromisoformat).\n(In Python a datetime object is valid, too)."
    },
    "training_data": {
      "anyOf": [
        {
          "$ref": "#/$defs/LinkedDataset"
        },
        {
          "$ref": "#/$defs/bioimageio__spec__dataset__v0_3__DatasetDescr"
        },
        {
          "$ref": "#/$defs/bioimageio__spec__dataset__v0_2__DatasetDescr"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The dataset used to train this model",
      "title": "Training Data"
    },
    "weights": {
      "$ref": "#/$defs/WeightsDescr",
      "description": "The weights for this model.\nWeights can be given for different formats, but should otherwise be equivalent.\nThe available weight formats determine which consumers can use this model."
    },
    "config": {
      "$ref": "#/$defs/bioimageio__spec__model__v0_5__Config"
    }
  },
  "required": [
    "name",
    "format_version",
    "type",
    "inputs",
    "outputs",
    "weights"
  ],
  "title": "model 0.5.14",
  "type": "object"
}

Fields:

• _validation_summary (ValidationSummary | None)
• description (FAIR[Annotated[str, MaxLen(1024), warn(MaxLen(512), 'Description longer than 512 characters.')]])
• covers (list[_FileDescr_cover])
• id_emoji (Annotated[str, Len(min_length=1, max_length=2), Field(examples=['🦈', '🦥'])] | None)
• attachments (list[FileDescr_package])
• cite (FAIR[list[CiteEntry]])
• license (FAIR[Annotated[LicenseId | DeprecatedLicenseId | None | FileDescr_package, Field(union_mode='left_to_right', examples=['CC0-1.0', 'MIT', 'BSD-2-Clause'])]])
• git_repo (Annotated[HttpUrl | None, Field(examples=['https://github.com/bioimage-io/spec-bioimage-io/tree/main/example_descriptions/models/unet2d_nuclei_broad'])])
• icon (Annotated[str, Len(min_length=1, max_length=2)] | FileDescr_package | None)
• links (Annotated[list[str], Field(examples=[('ilastik/ilastik', 'deepimagej/deepimagej', 'zero/notebook_u-net_3d_zerocostdl4mic')])])
• uploader (Uploader | None)
• maintainers (list[Maintainer])
• tags (FAIR[Annotated[list[str], Field(examples=[('unet2d', 'pytorch', 'nucleus', 'segmentation', 'dsb2018')])]])
• version (Version | None)
• version_comment (Annotated[str, MaxLen(512)] | None)
• format_version (Literal['0.5.14'])
• type (Literal['model'])
• id (ModelId | None)
• authors (FAIR[list[Author]])
• documentation (FAIR[FileDescr_documentation | None])
• inputs (NotEmpty[Sequence[InputTensorDescr]])
• name (Annotated[str, RestrictCharacters(string.ascii_letters + string.digits + '_+- ()'), MinLen(5), MaxLen(128), warn(MaxLen(64), 'Name longer than 64 characters.', INFO)])
• outputs (NotEmpty[Sequence[OutputTensorDescr]])
• packaged_by (list[Author])
• parent (LinkedModel | None)
• run_mode (Annotated[RunMode | None, warn(None, "Run mode '{value}' has limited support across consumer softwares.")])
• timestamp (Datetime)
• training_data (Annotated[None | LinkedDataset | DatasetDescr | DatasetDescr02, Field(union_mode='left_to_right')])
• weights (Annotated[WeightsDescr, WrapSerializer(package_weights)])
• config (Config)

Validators:

• _validate_documentation → documentation
• _validate_input_axes → inputs
• _validate_test_tensors
• _validate_tensor_references_in_proc_kwargs
• _validate_tensor_ids → outputs
• _validate_output_axes → outputs
• _validate_parent_is_not_self
• _validate_reproducibility_tolerance_output_ids
• _add_default_cover
• _convert

attachments pydantic-field

attachments: list[FileDescr_package]

file attachments

authors pydantic-field

authors: FAIR[list[Author]]

The authors are the creators of the model RDF and the primary points of contact.

cite pydantic-field

cite: FAIR[list[CiteEntry]]

citations

config pydantic-field

config: Config

covers pydantic-field

covers: list[_FileDescr_cover]

Cover images.

description pydantic-field

description: FAIR[
    Annotated[
        str,
        MaxLen(1024),
        warn(
            MaxLen(512),
            "Description longer than 512 characters.",
        ),
    ]
] = ""

A string containing a brief description.

documentation pydantic-field

documentation: FAIR[FileDescr_documentation | None] = None

Additional model documentation. The recommended documentation source file name is README.md. An .md suffix is mandatory. The documentation should include a '#[#] Validation' (sub)section with details on how to quantitatively validate the model on unseen data.

file_name property

file_name: FileName | None

File name of the bioimageio.yaml file the description was loaded from.

format_version pydantic-field

format_version: Literal['0.5.14'] = '0.5.14'

git_repo pydantic-field

git_repo: Annotated[
    HttpUrl | None,
    Field(
        examples=[
            "https://github.com/bioimage-io/spec-bioimage-io/tree/main/example_descriptions/models/unet2d_nuclei_broad"
        ]
    ),
] = None

A URL to the Git repository where the resource is being developed.

icon pydantic-field

icon: (
    Annotated[str, Len(min_length=1, max_length=2)]
    | FileDescr_package
    | None
) = None

An icon for illustration, e.g. on bioimage.io

id pydantic-field

id: ModelId | None = None

bioimage.io-wide unique resource identifier assigned by bioimage.io; version unspecific.

id_emoji pydantic-field

id_emoji: (
    Annotated[
        str,
        Len(min_length=1, max_length=2),
        Field(examples=["🦈", "🦥"]),
    ]
    | None
) = None

UTF-8 emoji for display alongside the id.

implemented_format_version class-attribute

implemented_format_version: Literal['0.5.14'] = '0.5.14'

implemented_format_version_tuple class-attribute

implemented_format_version_tuple: tuple[int, int, int]

implemented_type class-attribute

implemented_type: Literal['model'] = 'model'

inputs pydantic-field

inputs: NotEmpty[Sequence[InputTensorDescr]]

Describes the input tensors expected by this model.

license pydantic-field

license: FAIR[
    Annotated[
        LicenseId
        | DeprecatedLicenseId
        | None
        | FileDescr_package,
        Field(
            union_mode="left_to_right",
            examples=["CC0-1.0", "MIT", "BSD-2-Clause"],
        ),
    ]
] = None

A SPDX license identifier or a custom license file.

links pydantic-field

links: Annotated[
    list[str],
    Field(
        examples=[
            (
                "ilastik/ilastik",
                "deepimagej/deepimagej",
                "zero/notebook_u-net_3d_zerocostdl4mic",
            )
        ]
    ),
]

IDs of other bioimage.io resources

maintainers pydantic-field

maintainers: list[Maintainer]

Maintainers of this resource. If not specified, authors are maintainers and at least some of them has to specify their github_user name

name pydantic-field

name: Annotated[
    str,
    RestrictCharacters(
        string.ascii_letters + string.digits + "_+- ()"
    ),
    MinLen(5),
    MaxLen(128),
    warn(
        MaxLen(64), "Name longer than 64 characters.", INFO
    ),
]

A human-readable name of this model. It should be no longer than 64 characters and may only contain letter, number, underscore, minus, parentheses and spaces. We recommend to chose a name that refers to the model's task and image modality.

outputs pydantic-field

outputs: NotEmpty[Sequence[OutputTensorDescr]]

Describes the output tensors.

packaged_by pydantic-field

packaged_by: list[Author]

The persons that have packaged and uploaded this model. Only required if those persons differ from the authors.

parent pydantic-field

parent: LinkedModel | None = None

The model from which this model is derived, e.g. by fine-tuning the weights.

root property

root: RootHttpUrl | DirectoryPath | ZipFile

The URL/Path prefix to resolve any relative paths with.

run_mode pydantic-field

run_mode: Annotated[
    RunMode | None,
    warn(
        None,
        "Run mode '{value}' has limited support across consumer softwares.",
    ),
] = None

Custom run mode for this model: for more complex prediction procedures like test time data augmentation that currently cannot be expressed in the specification. No standard run modes are defined yet.

tags pydantic-field

tags: FAIR[
    Annotated[
        list[str],
        Field(
            examples=[
                (
                    "unet2d",
                    "pytorch",
                    "nucleus",
                    "segmentation",
                    "dsb2018",
                )
            ]
        ),
    ]
]

Associated tags

timestamp pydantic-field

timestamp: Datetime

Timestamp in ISO 8601 format with a few restrictions listed here. (In Python a datetime object is valid, too).

training_data pydantic-field

training_data: Annotated[
    None | LinkedDataset | DatasetDescr | DatasetDescr02,
    Field(union_mode="left_to_right"),
] = None

The dataset used to train this model

type pydantic-field

type: Literal['model'] = 'model'

uploader pydantic-field

uploader: Uploader | None = None

The person who uploaded the model (e.g. to bioimage.io)

validation_summary property

validation_summary: ValidationSummary

version pydantic-field

version: Version | None = None

The version of the resource following SemVer 2.0.

version_comment pydantic-field

version_comment: Annotated[str, MaxLen(512)] | None = None

A comment on the version of the resource.

weights pydantic-field

weights: Annotated[
    WeightsDescr, WrapSerializer(package_weights)
]

The weights for this model. Weights can be given for different formats, but should otherwise be equivalent. The available weight formats determine which consumers can use this model.

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any)

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any):
    super().__pydantic_init_subclass__(**kwargs)
    # set classvar implemented_format_version_tuple
    if "format_version" in cls.model_fields:
        if "." not in cls.implemented_format_version:
            cls.implemented_format_version_tuple = (0, 0, 0)
        else:
            fv_tuple = get_format_version_tuple(cls.implemented_format_version)
            assert fv_tuple is not None, (
                f"failed to cast '{cls.implemented_format_version}' to tuple"
            )
            cls.implemented_format_version_tuple = fv_tuple

convert_from_old_format_wo_validation classmethod

convert_from_old_format_wo_validation(
    data: dict[str, Any],
) -> None

Convert metadata following an older format version to this classes' format without validating the result.

Source code in src/bioimageio/spec/model/v0_5.py

@classmethod
def convert_from_old_format_wo_validation(cls, data: dict[str, Any]) -> None:
    """Convert metadata following an older format version to this classes' format
    without validating the result.
    """
    if (
        data.get("type") == "model"
        and isinstance(fv := data.get("format_version"), str)
        and fv.count(".") == 2
    ):
        fv_parts = fv.split(".")
        if any(not p.isdigit() for p in fv_parts):
            return

        fv_tuple = tuple(map(int, fv_parts))

        assert cls.implemented_format_version_tuple[0:2] == (0, 5)
        if fv_tuple[:2] in ((0, 3), (0, 4)):
            m04 = _ModelDescr_v0_4.load(data)
            if isinstance(m04, InvalidDescr):
                try:
                    updated = _model_conv.convert_as_dict(
                        m04  # pyright: ignore[reportArgumentType]
                    )
                except Exception as e:
                    logger.error(
                        "Failed to convert from invalid model 0.4 description."
                        + f"\nerror: {e}"
                        + "\nProceeding with model 0.5 validation without conversion."
                    )
                    updated = None
            else:
                updated = _model_conv.convert_as_dict(m04)

            if updated is not None:
                data.clear()
                data.update(updated)

        elif fv_tuple[:2] == (0, 5):
            # bump patch version
            data["format_version"] = cls.implemented_format_version

        if fv_tuple[:2] in ((0, 3), (0, 4)) or (
            fv_tuple[:2] == (0, 5) and fv_tuple[2] < 11
        ):
            convert_plain_covers_and_docs_and_icon(data)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get_axis_sizes

get_axis_sizes(
    ns: Mapping[
        tuple[TensorId, AxisId], ParameterizedSize_N
    ],
    batch_size: int | None = None,
    *,
    max_input_shape: Mapping[tuple[TensorId, AxisId], int]
    | None = None,
) -> _AxisSizes

Determine input and output block shape for scale factors ns of parameterized input sizes.

Parameters:

Name	Type	Description	Default
ns	Mapping[tuple[TensorId, AxisId], ParameterizedSize_N]	Scale factor n for each axis (keyed by (tensor_id, axis_id)) that is parameterized as size = min + n * step.	required
batch_size	int | None	The desired size of the batch dimension. If given batch_size overwrites any batch size present in max_input_shape. Default 1.	None
max_input_shape	Mapping[tuple[TensorId, AxisId], int] | None	Limits the derived block shapes. Each axis for which the input size, parameterized by n, is larger than max_input_shape is set to the minimal value n_min for which this is still true. Use this for small input samples or large values of ns. Or simply whenever you know the full input shape.	None

Returns:

Type	Description
_AxisSizes	Resolved axis sizes for model inputs and outputs.

Source code in src/bioimageio/spec/model/v0_5.py

def get_axis_sizes(
    self,
    ns: Mapping[tuple[TensorId, AxisId], ParameterizedSize_N],
    batch_size: int | None = None,
    *,
    max_input_shape: Mapping[tuple[TensorId, AxisId], int] | None = None,
) -> _AxisSizes:
    """Determine input and output block shape for scale factors **ns**
    of parameterized input sizes.

    Args:
        ns: Scale factor `n` for each axis (keyed by (tensor_id, axis_id))
            that is parameterized as `size = min + n * step`.
        batch_size: The desired size of the batch dimension.
            If given **batch_size** overwrites any batch size present in
            **max_input_shape**. Default 1.
        max_input_shape: Limits the derived block shapes.
            Each axis for which the input size, parameterized by `n`, is larger
            than **max_input_shape** is set to the minimal value `n_min` for which
            this is still true.
            Use this for small input samples or large values of **ns**.
            Or simply whenever you know the full input shape.

    Returns:
        Resolved axis sizes for model inputs and outputs.
    """
    max_input_shape = max_input_shape or {}
    if batch_size is None:
        for (_t_id, a_id), s in max_input_shape.items():
            if a_id == BATCH_AXIS_ID:
                batch_size = s
                break
        else:
            batch_size = 1

    all_axes = {
        t.id: {a.id: a for a in t.axes} for t in chain(self.inputs, self.outputs)
    }

    inputs: dict[tuple[TensorId, AxisId], int] = {}
    outputs: dict[tuple[TensorId, AxisId], int | _DataDepSize] = {}

    def get_axis_size(a: InputAxis | OutputAxis):
        if isinstance(a, BatchAxis):
            if (t_descr.id, a.id) in ns:
                logger.warning(
                    "Ignoring unexpected size increment factor (n) for batch axis"
                    + " of tensor '{}'.",
                    t_descr.id,
                )
            return batch_size
        elif isinstance(a.size, int):
            if (t_descr.id, a.id) in ns:
                logger.warning(
                    "Ignoring unexpected size increment factor (n) for fixed size"
                    + " axis '{}' of tensor '{}'.",
                    a.id,
                    t_descr.id,
                )
            return a.size
        elif isinstance(a.size, ParameterizedSize):
            if (t_descr.id, a.id) not in ns:
                raise ValueError(
                    "Size increment factor (n) missing for parametrized axis"
                    + f" '{a.id}' of tensor '{t_descr.id}'."
                )
            n = ns[(t_descr.id, a.id)]
            s_max = max_input_shape.get((t_descr.id, a.id))
            if s_max is not None:
                n = min(n, a.size.get_n(s_max))

            return a.size.get_size(n)

        elif isinstance(a.size, SizeReference):
            if (t_descr.id, a.id) in ns:
                logger.warning(
                    "Ignoring unexpected size increment factor (n) for axis '{}'"
                    + " of tensor '{}' with size reference.",
                    a.id,
                    t_descr.id,
                )
            assert not isinstance(a, BatchAxis)
            ref_axis = all_axes[a.size.tensor_id][a.size.axis_id]
            assert not isinstance(ref_axis, BatchAxis)
            ref_key = (a.size.tensor_id, a.size.axis_id)
            ref_size = inputs.get(ref_key, outputs.get(ref_key))
            assert ref_size is not None, ref_key
            assert not isinstance(ref_size, _DataDepSize), ref_key
            return a.size.get_size(
                axis=a,
                ref_axis=ref_axis,
                ref_size=ref_size,
            )
        elif isinstance(a.size, DataDependentSize):
            if (t_descr.id, a.id) in ns:
                logger.warning(
                    "Ignoring unexpected increment factor (n) for data dependent"
                    + " size axis '{}' of tensor '{}'.",
                    a.id,
                    t_descr.id,
                )
            return _DataDepSize(a.size.min, a.size.max)
        else:
            assert_never(a.size)

    # first resolve all , but the `SizeReference` input sizes
    for t_descr in self.inputs:
        for a in t_descr.axes:
            if not isinstance(a.size, SizeReference):
                s = get_axis_size(a)
                assert not isinstance(s, _DataDepSize)
                inputs[t_descr.id, a.id] = s

    # resolve all other input axis sizes
    for t_descr in self.inputs:
        for a in t_descr.axes:
            if isinstance(a.size, SizeReference):
                s = get_axis_size(a)
                assert not isinstance(s, _DataDepSize)
                inputs[t_descr.id, a.id] = s

    # resolve all output axis sizes
    for t_descr in self.outputs:
        for a in t_descr.axes:
            assert not isinstance(a.size, ParameterizedSize)
            s = get_axis_size(a)
            outputs[t_descr.id, a.id] = s

    return _AxisSizes(inputs=inputs, outputs=outputs)

get_batch_size staticmethod

get_batch_size(
    tensor_sizes: Mapping[TensorId, Mapping[AxisId, int]],
) -> int

Source code in src/bioimageio/spec/model/v0_5.py

@staticmethod
def get_batch_size(tensor_sizes: Mapping[TensorId, Mapping[AxisId, int]]) -> int:
    batch_size = 1
    tensor_with_batchsize: TensorId | None = None
    for tid in tensor_sizes:
        for aid, s in tensor_sizes[tid].items():
            if aid != BATCH_AXIS_ID or s == 1 or s == batch_size:
                continue

            if batch_size != 1:
                assert tensor_with_batchsize is not None
                raise ValueError(
                    f"batch size mismatch for tensors '{tensor_with_batchsize}' ({batch_size}) and '{tid}' ({s})"
                )

            batch_size = s
            tensor_with_batchsize = tid

    return batch_size

get_input_test_arrays

get_input_test_arrays() -> list[NDArray[Any]]

Source code in src/bioimageio/spec/model/v0_5.py

def get_input_test_arrays(self) -> list[NDArray[Any]]:
    return self._get_test_arrays(self.inputs)

get_ns

get_ns(
    input_sizes: Mapping[TensorId, Mapping[AxisId, int]],
)

get parameter n for each parameterized axis such that the valid input size is >= the given input size

Source code in src/bioimageio/spec/model/v0_5.py

def get_ns(self, input_sizes: Mapping[TensorId, Mapping[AxisId, int]]):
    """get parameter `n` for each parameterized axis
    such that the valid input size is >= the given input size"""
    ret: dict[tuple[TensorId, AxisId], ParameterizedSize_N] = {}
    axes = {t.id: {a.id: a for a in t.axes} for t in self.inputs}
    for tid in input_sizes:
        for aid, s in input_sizes[tid].items():
            size_descr = axes[tid][aid].size
            if isinstance(size_descr, ParameterizedSize):
                ret[(tid, aid)] = size_descr.get_n(s)
            elif size_descr is None or isinstance(size_descr, (int, SizeReference)):
                pass
            else:
                assert_never(size_descr)

    return ret

get_output_tensor_sizes

get_output_tensor_sizes(
    input_sizes: Mapping[TensorId, Mapping[AxisId, int]],
) -> dict[TensorId, dict[AxisId, int | _DataDepSize]]

Returns the tensor output sizes for given input_sizes. Only if input_sizes has a valid input shape, the tensor output size is exact. Otherwise it might be larger than the actual (valid) output

Source code in src/bioimageio/spec/model/v0_5.py

def get_output_tensor_sizes(
    self, input_sizes: Mapping[TensorId, Mapping[AxisId, int]]
) -> dict[TensorId, dict[AxisId, int | _DataDepSize]]:
    """Returns the tensor output sizes for given **input_sizes**.
    Only if **input_sizes** has a valid input shape, the tensor output size is exact.
    Otherwise it might be larger than the actual (valid) output"""
    batch_size = self.get_batch_size(input_sizes)
    ns = self.get_ns(input_sizes)

    tensor_sizes = self.get_tensor_sizes(ns, batch_size=batch_size)
    return tensor_sizes.outputs

get_output_test_arrays

get_output_test_arrays() -> list[NDArray[Any]]

Source code in src/bioimageio/spec/model/v0_5.py

def get_output_test_arrays(self) -> list[NDArray[Any]]:
    return self._get_test_arrays(self.outputs)

get_package_content

get_package_content(
    local_files_only: bool = False,
) -> dict[FileName, FileDescr | BioimageioYamlContent]

Returns package content without creating the package.

Source code in src/bioimageio/spec/_internal/common_nodes.py

def get_package_content(
    self, local_files_only: bool = False
) -> dict[FileName, FileDescr | BioimageioYamlContent]:
    """Returns package content without creating the package."""
    content: dict[FileName, FileDescr] = {}
    with PackagingContext(
        bioimageio_yaml_file_name=BIOIMAGEIO_YAML,
        file_sources=content,
        local_files_only=local_files_only,
    ):
        rdf_content: BioimageioYamlContent = self.model_dump(
            mode="json", exclude_unset=True
        )

    _ = rdf_content.pop("rdf_source", None)

    return {**content, BIOIMAGEIO_YAML: rdf_content}

get_tensor_sizes

get_tensor_sizes(
    ns: Mapping[
        tuple[TensorId, AxisId], ParameterizedSize_N
    ],
    batch_size: int,
    max_input_shape: Mapping[TensorId, Mapping[AxisId, int]]
    | None = None,
) -> _TensorSizes

Source code in src/bioimageio/spec/model/v0_5.py

def get_tensor_sizes(
    self,
    ns: Mapping[tuple[TensorId, AxisId], ParameterizedSize_N],
    batch_size: int,
    max_input_shape: Mapping[TensorId, Mapping[AxisId, int]] | None = None,
) -> _TensorSizes:
    max_axis_sizes: dict[tuple[TensorId, AxisId], int] = {}
    for m, this_max_axis_sizes in (max_input_shape or {}).items():
        for a, s in this_max_axis_sizes.items():
            max_axis_sizes[(m, a)] = s

    axis_sizes = self.get_axis_sizes(
        ns, batch_size=batch_size, max_input_shape=max_axis_sizes
    )
    return _TensorSizes(
        {
            t: {
                aa: axis_sizes.inputs[(tt, aa)]
                for tt, aa in axis_sizes.inputs
                if tt == t
            }
            for t in {tt for tt, _ in axis_sizes.inputs}
        },
        {
            t: {
                aa: axis_sizes.outputs[(tt, aa)]
                for tt, aa in axis_sizes.outputs
                if tt == t
            }
            for t in {tt for tt, _ in axis_sizes.outputs}
        },
    )

load classmethod

load(
    data: IncompleteDescrView,
    context: ValidationContext | None = None,
) -> Self | InvalidDescr

factory method to create a resource description object

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def load(
    cls,
    data: IncompleteDescrView,
    context: ValidationContext | None = None,
) -> Self | InvalidDescr:
    """factory method to create a resource description object"""

    context = context or get_validation_context()
    if context.perform_io_checks:
        file_descrs = extract_file_descrs(data)
        populate_cache(file_descrs)  # TODO: add progress bar

    with context.replace(log_warnings=context.warning_level <= INFO):
        rd, errors, val_warnings = cls._load_impl(deepcopy_incomplete_descr(data))

    if context.warning_level > INFO:
        all_warnings_context = context.replace(
            warning_level=INFO, log_warnings=False, raise_errors=False
        )
        # raise all validation warnings by reloading
        with all_warnings_context:
            _, _, val_warnings = cls._load_impl(deepcopy_incomplete_descr(data))

    format_status = "failed" if errors else "passed"
    rd.validation_summary.add_detail(
        ValidationDetail(
            errors=errors,
            name=(
                "bioimageio.spec format validation"
                f" {rd.type} {cls.implemented_format_version}"
            ),
            status=format_status,
            warnings=val_warnings,
        ),
        update_status=False,  # this special validation detail needs manual format updating below
    )
    assert format_status != "failed" or isinstance(rd, InvalidDescr)

    return rd

load_from_kwargs classmethod

load_from_kwargs(
    context: ValidationContext | None = None,
    *args: P.args,
    **kwargs: P.kwargs,
) -> T | InvalidDescr

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def load_from_kwargs(
    cls: Callable[P, T],
    context: ValidationContext | None = None,
    *args: P.args,
    **kwargs: P.kwargs,
) -> T | InvalidDescr:
    sig = signature(cls)
    bound = sig.bind_partial(*args, **kwargs)
    return cls.load(dict(bound.arguments), context=context)  # pyright: ignore[reportFunctionMemberAccess]

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

package

package(
    dest: ZipFile | IO[bytes] | Path | str | None = None,
    /,
    local_files_only: bool = False,
) -> ZipFile

package the described resource as a zip archive

Parameters:

Name	Type	Description	Default
dest	ZipFile | IO[bytes] | Path | str | None	(path/bytes stream of) destination zipfile	None

Source code in src/bioimageio/spec/_internal/common_nodes.py

def package(
    self,
    dest: ZipFile | IO[bytes] | Path | str | None = None,
    /,
    local_files_only: bool = False,
) -> ZipFile:
    """package the described resource as a zip archive

    Args:
        dest: (path/bytes stream of) destination zipfile
    """
    if dest is None:
        dest = BytesIO()

    if isinstance(dest, ZipFile):
        zip = dest
        if "r" in zip.mode:
            raise ValueError(
                f"zip file {dest} opened in '{zip.mode}' mode,"
                + " but write access is needed for packaging."
            )
    else:
        zip = ZipFile(dest, mode="w")

    if zip.filename is None:
        zip.filename = (
            str(getattr(self, "id", getattr(self, "name", "bioimageio"))) + ".zip"
        )

    content = self.get_package_content(local_files_only=local_files_only)
    write_content_to_zip(content, zip)
    return zip

validate_input_tensors

validate_input_tensors(
    sources: Sequence[NDArray[Any]]
    | Mapping[TensorId, NDArray[Any] | None],
    *,
    pad_inputs: bool | Literal["allow"] = True,
    crop_outputs: bool | Literal["allow"] = True,
) -> Mapping[TensorId, NDArray[Any] | None]

Check if the given input tensors match the model's input tensor descriptions. This includes checks of tensor shapes and dtypes, but not of the actual values.

Source code in src/bioimageio/spec/model/v0_5.py

def validate_input_tensors(
    self,
    sources: Sequence[NDArray[Any]] | Mapping[TensorId, NDArray[Any] | None],
    *,
    pad_inputs: bool | Literal["allow"] = True,
    crop_outputs: bool | Literal["allow"] = True,
) -> Mapping[TensorId, NDArray[Any] | None]:
    """Check if the given input tensors match the model's input tensor descriptions.
    This includes checks of tensor shapes and dtypes, but not of the actual values.
    """
    if not isinstance(sources, collections.abc.Mapping):
        sources = {descr.id: tensor for descr, tensor in zip(self.inputs, sources)}

    tensors = {
        **{descr.id: (descr, sources.get(descr.id)) for descr in self.inputs},
        **{  # outputs are required for halo
            descr.id: (descr, None) for descr in self.outputs
        },
    }
    validate_tensors(tensors, pad_inputs=pad_inputs, crop_outputs=crop_outputs)

    return sources

warn_about_tag_categories classmethod

warn_about_tag_categories(
    value: list[str], info: ValidationInfo
) -> list[str]

Source code in src/bioimageio/spec/generic/v0_3.py

@as_warning
@field_validator("tags")
@classmethod
def warn_about_tag_categories(
    cls, value: list[str], info: ValidationInfo
) -> list[str]:
    categories = TAG_CATEGORIES.get(info.data["type"], {})
    missing_categories: list[dict[str, Sequence[str]]] = []
    for cat, entries in categories.items():
        if not any(e in value for e in entries):
            missing_categories.append({cat: entries})

    if missing_categories:
        raise ValueError(
            f"Missing tags from bioimage.io categories: {missing_categories}"
        )

    return value

ModelId

Bases: ResourceId

              flowchart TD
              bioimageio.spec.model.v0_5.ModelId[ModelId]
              bioimageio.spec.generic.v0_3.ResourceId[ResourceId]
              bioimageio.spec._internal.validated_string.ValidatedString[ValidatedString]

                              bioimageio.spec.generic.v0_3.ResourceId --> bioimageio.spec.model.v0_5.ModelId
                                bioimageio.spec._internal.validated_string.ValidatedString --> bioimageio.spec.generic.v0_3.ResourceId
                



              click bioimageio.spec.model.v0_5.ModelId href "" "bioimageio.spec.model.v0_5.ModelId"
              click bioimageio.spec.generic.v0_3.ResourceId href "" "bioimageio.spec.generic.v0_3.ResourceId"
              click bioimageio.spec._internal.validated_string.ValidatedString href "" "bioimageio.spec._internal.validated_string.ValidatedString"
            

Methods:

Name	Description
__get_pydantic_core_schema__
__get_pydantic_json_schema__
__new__

Attributes:

Name	Type	Description
root_model	type[RootModel[Any]]	the pydantic root model to validate the string

root_model class-attribute

root_model: type[RootModel[Any]] = RootModel[
    Annotated[
        NotEmpty[str],
        RestrictCharacters(
            string.ascii_lowercase + string.digits + "_-/."
        ),
        annotated_types.Predicate(
            lambda s: (
                not (
                    s.startswith("/") or s.endswith("/")
                )
            )
        ),
    ]
]

the pydantic root model to validate the string

__get_pydantic_core_schema__ classmethod

__get_pydantic_core_schema__(
    source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_core_schema__(
    cls, source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema:
    return no_info_after_validator_function(cls, handler(str))

__get_pydantic_json_schema__ classmethod

__get_pydantic_json_schema__(
    core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_json_schema__(
    cls, core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue:
    json_schema = cls.root_model.model_json_schema(mode=handler.mode)
    json_schema["title"] = cls.__name__.strip("_")
    if cls.__doc__:
        json_schema["description"] = cls.__doc__

    return json_schema

__new__

__new__(object: object)

Source code in src/bioimageio/spec/_internal/validated_string.py

def __new__(cls, object: object):
    _validated = cls.root_model.model_validate(object).root
    self = super().__new__(cls, _validated)
    self._validated = _validated
    return self._after_validator()

NominalOrOrdinalDataDescr pydantic-model

Bases: Node

Show JSON schema:

{
  "additionalProperties": false,
  "properties": {
    "values": {
      "anyOf": [
        {
          "items": {
            "type": "integer"
          },
          "minItems": 1,
          "type": "array"
        },
        {
          "items": {
            "type": "number"
          },
          "minItems": 1,
          "type": "array"
        },
        {
          "items": {
            "type": "boolean"
          },
          "minItems": 1,
          "type": "array"
        },
        {
          "items": {
            "type": "string"
          },
          "minItems": 1,
          "type": "array"
        }
      ],
      "description": "A fixed set of nominal or an ascending sequence of ordinal values.\nIn this case `data.type` is required to be an unsigend integer type, e.g. 'uint8'.\nString `values` are interpreted as labels for tensor values 0, ..., N.\nNote: as YAML 1.2 does not natively support a \"set\" datatype,\nnominal values should be given as a sequence (aka list/array) as well.",
      "title": "Values"
    },
    "type": {
      "default": "uint8",
      "enum": [
        "float32",
        "float64",
        "uint8",
        "int8",
        "uint16",
        "int16",
        "uint32",
        "int32",
        "uint64",
        "int64",
        "bool"
      ],
      "examples": [
        "float32",
        "uint8",
        "uint16",
        "int64",
        "bool"
      ],
      "title": "Type",
      "type": "string"
    },
    "unit": {
      "anyOf": [
        {
          "const": "arbitrary unit",
          "type": "string"
        },
        {
          "description": "An SI unit",
          "minLength": 1,
          "pattern": "^(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^[+-]?[1-9]\\d*)?((\u00b7(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^[+-]?[1-9]\\d*)?)|(/(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^+?[1-9]\\d*)?))*$",
          "title": "SiUnit",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Unit"
    }
  },
  "required": [
    "values"
  ],
  "title": "model.v0_5.NominalOrOrdinalDataDescr",
  "type": "object"
}

Fields:

• values (TVs)
• type (Annotated[NominalOrOrdinalDType, Field(examples=['float32', 'uint8', 'uint16', 'int64', 'bool'])])
• unit (Literal['arbitrary unit'] | SiUnit | None)

Validators:

• _validate_values_match_type

range property

range

type pydantic-field

type: Annotated[
    NominalOrOrdinalDType,
    Field(
        examples=[
            "float32",
            "uint8",
            "uint16",
            "int64",
            "bool",
        ]
    ),
] = "uint8"

unit pydantic-field

unit: Literal['arbitrary unit'] | SiUnit | None = None

values pydantic-field

values: TVs

A fixed set of nominal or an ascending sequence of ordinal values. In this case data.type is required to be an unsigend integer type, e.g. 'uint8'. String values are interpreted as labels for tensor values 0, ..., N. Note: as YAML 1.2 does not natively support a "set" datatype, nominal values should be given as a sequence (aka list/array) as well.

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

OnnxWeightsDescr pydantic-model

Bases: WeightsEntryDescrBase

Show JSON schema:

{
  "$defs": {
    "Author": {
      "additionalProperties": false,
      "properties": {
        "affiliation": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Affiliation",
          "title": "Affiliation"
        },
        "email": {
          "anyOf": [
            {
              "format": "email",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Email",
          "title": "Email"
        },
        "orcid": {
          "anyOf": [
            {
              "description": "An ORCID identifier, see https://orcid.org/",
              "title": "OrcidId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "An [ORCID iD](https://support.orcid.org/hc/en-us/sections/360001495313-What-is-ORCID\n) in hyphenated groups of 4 digits, (and [valid](\nhttps://support.orcid.org/hc/en-us/articles/360006897674-Structure-of-the-ORCID-Identifier\n) as per ISO 7064 11,2.)",
          "examples": [
            "0000-0001-2345-6789"
          ],
          "title": "Orcid"
        },
        "name": {
          "title": "Name",
          "type": "string"
        },
        "github_user": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Github User"
        }
      },
      "required": [
        "name"
      ],
      "title": "generic.v0_3.Author",
      "type": "object"
    },
    "FileDescr": {
      "additionalProperties": false,
      "description": "A file description",
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "File source",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        }
      },
      "required": [
        "source"
      ],
      "title": "_internal.io.FileDescr",
      "type": "object"
    },
    "RelativeFilePath": {
      "description": "A path relative to the `rdf.yaml` file (also if the RDF source is a URL).",
      "format": "path",
      "title": "RelativeFilePath",
      "type": "string"
    }
  },
  "additionalProperties": false,
  "properties": {
    "source": {
      "anyOf": [
        {
          "description": "A URL with the HTTP or HTTPS scheme.",
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "title": "HttpUrl",
          "type": "string"
        },
        {
          "$ref": "#/$defs/RelativeFilePath"
        },
        {
          "format": "file-path",
          "title": "FilePath",
          "type": "string"
        }
      ],
      "description": "Source of the weights file.",
      "title": "FileSource"
    },
    "sha256": {
      "anyOf": [
        {
          "description": "A SHA-256 hash value",
          "maxLength": 64,
          "minLength": 64,
          "title": "Sha256",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "SHA256 hash value of the **source** file.",
      "title": "Sha256"
    },
    "authors": {
      "anyOf": [
        {
          "items": {
            "$ref": "#/$defs/Author"
          },
          "type": "array"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
      "title": "Authors"
    },
    "parent": {
      "anyOf": [
        {
          "enum": [
            "keras_hdf5",
            "keras_v3",
            "onnx",
            "pytorch_state_dict",
            "tensorflow_js",
            "tensorflow_saved_model_bundle",
            "torchscript"
          ],
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
      "examples": [
        "pytorch_state_dict"
      ],
      "title": "Parent"
    },
    "comment": {
      "default": "",
      "description": "A comment about this weights entry, for example how these weights were created.",
      "title": "Comment",
      "type": "string"
    },
    "opset_version": {
      "description": "ONNX opset version",
      "minimum": 7,
      "title": "Opset Version",
      "type": "integer"
    },
    "external_data": {
      "anyOf": [
        {
          "$ref": "#/$defs/FileDescr",
          "examples": [
            {
              "source": "weights.onnx.data"
            }
          ]
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Source of the external ONNX data file holding the weights.\n(If present **source** holds the ONNX architecture without weights)."
    }
  },
  "required": [
    "source",
    "opset_version"
  ],
  "title": "model.v0_5.OnnxWeightsDescr",
  "type": "object"
}

Fields:

• source (Annotated[FileSource, AfterValidator(wo_special_file_name)])
• sha256 (Sha256 | None)
• authors (list[Author] | None)
• parent (Annotated[WeightsFormat | None, Field(examples=['pytorch_state_dict'])])
• comment (str)
• opset_version (Annotated[int, Ge(7)])
• external_data (FileDescr_external_data | None)

Validators:

• _validate
• _validate_external_data_unique_file_name

authors pydantic-field

authors: list[Author] | None = None

Authors Either the person(s) that have trained this model resulting in the original weights file. (If this is the initial weights entry, i.e. it does not have a parent) Or the person(s) who have converted the weights to this weights format. (If this is a child weight, i.e. it has a parent field)

comment pydantic-field

comment: str = ''

A comment about this weights entry, for example how these weights were created.

external_data pydantic-field

external_data: FileDescr_external_data | None = None

Source of the external ONNX data file holding the weights. (If present source holds the ONNX architecture without weights).

opset_version pydantic-field

opset_version: Annotated[int, Ge(7)]

ONNX opset version

parent pydantic-field

parent: Annotated[
    WeightsFormat | None,
    Field(examples=["pytorch_state_dict"]),
] = None

The source weights these weights were converted from. For example, if a model's weights were converted from the pytorch_state_dict format to torchscript, The pytorch_state_dict weights entry has no parent and is the parent of the torchscript weights. All weight entries except one (the initial set of weights resulting from training the model), need to have this field.

sha256 pydantic-field

sha256: Sha256 | None = None

SHA256 hash value of the source file.

source pydantic-field

source: Annotated[
    FileSource, AfterValidator(wo_special_file_name)
]

Source of the weights file.

suffix property

suffix: str

type class-attribute

type: WeightsFormat = 'onnx'

weights_format_name class-attribute

weights_format_name: str = 'ONNX'

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

download

download(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

alias for .get_reader

Source code in src/bioimageio/spec/_internal/io.py

def download(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """alias for `.get_reader`"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

get_reader

get_reader(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

open the file source (download if needed)

Source code in src/bioimageio/spec/_internal/io.py

def get_reader(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """open the file source (download if needed)"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

validate_sha256

validate_sha256(force_recompute: bool = False) -> None

validate the sha256 hash value of the source file

Source code in src/bioimageio/spec/_internal/io.py

def validate_sha256(self, force_recompute: bool = False) -> None:
    """validate the sha256 hash value of the **source** file"""
    context = get_validation_context()
    src_str = str(self.source)
    if force_recompute:
        actual_sha = None
    else:
        actual_sha = context.known_files.get(src_str)

    if actual_sha is None:
        if context.perform_io_checks or force_recompute:
            reader = get_reader(self.source, sha256=self.sha256)
            if force_recompute:
                actual_sha = get_sha256(reader)
            else:
                actual_sha = reader.sha256

            context.known_files[src_str] = actual_sha
        elif context.known_files and src_str not in context.known_files:
            # perform_io_checks is False, but known files were given,
            # so we expect all file references to be in there
            raise ValueError(f"File {src_str} not found in `known_files`.")

    if actual_sha is None or self.sha256 == actual_sha:
        return
    elif self.sha256 is None or context.update_hashes:
        self.sha256 = actual_sha
    elif self.sha256 != actual_sha:
        raise ValueError(
            f"Sha256 mismatch for {self.source}. Expected {self.sha256}, got "
            + f"{actual_sha}. Update expected `sha256` or point to the matching "
            + "file."
        )

OrcidId

Bases: ValidatedString

              flowchart TD
              bioimageio.spec.model.v0_5.OrcidId[OrcidId]
              bioimageio.spec._internal.validated_string.ValidatedString[ValidatedString]

                              bioimageio.spec._internal.validated_string.ValidatedString --> bioimageio.spec.model.v0_5.OrcidId
                


              click bioimageio.spec.model.v0_5.OrcidId href "" "bioimageio.spec.model.v0_5.OrcidId"
              click bioimageio.spec._internal.validated_string.ValidatedString href "" "bioimageio.spec._internal.validated_string.ValidatedString"
            

An ORCID identifier, see https://orcid.org/

Methods:

Name	Description
__get_pydantic_core_schema__
__get_pydantic_json_schema__
__new__

Attributes:

Name	Type	Description
root_model	type[RootModel[Any]]	the pydantic root model to validate the string

root_model class-attribute

root_model: type[RootModel[Any]] = RootModel[
    Annotated[str, AfterValidator(_validate_orcid_id)]
]

the pydantic root model to validate the string

__get_pydantic_core_schema__ classmethod

__get_pydantic_core_schema__(
    source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_core_schema__(
    cls, source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema:
    return no_info_after_validator_function(cls, handler(str))

__get_pydantic_json_schema__ classmethod

__get_pydantic_json_schema__(
    core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_json_schema__(
    cls, core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue:
    json_schema = cls.root_model.model_json_schema(mode=handler.mode)
    json_schema["title"] = cls.__name__.strip("_")
    if cls.__doc__:
        json_schema["description"] = cls.__doc__

    return json_schema

__new__

__new__(object: object)

Source code in src/bioimageio/spec/_internal/validated_string.py

def __new__(cls, object: object):
    _validated = cls.root_model.model_validate(object).root
    self = super().__new__(cls, _validated)
    self._validated = _validated
    return self._after_validator()

OutputTensorDescr pydantic-model

Bases: TensorDescrBase[OutputAxis]

Show JSON schema:

{
  "$defs": {
    "BatchAxis": {
      "additionalProperties": false,
      "properties": {
        "id": {
          "default": "batch",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "batch",
          "title": "Type",
          "type": "string"
        },
        "size": {
          "anyOf": [
            {
              "const": 1,
              "type": "integer"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The batch size may be fixed to 1,\notherwise (the default) it may be chosen arbitrarily depending on available memory",
          "title": "Size"
        }
      },
      "required": [
        "type"
      ],
      "title": "model.v0_5.BatchAxis",
      "type": "object"
    },
    "BinarizeAlongAxisKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [BinarizeDescr][]",
      "properties": {
        "threshold": {
          "description": "The fixed threshold values along `axis`",
          "items": {
            "type": "number"
          },
          "minItems": 1,
          "title": "Threshold",
          "type": "array"
        },
        "axis": {
          "description": "The `threshold` axis",
          "examples": [
            "channel"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        }
      },
      "required": [
        "threshold",
        "axis"
      ],
      "title": "model.v0_5.BinarizeAlongAxisKwargs",
      "type": "object"
    },
    "BinarizeDescr": {
      "additionalProperties": false,
      "description": "Binarize the tensor with a fixed threshold.\n\nValues above [BinarizeKwargs.threshold][]/[BinarizeAlongAxisKwargs.threshold][]\nwill be set to one, values below the threshold to zero.\n\nExamples:\n- in YAML\n    ```yaml\n    postprocessing:\n      - id: binarize\n        kwargs:\n          axis: 'channel'\n          threshold: [0.25, 0.5, 0.75]\n    ```\n- in Python:\n\n    >>> postprocessing = [BinarizeDescr(\n    ...   kwargs=BinarizeAlongAxisKwargs(\n    ...       axis=AxisId('channel'),\n    ...       threshold=[0.25, 0.5, 0.75],\n    ...   )\n    ... )]",
      "properties": {
        "id": {
          "const": "binarize",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "anyOf": [
            {
              "$ref": "#/$defs/BinarizeKwargs"
            },
            {
              "$ref": "#/$defs/BinarizeAlongAxisKwargs"
            }
          ],
          "title": "Kwargs"
        }
      },
      "required": [
        "id",
        "kwargs"
      ],
      "title": "model.v0_5.BinarizeDescr",
      "type": "object"
    },
    "BinarizeKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [BinarizeDescr][]",
      "properties": {
        "threshold": {
          "description": "The fixed threshold",
          "title": "Threshold",
          "type": "number"
        }
      },
      "required": [
        "threshold"
      ],
      "title": "model.v0_5.BinarizeKwargs",
      "type": "object"
    },
    "CellposeFlowDynamicsDescr": {
      "additionalProperties": false,
      "description": "Cellpose flow dynamics postprocessing as described in:\n- Carsen Stringer and Marius Pachitariu. [*Cellpose: a generalist algorithm for cellular segmentation*](https://www.nature.com/articles/s41592-020-01018-x). Nature Methods, 2021.\n\nNote: Only available if the `cellpose` package is installed.",
      "properties": {
        "id": {
          "const": "cellpose_flow_dynamics",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "$ref": "#/$defs/CellposeFlowDynamicsKwargs"
        }
      },
      "required": [
        "id",
        "kwargs"
      ],
      "title": "model.v0_5.CellposeFlowDynamicsDescr",
      "type": "object"
    },
    "CellposeFlowDynamicsKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [CellposeFlowDynamicsDescr][]",
      "properties": {
        "cellprob_threshold": {
          "title": "Cellprob Threshold",
          "type": "number"
        },
        "flow_threshold": {
          "title": "Flow Threshold",
          "type": "number"
        },
        "do_3D": {
          "title": "Do 3D",
          "type": "boolean"
        },
        "min_size": {
          "default": 15,
          "description": "Minimum size of objects to keep, in pixels. Default is 15, which is the default in Cellpose. Set to 0 to disable filtering by size.",
          "title": "Min Size",
          "type": "integer"
        },
        "output_dtype": {
          "default": "uint16",
          "enum": [
            "uint16",
            "uint32"
          ],
          "title": "Output Dtype",
          "type": "string"
        }
      },
      "required": [
        "cellprob_threshold",
        "flow_threshold",
        "do_3D"
      ],
      "title": "model.v0_5.CellposeFlowDynamicsKwargs",
      "type": "object"
    },
    "ChannelAxis": {
      "additionalProperties": false,
      "properties": {
        "id": {
          "default": "channel",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "channel",
          "title": "Type",
          "type": "string"
        },
        "channel_names": {
          "description": "Name/label for each channel. The number of channels is given by `len(channel_names)`.",
          "items": {
            "type": "string"
          },
          "minItems": 1,
          "title": "Channel Names",
          "type": "array"
        }
      },
      "required": [
        "type",
        "channel_names"
      ],
      "title": "model.v0_5.ChannelAxis",
      "type": "object"
    },
    "ClipDescr": {
      "additionalProperties": false,
      "description": "Set tensor values below min to min and above max to max.\n\nSee `ScaleRangeDescr` for examples.",
      "properties": {
        "id": {
          "const": "clip",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "$ref": "#/$defs/ClipKwargs"
        }
      },
      "required": [
        "id",
        "kwargs"
      ],
      "title": "model.v0_5.ClipDescr",
      "type": "object"
    },
    "ClipKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [ClipDescr][]",
      "properties": {
        "min": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Minimum value for clipping.\n\nExclusive with [min_percentile][]",
          "title": "Min"
        },
        "min_percentile": {
          "anyOf": [
            {
              "exclusiveMaximum": 100,
              "minimum": 0,
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Minimum percentile for clipping.\n\nExclusive with [min][].\n\nIn range [0, 100).",
          "title": "Min Percentile"
        },
        "max": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Maximum value for clipping.\n\nExclusive with `max_percentile`.",
          "title": "Max"
        },
        "max_percentile": {
          "anyOf": [
            {
              "exclusiveMinimum": 1,
              "maximum": 100,
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Maximum percentile for clipping.\n\nExclusive with `max`.\n\nIn range (1, 100].",
          "title": "Max Percentile"
        },
        "axes": {
          "anyOf": [
            {
              "items": {
                "maxLength": 16,
                "minLength": 1,
                "title": "AxisId",
                "type": "string"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The subset of axes to determine percentiles jointly,\n\ni.e. axes to reduce to compute min/max from `min_percentile`/`max_percentile`.\nFor example to clip 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x')\nresulting in a tensor of equal shape with clipped values per channel, specify `axes=('batch', 'x', 'y')`.\nTo clip samples independently, leave out the 'batch' axis.\n\nOnly valid if `min_percentile` and/or `max_percentile` are set.\n\nDefault: Compute percentiles over all axes jointly.",
          "examples": [
            [
              "batch",
              "x",
              "y"
            ]
          ],
          "title": "Axes"
        }
      },
      "title": "model.v0_5.ClipKwargs",
      "type": "object"
    },
    "CustomProcessingDescr": {
      "additionalProperties": false,
      "description": "Custom (post)processing op \u2014 source file shipped inline with the model.\n\nSupports (post)processing that cannot be expressed by the built-in named\noperations (watershed, connected components, etc.)\nusing a simple Python callable interface.\n\nThe op is implemented in a ``.py`` file packaged alongside the model weights.\nTwo styles are supported:\n\n*Callable class* \u2014 kwargs go to ``__init__``, tensors arrive in ``__call__``:\n\n.. code-block:: python\n\n    # my_postprocess.py\n    import numpy as np\n\n    class my_postprocess:\n        def __init__(self, threshold: float = 0.5) -> None:\n            self.threshold = threshold\n        def __call__(self, *arrays: np.ndarray) -> np.ndarray:\n            # arrays = model output tensors in rdf.yaml declaration order\n            return (arrays[0] > self.threshold).astype(np.uint8)\n\n*Factory function* \u2014 alternative closure style, identical runtime behaviour:\n\n.. code-block:: python\n\n    # my_postprocess.py\n    import numpy as np\n\n    def my_postprocess(threshold: float = 0.5):\n        def run(*arrays: np.ndarray) -> np.ndarray:\n            return (arrays[0] > threshold).astype(np.uint8)\n        return run\n\nReference it in ``rdf.yaml`` with the source file included in the package:\n\n.. code-block:: yaml\n\n    postprocessing:\n      - id: custom\n        callable: my_postprocess   # class or function name in source\n        source: my_postprocess.py  # packaged alongside weights\n        sha256: <hash>             # sha256 of the source file\n        kwargs:                    # forwarded to __init__ / factory\n          threshold: 0.5\n\n**Security:** source files are SHA-256 verified before execution.\nExecution requires explicit opt-in in bioimageio.core and curator\nreview before Zoo publication.",
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "Python source file (included when packaging the model).",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        },
        "id": {
          "const": "custom",
          "title": "Id",
          "type": "string"
        },
        "callable": {
          "description": "Name of the callable class or factory function defined in ``source``.\n\nAt runtime: ``op = callable(**kwargs)``, then ``result = op(*output_tensors)``\nper image.  Both a class with ``__call__`` and a factory function returning\na callable satisfy this protocol.",
          "examples": [
            "my_postprocess_factory",
            "MyPostprocessClass"
          ],
          "title": "Callable",
          "type": "string"
        },
        "kwargs": {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "description": "Keyword arguments forwarded to the callable (``__init__`` or factory).",
          "title": "Kwargs",
          "type": "object"
        }
      },
      "required": [
        "source",
        "id",
        "callable"
      ],
      "title": "model.v0_5.CustomProcessingDescr",
      "type": "object"
    },
    "DataDependentSize": {
      "additionalProperties": false,
      "properties": {
        "min": {
          "default": 1,
          "exclusiveMinimum": 0,
          "title": "Min",
          "type": "integer"
        },
        "max": {
          "anyOf": [
            {
              "exclusiveMinimum": 1,
              "type": "integer"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Max"
        }
      },
      "title": "model.v0_5.DataDependentSize",
      "type": "object"
    },
    "EnsureDtypeDescr": {
      "additionalProperties": false,
      "description": "Cast the tensor data type to `EnsureDtypeKwargs.dtype` (if not matching).\n\nThis can for example be used to ensure the inner neural network model gets a\ndifferent input tensor data type than the fully described bioimage.io model does.\n\nExamples:\n    The described bioimage.io model (incl. preprocessing) accepts any\n    float32-compatible tensor, normalizes it with percentiles and clipping and then\n    casts it to uint8, which is what the neural network in this example expects.\n    - in YAML\n        ```yaml\n        inputs:\n        - data:\n            type: float32  # described bioimage.io model is compatible with any float32 input tensor\n          preprocessing:\n          - id: scale_range\n              kwargs:\n              axes: ['y', 'x']\n              max_percentile: 99.8\n              min_percentile: 5.0\n          - id: clip\n              kwargs:\n              min: 0.0\n              max: 1.0\n          - id: ensure_dtype  # the neural network of the model requires uint8\n              kwargs:\n              dtype: uint8\n        ```\n    - in Python:\n        >>> preprocessing = [\n        ...     ScaleRangeDescr(\n        ...         kwargs=ScaleRangeKwargs(\n        ...           axes= (AxisId('y'), AxisId('x')),\n        ...           max_percentile= 99.8,\n        ...           min_percentile= 5.0,\n        ...         )\n        ...     ),\n        ...     ClipDescr(kwargs=ClipKwargs(min=0.0, max=1.0)),\n        ...     EnsureDtypeDescr(kwargs=EnsureDtypeKwargs(dtype=\"uint8\")),\n        ... ]",
      "properties": {
        "id": {
          "const": "ensure_dtype",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "$ref": "#/$defs/EnsureDtypeKwargs"
        }
      },
      "required": [
        "id",
        "kwargs"
      ],
      "title": "model.v0_5.EnsureDtypeDescr",
      "type": "object"
    },
    "EnsureDtypeKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [EnsureDtypeDescr][]",
      "properties": {
        "dtype": {
          "enum": [
            "float32",
            "float64",
            "uint8",
            "int8",
            "uint16",
            "int16",
            "uint32",
            "int32",
            "uint64",
            "int64",
            "bool"
          ],
          "title": "Dtype",
          "type": "string"
        }
      },
      "required": [
        "dtype"
      ],
      "title": "model.v0_5.EnsureDtypeKwargs",
      "type": "object"
    },
    "FileDescr": {
      "additionalProperties": false,
      "description": "A file description",
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "File source",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        }
      },
      "required": [
        "source"
      ],
      "title": "_internal.io.FileDescr",
      "type": "object"
    },
    "FixedZeroMeanUnitVarianceAlongAxisKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [FixedZeroMeanUnitVarianceDescr][]",
      "properties": {
        "mean": {
          "description": "The mean value(s) to normalize with.",
          "items": {
            "type": "number"
          },
          "minItems": 1,
          "title": "Mean",
          "type": "array"
        },
        "std": {
          "description": "The standard deviation value(s) to normalize with.\nSize must match `mean` values.",
          "items": {
            "minimum": 1e-06,
            "type": "number"
          },
          "minItems": 1,
          "title": "Std",
          "type": "array"
        },
        "axis": {
          "description": "The axis of the mean/std values to normalize each entry along that dimension\nseparately.",
          "examples": [
            "channel",
            "index"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        }
      },
      "required": [
        "mean",
        "std",
        "axis"
      ],
      "title": "model.v0_5.FixedZeroMeanUnitVarianceAlongAxisKwargs",
      "type": "object"
    },
    "FixedZeroMeanUnitVarianceDescr": {
      "additionalProperties": false,
      "description": "Subtract a given mean and divide by the standard deviation.\n\nNormalize with fixed, precomputed values for\n`FixedZeroMeanUnitVarianceKwargs.mean` and `FixedZeroMeanUnitVarianceKwargs.std`\nUse `FixedZeroMeanUnitVarianceAlongAxisKwargs` for independent scaling along given\naxes.\n\nExamples:\n1. scalar value for whole tensor\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: fixed_zero_mean_unit_variance\n        kwargs:\n          mean: 103.5\n          std: 13.7\n    ```\n    - in Python\n    >>> preprocessing = [FixedZeroMeanUnitVarianceDescr(\n    ...   kwargs=FixedZeroMeanUnitVarianceKwargs(mean=103.5, std=13.7)\n    ... )]\n\n2. independently along an axis\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: fixed_zero_mean_unit_variance\n        kwargs:\n          axis: channel\n          mean: [101.5, 102.5, 103.5]\n          std: [11.7, 12.7, 13.7]\n    ```\n    - in Python\n    >>> preprocessing = [FixedZeroMeanUnitVarianceDescr(\n    ...   kwargs=FixedZeroMeanUnitVarianceAlongAxisKwargs(\n    ...     axis=AxisId(\"channel\"),\n    ...     mean=[101.5, 102.5, 103.5],\n    ...     std=[11.7, 12.7, 13.7],\n    ...   )\n    ... )]",
      "properties": {
        "id": {
          "const": "fixed_zero_mean_unit_variance",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "anyOf": [
            {
              "$ref": "#/$defs/FixedZeroMeanUnitVarianceKwargs"
            },
            {
              "$ref": "#/$defs/FixedZeroMeanUnitVarianceAlongAxisKwargs"
            }
          ],
          "title": "Kwargs"
        }
      },
      "required": [
        "id",
        "kwargs"
      ],
      "title": "model.v0_5.FixedZeroMeanUnitVarianceDescr",
      "type": "object"
    },
    "FixedZeroMeanUnitVarianceKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [FixedZeroMeanUnitVarianceDescr][]",
      "properties": {
        "mean": {
          "description": "The mean value to normalize with.",
          "title": "Mean",
          "type": "number"
        },
        "std": {
          "description": "The standard deviation value to normalize with.",
          "minimum": 1e-06,
          "title": "Std",
          "type": "number"
        }
      },
      "required": [
        "mean",
        "std"
      ],
      "title": "model.v0_5.FixedZeroMeanUnitVarianceKwargs",
      "type": "object"
    },
    "IndexOutputAxis": {
      "additionalProperties": false,
      "properties": {
        "id": {
          "default": "index",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "index",
          "title": "Type",
          "type": "string"
        },
        "size": {
          "anyOf": [
            {
              "exclusiveMinimum": 0,
              "type": "integer"
            },
            {
              "$ref": "#/$defs/SizeReference"
            },
            {
              "$ref": "#/$defs/DataDependentSize"
            }
          ],
          "description": "The size/length of this axis can be specified as\n- fixed integer\n- reference to another axis with an optional offset ([SizeReference][])\n- data dependent size using [DataDependentSize][] (size is only known after model inference)",
          "examples": [
            10,
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ],
          "title": "Size"
        }
      },
      "required": [
        "type",
        "size"
      ],
      "title": "model.v0_5.IndexOutputAxis",
      "type": "object"
    },
    "IntervalOrRatioDataDescr": {
      "additionalProperties": false,
      "properties": {
        "type": {
          "default": "float32",
          "enum": [
            "float32",
            "float64",
            "uint8",
            "int8",
            "uint16",
            "int16",
            "uint32",
            "int32",
            "uint64",
            "int64"
          ],
          "examples": [
            "float32",
            "float64",
            "uint8",
            "uint16"
          ],
          "title": "Type",
          "type": "string"
        },
        "range": {
          "default": [
            null,
            null
          ],
          "description": "Tuple `(minimum, maximum)` specifying the allowed range of the data in this tensor.\n`None` corresponds to min/max of what can be expressed by **type**.",
          "maxItems": 2,
          "minItems": 2,
          "prefixItems": [
            {
              "anyOf": [
                {
                  "type": "number"
                },
                {
                  "type": "null"
                }
              ]
            },
            {
              "anyOf": [
                {
                  "type": "number"
                },
                {
                  "type": "null"
                }
              ]
            }
          ],
          "title": "Range",
          "type": "array"
        },
        "unit": {
          "anyOf": [
            {
              "const": "arbitrary unit",
              "type": "string"
            },
            {
              "description": "An SI unit",
              "minLength": 1,
              "pattern": "^(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^[+-]?[1-9]\\d*)?((\u00b7(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^[+-]?[1-9]\\d*)?)|(/(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^+?[1-9]\\d*)?))*$",
              "title": "SiUnit",
              "type": "string"
            }
          ],
          "default": "arbitrary unit",
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "description": "Scale for data on an interval (or ratio) scale.",
          "title": "Scale",
          "type": "number"
        },
        "offset": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Offset for data on a ratio scale.",
          "title": "Offset"
        }
      },
      "title": "model.v0_5.IntervalOrRatioDataDescr",
      "type": "object"
    },
    "NominalOrOrdinalDataDescr": {
      "additionalProperties": false,
      "properties": {
        "values": {
          "anyOf": [
            {
              "items": {
                "type": "integer"
              },
              "minItems": 1,
              "type": "array"
            },
            {
              "items": {
                "type": "number"
              },
              "minItems": 1,
              "type": "array"
            },
            {
              "items": {
                "type": "boolean"
              },
              "minItems": 1,
              "type": "array"
            },
            {
              "items": {
                "type": "string"
              },
              "minItems": 1,
              "type": "array"
            }
          ],
          "description": "A fixed set of nominal or an ascending sequence of ordinal values.\nIn this case `data.type` is required to be an unsigend integer type, e.g. 'uint8'.\nString `values` are interpreted as labels for tensor values 0, ..., N.\nNote: as YAML 1.2 does not natively support a \"set\" datatype,\nnominal values should be given as a sequence (aka list/array) as well.",
          "title": "Values"
        },
        "type": {
          "default": "uint8",
          "enum": [
            "float32",
            "float64",
            "uint8",
            "int8",
            "uint16",
            "int16",
            "uint32",
            "int32",
            "uint64",
            "int64",
            "bool"
          ],
          "examples": [
            "float32",
            "uint8",
            "uint16",
            "int64",
            "bool"
          ],
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "const": "arbitrary unit",
              "type": "string"
            },
            {
              "description": "An SI unit",
              "minLength": 1,
              "pattern": "^(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^[+-]?[1-9]\\d*)?((\u00b7(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^[+-]?[1-9]\\d*)?)|(/(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^+?[1-9]\\d*)?))*$",
              "title": "SiUnit",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        }
      },
      "required": [
        "values"
      ],
      "title": "model.v0_5.NominalOrOrdinalDataDescr",
      "type": "object"
    },
    "RelativeFilePath": {
      "description": "A path relative to the `rdf.yaml` file (also if the RDF source is a URL).",
      "format": "path",
      "title": "RelativeFilePath",
      "type": "string"
    },
    "ScaleLinearAlongAxisKwargs": {
      "additionalProperties": false,
      "description": "Key word arguments for [ScaleLinearDescr][]",
      "properties": {
        "axis": {
          "description": "The axis of gain and offset values.",
          "examples": [
            "channel"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "gain": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "items": {
                "type": "number"
              },
              "minItems": 1,
              "type": "array"
            }
          ],
          "default": 1.0,
          "description": "multiplicative factor",
          "title": "Gain"
        },
        "offset": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "items": {
                "type": "number"
              },
              "minItems": 1,
              "type": "array"
            }
          ],
          "default": 0.0,
          "description": "additive term",
          "title": "Offset"
        }
      },
      "required": [
        "axis"
      ],
      "title": "model.v0_5.ScaleLinearAlongAxisKwargs",
      "type": "object"
    },
    "ScaleLinearDescr": {
      "additionalProperties": false,
      "description": "Fixed linear scaling.\n\nExamples:\n  1. Scale with scalar gain and offset\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: scale_linear\n        kwargs:\n          gain: 2.0\n          offset: 3.0\n    ```\n    - in Python:\n\n    >>> preprocessing = [\n    ...     ScaleLinearDescr(kwargs=ScaleLinearKwargs(gain= 2.0, offset=3.0))\n    ... ]\n\n  2. Independent scaling along an axis\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: scale_linear\n        kwargs:\n          axis: 'channel'\n          gain: [1.0, 2.0, 3.0]\n    ```\n    - in Python:\n\n    >>> preprocessing = [\n    ...     ScaleLinearDescr(\n    ...         kwargs=ScaleLinearAlongAxisKwargs(\n    ...             axis=AxisId(\"channel\"),\n    ...             gain=[1.0, 2.0, 3.0],\n    ...         )\n    ...     )\n    ... ]",
      "properties": {
        "id": {
          "const": "scale_linear",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "anyOf": [
            {
              "$ref": "#/$defs/ScaleLinearKwargs"
            },
            {
              "$ref": "#/$defs/ScaleLinearAlongAxisKwargs"
            }
          ],
          "title": "Kwargs"
        }
      },
      "required": [
        "id",
        "kwargs"
      ],
      "title": "model.v0_5.ScaleLinearDescr",
      "type": "object"
    },
    "ScaleLinearKwargs": {
      "additionalProperties": false,
      "description": "Key word arguments for [ScaleLinearDescr][]",
      "properties": {
        "gain": {
          "default": 1.0,
          "description": "multiplicative factor",
          "title": "Gain",
          "type": "number"
        },
        "offset": {
          "default": 0.0,
          "description": "additive term",
          "title": "Offset",
          "type": "number"
        }
      },
      "title": "model.v0_5.ScaleLinearKwargs",
      "type": "object"
    },
    "ScaleMeanVarianceDescr": {
      "additionalProperties": false,
      "description": "Scale a tensor's data distribution to match another tensor's mean/std.\n`out  = (tensor - mean) / (std + eps) * (ref_std + eps) + ref_mean.`",
      "properties": {
        "id": {
          "const": "scale_mean_variance",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "$ref": "#/$defs/ScaleMeanVarianceKwargs"
        }
      },
      "required": [
        "id",
        "kwargs"
      ],
      "title": "model.v0_5.ScaleMeanVarianceDescr",
      "type": "object"
    },
    "ScaleMeanVarianceKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [ScaleMeanVarianceKwargs][]",
      "properties": {
        "reference_tensor": {
          "description": "ID of unprocessed input tensor to match.",
          "maxLength": 32,
          "minLength": 1,
          "title": "TensorId",
          "type": "string"
        },
        "axes": {
          "anyOf": [
            {
              "items": {
                "maxLength": 16,
                "minLength": 1,
                "title": "AxisId",
                "type": "string"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The subset of axes to normalize jointly, i.e. axes to reduce to compute mean/std.\nFor example to normalize 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x')\nresulting in a tensor of equal shape normalized per channel, specify `axes=('batch', 'x', 'y')`.\nTo normalize samples independently, leave out the 'batch' axis.\nDefault: Scale all axes jointly.",
          "examples": [
            [
              "batch",
              "x",
              "y"
            ]
          ],
          "title": "Axes"
        },
        "eps": {
          "default": 1e-06,
          "description": "Epsilon for numeric stability:\n`out  = (tensor - mean) / (std + eps) * (ref_std + eps) + ref_mean.`",
          "exclusiveMinimum": 0,
          "maximum": 0.1,
          "title": "Eps",
          "type": "number"
        }
      },
      "required": [
        "reference_tensor"
      ],
      "title": "model.v0_5.ScaleMeanVarianceKwargs",
      "type": "object"
    },
    "ScaleRangeDescr": {
      "additionalProperties": false,
      "description": "Scale with percentiles.\n\nExamples:\n1. Scale linearly to map 5th percentile to 0 and 99.8th percentile to 1.0\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: scale_range\n        kwargs:\n          axes: ['y', 'x']\n          max_percentile: 99.8\n          min_percentile: 5.0\n    ```\n    - in Python\n\n    >>> preprocessing = [\n    ...     ScaleRangeDescr(\n    ...         kwargs=ScaleRangeKwargs(\n    ...           axes= (AxisId('y'), AxisId('x')),\n    ...           max_percentile= 99.8,\n    ...           min_percentile= 5.0,\n    ...         )\n    ...     )\n    ... ]\n\n  2. Combine the above scaling with additional clipping to clip values outside the range given by the percentiles.\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: scale_range\n        kwargs:\n          axes: ['y', 'x']\n          max_percentile: 99.8\n          min_percentile: 5.0\n       - id: clip\n         kwargs:\n          min: 0.0\n          max: 1.0\n    ```\n    - in Python\n\n    >>> preprocessing = [\n    ...   ScaleRangeDescr(\n    ...     kwargs=ScaleRangeKwargs(\n    ...       axes= (AxisId('y'), AxisId('x')),\n    ...       max_percentile= 99.8,\n    ...       min_percentile= 5.0,\n    ...     )\n    ...   ),\n    ...   ClipDescr(\n    ...     kwargs=ClipKwargs(\n    ...       min=0.0,\n    ...       max=1.0,\n    ...     )\n    ...   ),\n    ... ]",
      "properties": {
        "id": {
          "const": "scale_range",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "$ref": "#/$defs/ScaleRangeKwargs"
        }
      },
      "required": [
        "id"
      ],
      "title": "model.v0_5.ScaleRangeDescr",
      "type": "object"
    },
    "ScaleRangeKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [ScaleRangeDescr][]\n\nFor `min_percentile`=0.0 (the default) and `max_percentile`=100 (the default)\nthis processing step normalizes data to the [0, 1] intervall.\nFor other percentiles the normalized values will partially be outside the [0, 1]\nintervall. Use `ScaleRange` followed by `ClipDescr` if you want to limit the\nnormalized values to a range.",
      "properties": {
        "axes": {
          "anyOf": [
            {
              "items": {
                "maxLength": 16,
                "minLength": 1,
                "title": "AxisId",
                "type": "string"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The subset of axes to normalize jointly, i.e. axes to reduce to compute the min/max percentile value.\nFor example to normalize 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x')\nresulting in a tensor of equal shape normalized per channel, specify `axes=('batch', 'x', 'y')`.\nTo normalize samples independently, leave out the \"batch\" axis.\nDefault: Scale all axes jointly.",
          "examples": [
            [
              "batch",
              "x",
              "y"
            ]
          ],
          "title": "Axes"
        },
        "min_percentile": {
          "default": 0.0,
          "description": "The lower percentile used to determine the value to align with zero.",
          "exclusiveMaximum": 100,
          "minimum": 0,
          "title": "Min Percentile",
          "type": "number"
        },
        "max_percentile": {
          "default": 100.0,
          "description": "The upper percentile used to determine the value to align with one.\nHas to be bigger than `min_percentile`.\nThe range is 1 to 100 instead of 0 to 100 to avoid mistakenly\naccepting percentiles specified in the range 0.0 to 1.0.",
          "exclusiveMinimum": 1,
          "maximum": 100,
          "title": "Max Percentile",
          "type": "number"
        },
        "eps": {
          "default": 1e-06,
          "description": "Epsilon for numeric stability.\n`out = (tensor - v_lower) / (v_upper - v_lower + eps)`;\nwith `v_lower,v_upper` values at the respective percentiles.",
          "exclusiveMinimum": 0,
          "maximum": 0.1,
          "title": "Eps",
          "type": "number"
        },
        "reference_tensor": {
          "anyOf": [
            {
              "maxLength": 32,
              "minLength": 1,
              "title": "TensorId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "ID of the unprocessed input tensor to compute the percentiles from.\nDefault: The tensor itself.",
          "title": "Reference Tensor"
        }
      },
      "title": "model.v0_5.ScaleRangeKwargs",
      "type": "object"
    },
    "SigmoidDescr": {
      "additionalProperties": false,
      "description": "The logistic sigmoid function, a.k.a. expit function.\n\nExamples:\n- in YAML\n    ```yaml\n    postprocessing:\n      - id: sigmoid\n    ```\n- in Python:\n\n    >>> postprocessing = [SigmoidDescr()]",
      "properties": {
        "id": {
          "const": "sigmoid",
          "title": "Id",
          "type": "string"
        }
      },
      "required": [
        "id"
      ],
      "title": "model.v0_5.SigmoidDescr",
      "type": "object"
    },
    "SizeReference": {
      "additionalProperties": false,
      "description": "A tensor axis size (extent in pixels/frames) defined in relation to a reference axis.\n\n`axis.size = reference.size * reference.scale / axis.scale + offset`\n\nNote:\n1. The axis and the referenced axis need to have the same unit (or no unit).\n2. Batch axes may not be referenced.\n3. Fractions are rounded down.\n4. If the reference axis is `concatenable` the referencing axis is assumed to be\n    `concatenable` as well with the same block order.\n\nExample:\nAn unisotropic input image of w*h=100*49 pixels depicts a phsical space of 200*196mm\u00b2.\nLet's assume that we want to express the image height h in relation to its width w\ninstead of only accepting input images of exactly 100*49 pixels\n(for example to express a range of valid image shapes by parametrizing w, see `ParameterizedSize`).\n\n>>> w = SpaceInputAxis(id=AxisId(\"w\"), size=100, unit=\"millimeter\", scale=2)\n>>> h = SpaceInputAxis(\n...     id=AxisId(\"h\"),\n...     size=SizeReference(tensor_id=TensorId(\"input\"), axis_id=AxisId(\"w\"), offset=-1),\n...     unit=\"millimeter\",\n...     scale=4,\n... )\n>>> print(h.size.get_size(h, w))\n49\n\n\u21d2 h = w * w.scale / h.scale + offset = 100 * 2mm / 4mm - 1 = 49",
      "properties": {
        "tensor_id": {
          "description": "tensor id of the reference axis",
          "maxLength": 32,
          "minLength": 1,
          "title": "TensorId",
          "type": "string"
        },
        "axis_id": {
          "description": "axis id of the reference axis",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "offset": {
          "default": 0,
          "title": "Offset",
          "type": "integer"
        }
      },
      "required": [
        "tensor_id",
        "axis_id"
      ],
      "title": "model.v0_5.SizeReference",
      "type": "object"
    },
    "SoftmaxDescr": {
      "additionalProperties": false,
      "description": "The softmax function.\n\nExamples:\n- in YAML\n    ```yaml\n    postprocessing:\n      - id: softmax\n        kwargs:\n          axis: channel\n    ```\n- in Python:\n\n    >>> postprocessing = [SoftmaxDescr(kwargs=SoftmaxKwargs(axis=AxisId(\"channel\")))]",
      "properties": {
        "id": {
          "const": "softmax",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "$ref": "#/$defs/SoftmaxKwargs"
        }
      },
      "required": [
        "id"
      ],
      "title": "model.v0_5.SoftmaxDescr",
      "type": "object"
    },
    "SoftmaxKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [SoftmaxDescr][]",
      "properties": {
        "axis": {
          "default": "channel",
          "description": "The axis to apply the softmax function along.\nNote:\n    Defaults to 'channel' axis\n    (which may not exist, in which case\n    a different axis id has to be specified).",
          "examples": [
            "channel"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        }
      },
      "title": "model.v0_5.SoftmaxKwargs",
      "type": "object"
    },
    "SpaceOutputAxis": {
      "additionalProperties": false,
      "properties": {
        "size": {
          "anyOf": [
            {
              "exclusiveMinimum": 0,
              "type": "integer"
            },
            {
              "$ref": "#/$defs/SizeReference"
            }
          ],
          "description": "The size/length of this axis can be specified as\n- fixed integer\n- reference to another axis with an optional offset (see [SizeReference][])",
          "examples": [
            10,
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ],
          "title": "Size"
        },
        "id": {
          "default": "x",
          "examples": [
            "x",
            "y",
            "z"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "space",
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "enum": [
                "attometer",
                "angstrom",
                "centimeter",
                "decimeter",
                "exameter",
                "femtometer",
                "foot",
                "gigameter",
                "hectometer",
                "inch",
                "kilometer",
                "megameter",
                "meter",
                "micrometer",
                "mile",
                "millimeter",
                "nanometer",
                "parsec",
                "petameter",
                "picometer",
                "terameter",
                "yard",
                "yoctometer",
                "yottameter",
                "zeptometer",
                "zettameter"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "exclusiveMinimum": 0,
          "title": "Scale",
          "type": "number"
        }
      },
      "required": [
        "size",
        "type"
      ],
      "title": "model.v0_5.SpaceOutputAxis",
      "type": "object"
    },
    "SpaceOutputAxisWithHalo": {
      "additionalProperties": false,
      "properties": {
        "halo": {
          "description": "The halo should be cropped from the output tensor to avoid boundary effects.\nIt is to be cropped from both sides, i.e. `size_after_crop = size - 2 * halo`.\nTo document a halo that is already cropped by the model use `size.offset` instead.",
          "minimum": 1,
          "title": "Halo",
          "type": "integer"
        },
        "size": {
          "$ref": "#/$defs/SizeReference",
          "description": "reference to another axis with an optional offset (see [SizeReference][])",
          "examples": [
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ]
        },
        "id": {
          "default": "x",
          "examples": [
            "x",
            "y",
            "z"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "space",
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "enum": [
                "attometer",
                "angstrom",
                "centimeter",
                "decimeter",
                "exameter",
                "femtometer",
                "foot",
                "gigameter",
                "hectometer",
                "inch",
                "kilometer",
                "megameter",
                "meter",
                "micrometer",
                "mile",
                "millimeter",
                "nanometer",
                "parsec",
                "petameter",
                "picometer",
                "terameter",
                "yard",
                "yoctometer",
                "yottameter",
                "zeptometer",
                "zettameter"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "exclusiveMinimum": 0,
          "title": "Scale",
          "type": "number"
        }
      },
      "required": [
        "halo",
        "size",
        "type"
      ],
      "title": "model.v0_5.SpaceOutputAxisWithHalo",
      "type": "object"
    },
    "StardistPostprocessingDescr": {
      "additionalProperties": false,
      "description": "Stardist postprocessing including non-maximum suppression and converting polygon representations to instance labels\n\nas described in:\n- Uwe Schmidt, Martin Weigert, Coleman Broaddus, and Gene Myers.\n[*Cell Detection with Star-convex Polygons*](https://arxiv.org/abs/1806.03535).\nInternational Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI), Granada, Spain, September 2018.\n- Martin Weigert, Uwe Schmidt, Robert Haase, Ko Sugawara, and Gene Myers.\n[*Star-convex Polyhedra for 3D Object Detection and Segmentation in Microscopy*](http://openaccess.thecvf.com/content_WACV_2020/papers/Weigert_Star-convex_Polyhedra_for_3D_Object_Detection_and_Segmentation_in_Microscopy_WACV_2020_paper.pdf).\nThe IEEE Winter Conference on Applications of Computer Vision (WACV), Snowmass Village, Colorado, March 2020.\n\nNote: Only available if the `stardist` package is installed.",
      "properties": {
        "id": {
          "const": "stardist_postprocessing",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "anyOf": [
            {
              "$ref": "#/$defs/StardistPostprocessingKwargs2D"
            },
            {
              "$ref": "#/$defs/StardistPostprocessingKwargs3D"
            }
          ],
          "title": "Kwargs"
        }
      },
      "required": [
        "id",
        "kwargs"
      ],
      "title": "model.v0_5.StardistPostprocessingDescr",
      "type": "object"
    },
    "StardistPostprocessingKwargs2D": {
      "additionalProperties": false,
      "properties": {
        "prob_threshold": {
          "description": "The probability threshold for object candidate selection.",
          "title": "Prob Threshold",
          "type": "number"
        },
        "nms_threshold": {
          "description": "The IoU threshold for non-maximum suppression.",
          "title": "Nms Threshold",
          "type": "number"
        },
        "n_rays": {
          "description": "Number of radial lines (rays) cast from the center of an object to its boundary.",
          "title": "N Rays",
          "type": "integer"
        },
        "grid": {
          "description": "Grid size of network predictions.",
          "maxItems": 2,
          "minItems": 2,
          "prefixItems": [
            {
              "type": "integer"
            },
            {
              "type": "integer"
            }
          ],
          "title": "Grid",
          "type": "array"
        },
        "b": {
          "anyOf": [
            {
              "type": "integer"
            },
            {
              "maxItems": 2,
              "minItems": 2,
              "prefixItems": [
                {
                  "maxItems": 2,
                  "minItems": 2,
                  "prefixItems": [
                    {
                      "type": "integer"
                    },
                    {
                      "type": "integer"
                    }
                  ],
                  "type": "array"
                },
                {
                  "maxItems": 2,
                  "minItems": 2,
                  "prefixItems": [
                    {
                      "type": "integer"
                    },
                    {
                      "type": "integer"
                    }
                  ],
                  "type": "array"
                }
              ],
              "type": "array"
            }
          ],
          "description": "Border region in which object probability is set to zero.",
          "title": "B"
        }
      },
      "required": [
        "prob_threshold",
        "nms_threshold",
        "n_rays",
        "grid",
        "b"
      ],
      "title": "model.v0_5.StardistPostprocessingKwargs2D",
      "type": "object"
    },
    "StardistPostprocessingKwargs3D": {
      "additionalProperties": false,
      "properties": {
        "prob_threshold": {
          "description": "The probability threshold for object candidate selection.",
          "title": "Prob Threshold",
          "type": "number"
        },
        "nms_threshold": {
          "description": "The IoU threshold for non-maximum suppression.",
          "title": "Nms Threshold",
          "type": "number"
        },
        "n_rays": {
          "description": "Number of radial lines (rays) cast from the center of an object to its boundary.",
          "title": "N Rays",
          "type": "integer"
        },
        "grid": {
          "description": "Grid size of network predictions.",
          "maxItems": 3,
          "minItems": 3,
          "prefixItems": [
            {
              "type": "integer"
            },
            {
              "type": "integer"
            },
            {
              "type": "integer"
            }
          ],
          "title": "Grid",
          "type": "array"
        },
        "b": {
          "anyOf": [
            {
              "type": "integer"
            },
            {
              "maxItems": 3,
              "minItems": 3,
              "prefixItems": [
                {
                  "maxItems": 2,
                  "minItems": 2,
                  "prefixItems": [
                    {
                      "type": "integer"
                    },
                    {
                      "type": "integer"
                    }
                  ],
                  "type": "array"
                },
                {
                  "maxItems": 2,
                  "minItems": 2,
                  "prefixItems": [
                    {
                      "type": "integer"
                    },
                    {
                      "type": "integer"
                    }
                  ],
                  "type": "array"
                },
                {
                  "maxItems": 2,
                  "minItems": 2,
                  "prefixItems": [
                    {
                      "type": "integer"
                    },
                    {
                      "type": "integer"
                    }
                  ],
                  "type": "array"
                }
              ],
              "type": "array"
            }
          ],
          "description": "Border region in which object probability is set to zero.",
          "title": "B"
        },
        "anisotropy": {
          "description": "Anisotropy factors for 3D star-convex polyhedra, i.e. the physical pixel size along each spatial axis.",
          "maxItems": 3,
          "minItems": 3,
          "prefixItems": [
            {
              "type": "number"
            },
            {
              "type": "number"
            },
            {
              "type": "number"
            }
          ],
          "title": "Anisotropy",
          "type": "array"
        },
        "overlap_label": {
          "anyOf": [
            {
              "type": "integer"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Optional label to apply to any area of overlapping predicted objects.",
          "title": "Overlap Label"
        }
      },
      "required": [
        "prob_threshold",
        "nms_threshold",
        "n_rays",
        "grid",
        "b",
        "anisotropy"
      ],
      "title": "model.v0_5.StardistPostprocessingKwargs3D",
      "type": "object"
    },
    "TimeOutputAxis": {
      "additionalProperties": false,
      "properties": {
        "size": {
          "anyOf": [
            {
              "exclusiveMinimum": 0,
              "type": "integer"
            },
            {
              "$ref": "#/$defs/SizeReference"
            }
          ],
          "description": "The size/length of this axis can be specified as\n- fixed integer\n- reference to another axis with an optional offset (see [SizeReference][])",
          "examples": [
            10,
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ],
          "title": "Size"
        },
        "id": {
          "default": "time",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "time",
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "enum": [
                "attosecond",
                "centisecond",
                "day",
                "decisecond",
                "exasecond",
                "femtosecond",
                "gigasecond",
                "hectosecond",
                "hour",
                "kilosecond",
                "megasecond",
                "microsecond",
                "millisecond",
                "minute",
                "nanosecond",
                "petasecond",
                "picosecond",
                "second",
                "terasecond",
                "yoctosecond",
                "yottasecond",
                "zeptosecond",
                "zettasecond"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "exclusiveMinimum": 0,
          "title": "Scale",
          "type": "number"
        }
      },
      "required": [
        "size",
        "type"
      ],
      "title": "model.v0_5.TimeOutputAxis",
      "type": "object"
    },
    "TimeOutputAxisWithHalo": {
      "additionalProperties": false,
      "properties": {
        "halo": {
          "description": "The halo should be cropped from the output tensor to avoid boundary effects.\nIt is to be cropped from both sides, i.e. `size_after_crop = size - 2 * halo`.\nTo document a halo that is already cropped by the model use `size.offset` instead.",
          "minimum": 1,
          "title": "Halo",
          "type": "integer"
        },
        "size": {
          "$ref": "#/$defs/SizeReference",
          "description": "reference to another axis with an optional offset (see [SizeReference][])",
          "examples": [
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ]
        },
        "id": {
          "default": "time",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "time",
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "enum": [
                "attosecond",
                "centisecond",
                "day",
                "decisecond",
                "exasecond",
                "femtosecond",
                "gigasecond",
                "hectosecond",
                "hour",
                "kilosecond",
                "megasecond",
                "microsecond",
                "millisecond",
                "minute",
                "nanosecond",
                "petasecond",
                "picosecond",
                "second",
                "terasecond",
                "yoctosecond",
                "yottasecond",
                "zeptosecond",
                "zettasecond"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "exclusiveMinimum": 0,
          "title": "Scale",
          "type": "number"
        }
      },
      "required": [
        "halo",
        "size",
        "type"
      ],
      "title": "model.v0_5.TimeOutputAxisWithHalo",
      "type": "object"
    },
    "YamlValue": {
      "anyOf": [
        {
          "type": "boolean"
        },
        {
          "format": "date",
          "type": "string"
        },
        {
          "format": "date-time",
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        },
        {
          "type": "string"
        },
        {
          "items": {
            "$ref": "#/$defs/YamlValue"
          },
          "type": "array"
        },
        {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "type": "object"
        },
        {
          "type": "null"
        }
      ]
    },
    "ZeroMeanUnitVarianceDescr": {
      "additionalProperties": false,
      "description": "Subtract mean and divide by variance.\n\nExamples:\n    Subtract tensor mean and variance\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: zero_mean_unit_variance\n    ```\n    - in Python\n    >>> preprocessing = [ZeroMeanUnitVarianceDescr()]",
      "properties": {
        "id": {
          "const": "zero_mean_unit_variance",
          "title": "Id",
          "type": "string"
        },
        "kwargs": {
          "$ref": "#/$defs/ZeroMeanUnitVarianceKwargs"
        }
      },
      "required": [
        "id"
      ],
      "title": "model.v0_5.ZeroMeanUnitVarianceDescr",
      "type": "object"
    },
    "ZeroMeanUnitVarianceKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [ZeroMeanUnitVarianceDescr][]",
      "properties": {
        "axes": {
          "anyOf": [
            {
              "items": {
                "maxLength": 16,
                "minLength": 1,
                "title": "AxisId",
                "type": "string"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The subset of axes to normalize jointly, i.e. axes to reduce to compute mean/std.\nFor example to normalize 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x')\nresulting in a tensor of equal shape normalized per channel, specify `axes=('batch', 'x', 'y')`.\nTo normalize each sample independently leave out the 'batch' axis.\nDefault: Scale all axes jointly.",
          "examples": [
            [
              "batch",
              "x",
              "y"
            ]
          ],
          "title": "Axes"
        },
        "eps": {
          "default": 1e-06,
          "description": "epsilon for numeric stability: `out = (tensor - mean) / (std + eps)`.",
          "exclusiveMinimum": 0,
          "maximum": 0.1,
          "title": "Eps",
          "type": "number"
        }
      },
      "title": "model.v0_5.ZeroMeanUnitVarianceKwargs",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "properties": {
    "id": {
      "default": "output",
      "description": "Output tensor id.\nNo duplicates are allowed across all inputs and outputs.",
      "maxLength": 32,
      "minLength": 1,
      "title": "TensorId",
      "type": "string"
    },
    "description": {
      "default": "",
      "description": "free text description",
      "maxLength": 128,
      "title": "Description",
      "type": "string"
    },
    "axes": {
      "description": "tensor axes",
      "items": {
        "discriminator": {
          "mapping": {
            "batch": "#/$defs/BatchAxis",
            "channel": "#/$defs/ChannelAxis",
            "index": "#/$defs/IndexOutputAxis",
            "space": {
              "oneOf": [
                {
                  "$ref": "#/$defs/SpaceOutputAxis"
                },
                {
                  "$ref": "#/$defs/SpaceOutputAxisWithHalo"
                }
              ]
            },
            "time": {
              "oneOf": [
                {
                  "$ref": "#/$defs/TimeOutputAxis"
                },
                {
                  "$ref": "#/$defs/TimeOutputAxisWithHalo"
                }
              ]
            }
          },
          "propertyName": "type"
        },
        "oneOf": [
          {
            "$ref": "#/$defs/BatchAxis"
          },
          {
            "$ref": "#/$defs/ChannelAxis"
          },
          {
            "$ref": "#/$defs/IndexOutputAxis"
          },
          {
            "oneOf": [
              {
                "$ref": "#/$defs/TimeOutputAxis"
              },
              {
                "$ref": "#/$defs/TimeOutputAxisWithHalo"
              }
            ]
          },
          {
            "oneOf": [
              {
                "$ref": "#/$defs/SpaceOutputAxis"
              },
              {
                "$ref": "#/$defs/SpaceOutputAxisWithHalo"
              }
            ]
          }
        ]
      },
      "minItems": 1,
      "title": "Axes",
      "type": "array"
    },
    "test_tensor": {
      "anyOf": [
        {
          "$ref": "#/$defs/FileDescr"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "An example tensor to use for testing.\nUsing the model with the test input tensors is expected to yield the test output tensors.\nEach test tensor has be a an ndarray in the\n[numpy.lib file format](https://numpy.org/doc/stable/reference/generated/numpy.lib.format.html#module-numpy.lib.format).\nThe file extension must be '.npy'."
    },
    "sample_tensor": {
      "anyOf": [
        {
          "$ref": "#/$defs/FileDescr"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "A sample tensor to illustrate a possible input/output for the model,\nThe sample image primarily serves to inform a human user about an example use case\nand is typically stored as .hdf5, .png or .tiff.\nIt has to be readable by the [imageio library](https://imageio.readthedocs.io/en/stable/formats/index.html#supported-formats)\n(numpy's `.npy` format is not supported).\nThe image dimensionality has to match the number of axes specified in this tensor description."
    },
    "data": {
      "anyOf": [
        {
          "$ref": "#/$defs/NominalOrOrdinalDataDescr"
        },
        {
          "$ref": "#/$defs/IntervalOrRatioDataDescr"
        },
        {
          "items": {
            "anyOf": [
              {
                "$ref": "#/$defs/NominalOrOrdinalDataDescr"
              },
              {
                "$ref": "#/$defs/IntervalOrRatioDataDescr"
              }
            ]
          },
          "minItems": 1,
          "type": "array"
        }
      ],
      "default": {
        "type": "float32",
        "range": [
          null,
          null
        ],
        "unit": "arbitrary unit",
        "scale": 1.0,
        "offset": null
      },
      "description": "Description of the tensor's data values, optionally per channel.\nIf specified per channel, the data `type` needs to match across channels.",
      "title": "Data"
    },
    "postprocessing": {
      "description": "Description of how this output should be postprocessed.\n\nnote: `postprocessing` always ends with an 'ensure_dtype' operation.\n      If not given this is added to cast to this tensor's `data.type`.",
      "items": {
        "discriminator": {
          "mapping": {
            "binarize": "#/$defs/BinarizeDescr",
            "cellpose_flow_dynamics": "#/$defs/CellposeFlowDynamicsDescr",
            "clip": "#/$defs/ClipDescr",
            "custom": "#/$defs/CustomProcessingDescr",
            "ensure_dtype": "#/$defs/EnsureDtypeDescr",
            "fixed_zero_mean_unit_variance": "#/$defs/FixedZeroMeanUnitVarianceDescr",
            "scale_linear": "#/$defs/ScaleLinearDescr",
            "scale_mean_variance": "#/$defs/ScaleMeanVarianceDescr",
            "scale_range": "#/$defs/ScaleRangeDescr",
            "sigmoid": "#/$defs/SigmoidDescr",
            "softmax": "#/$defs/SoftmaxDescr",
            "stardist_postprocessing": "#/$defs/StardistPostprocessingDescr",
            "zero_mean_unit_variance": "#/$defs/ZeroMeanUnitVarianceDescr"
          },
          "propertyName": "id"
        },
        "oneOf": [
          {
            "$ref": "#/$defs/BinarizeDescr"
          },
          {
            "$ref": "#/$defs/CellposeFlowDynamicsDescr"
          },
          {
            "$ref": "#/$defs/ClipDescr"
          },
          {
            "$ref": "#/$defs/CustomProcessingDescr"
          },
          {
            "$ref": "#/$defs/EnsureDtypeDescr"
          },
          {
            "$ref": "#/$defs/FixedZeroMeanUnitVarianceDescr"
          },
          {
            "$ref": "#/$defs/ScaleLinearDescr"
          },
          {
            "$ref": "#/$defs/ScaleMeanVarianceDescr"
          },
          {
            "$ref": "#/$defs/ScaleRangeDescr"
          },
          {
            "$ref": "#/$defs/SigmoidDescr"
          },
          {
            "$ref": "#/$defs/SoftmaxDescr"
          },
          {
            "$ref": "#/$defs/StardistPostprocessingDescr"
          },
          {
            "$ref": "#/$defs/ZeroMeanUnitVarianceDescr"
          }
        ]
      },
      "title": "Postprocessing",
      "type": "array"
    }
  },
  "required": [
    "axes"
  ],
  "title": "model.v0_5.OutputTensorDescr",
  "type": "object"
}

Fields:

• description (Annotated[str, MaxLen(128)])
• axes (NotEmpty[Sequence[IO_AxisT]])
• test_tensor (FAIR[FileDescr_package | None])
• sample_tensor (FAIR[FileDescr_package | None])
• data (TensorDataDescr | NotEmpty[Sequence[TensorDataDescr]])
• id (TensorId)
• postprocessing (list[PostprocessingDescr])

Validators:

• _validate_axes → axes
• _validate_sample_tensor
• _check_data_type_across_channels → data
• _check_data_matches_channelaxis
• _validate_postprocessing_kwargs

axes pydantic-field

axes: NotEmpty[Sequence[IO_AxisT]]

tensor axes

data pydantic-field

data: TensorDataDescr | NotEmpty[Sequence[TensorDataDescr]]

Description of the tensor's data values, optionally per channel. If specified per channel, the data type needs to match across channels.

description pydantic-field

description: Annotated[str, MaxLen(128)] = ''

free text description

dtype property

dtype: Literal[
    "float32",
    "float64",
    "uint8",
    "int8",
    "uint16",
    "int16",
    "uint32",
    "int32",
    "uint64",
    "int64",
    "bool",
]

dtype as specified under data.type or data[i].type

id pydantic-field

id: TensorId

Output tensor id. No duplicates are allowed across all inputs and outputs.

postprocessing pydantic-field

postprocessing: list[PostprocessingDescr]

Description of how this output should be postprocessed.

postprocessing always ends with an 'ensure_dtype' operation.

If not given this is added to cast to this tensor's data.type.

sample_tensor pydantic-field

sample_tensor: FAIR[FileDescr_package | None] = None

A sample tensor to illustrate a possible input/output for the model, The sample image primarily serves to inform a human user about an example use case and is typically stored as .hdf5, .png or .tiff. It has to be readable by the imageio library (numpy's .npy format is not supported). The image dimensionality has to match the number of axes specified in this tensor description.

shape property

shape

test_tensor pydantic-field

test_tensor: FAIR[FileDescr_package | None] = None

An example tensor to use for testing. Using the model with the test input tensors is expected to yield the test output tensors. Each test tensor has be a an ndarray in the numpy.lib file format. The file extension must be '.npy'.

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get_axis_sizes_for_array

get_axis_sizes_for_array(
    array: NDArray[Any],
) -> dict[AxisId, int]

Source code in src/bioimageio/spec/model/v0_5.py

def get_axis_sizes_for_array(self, array: NDArray[Any]) -> dict[AxisId, int]:
    if len(array.shape) != len(self.axes):
        raise ValueError(
            f"Dimension mismatch: array shape {array.shape} (#{len(array.shape)})"
            + f" incompatible with {len(self.axes)} axes."
        )
    return {a.id: array.shape[i] for i, a in enumerate(self.axes)}

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

ParameterizedSize pydantic-model

Bases: Node

Describes a range of valid tensor axis sizes as size = min + n*step.

• min and step are given by the model description.
• All blocksize paramters n = 0,1,2,... yield a valid size.
• A greater blocksize paramter n = 0,1,2,... results in a greater size. This allows to adjust the axis size more generically.

Show JSON schema:

{
  "additionalProperties": false,
  "description": "Describes a range of valid tensor axis sizes as `size = min + n*step`.\n\n- **min** and **step** are given by the model description.\n- All blocksize paramters n = 0,1,2,... yield a valid `size`.\n- A greater blocksize paramter n = 0,1,2,... results in a greater **size**.\n  This allows to adjust the axis size more generically.",
  "properties": {
    "min": {
      "exclusiveMinimum": 0,
      "title": "Min",
      "type": "integer"
    },
    "step": {
      "exclusiveMinimum": 0,
      "title": "Step",
      "type": "integer"
    }
  },
  "required": [
    "min",
    "step"
  ],
  "title": "model.v0_5.ParameterizedSize",
  "type": "object"
}

Fields:

• min (Annotated[int, Gt(0)])
• step (Annotated[int, Gt(0)])

N class-attribute

N: type[int] = ParameterizedSize_N

Positive integer to parameterize this axis

min pydantic-field

min: Annotated[int, Gt(0)]

step pydantic-field

step: Annotated[int, Gt(0)]

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get_n

get_n(s: int) -> ParameterizedSize_N

return smallest n parameterizing a size greater or equal than s

Source code in src/bioimageio/spec/model/v0_5.py

def get_n(self, s: int) -> ParameterizedSize_N:
    """return smallest n parameterizing a size greater or equal than `s`"""
    return ceil((s - self.min) / self.step)

get_size

get_size(n: ParameterizedSize_N) -> int

Source code in src/bioimageio/spec/model/v0_5.py

def get_size(self, n: ParameterizedSize_N) -> int:
    return self.min + self.step * n

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

validate_size

validate_size(size: int, msg_prefix: str = '') -> int

Source code in src/bioimageio/spec/model/v0_5.py

def validate_size(self, size: int, msg_prefix: str = "") -> int:
    if size < self.min:
        raise ValueError(
            f"{msg_prefix}size {size} < {self.min} (minimum axis size)"
        )
    if (size - self.min) % self.step != 0:
        raise ValueError(
            f"{msg_prefix}size {size} is not parameterized by `min + n*step` ="
            + f" `{self.min} + n*{self.step}`"
        )

    return size

PytorchStateDictWeightsDescr pydantic-model

Bases: WeightsEntryDescrBase

Show JSON schema:

{
  "$defs": {
    "ArchitectureFromFileDescr": {
      "additionalProperties": false,
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "Architecture source file",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        },
        "callable": {
          "description": "Identifier of the callable that returns a torch.nn.Module instance.",
          "examples": [
            "MyNetworkClass",
            "get_my_model"
          ],
          "minLength": 1,
          "title": "Identifier",
          "type": "string"
        },
        "kwargs": {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "description": "key word arguments for the `callable`",
          "title": "Kwargs",
          "type": "object"
        }
      },
      "required": [
        "source",
        "callable"
      ],
      "title": "model.v0_5.ArchitectureFromFileDescr",
      "type": "object"
    },
    "ArchitectureFromLibraryDescr": {
      "additionalProperties": false,
      "properties": {
        "callable": {
          "description": "Identifier of the callable that returns a torch.nn.Module instance.",
          "examples": [
            "MyNetworkClass",
            "get_my_model"
          ],
          "minLength": 1,
          "title": "Identifier",
          "type": "string"
        },
        "kwargs": {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "description": "key word arguments for the `callable`",
          "title": "Kwargs",
          "type": "object"
        },
        "import_from": {
          "description": "Where to import the callable from, i.e. `from <import_from> import <callable>`",
          "title": "Import From",
          "type": "string"
        }
      },
      "required": [
        "callable",
        "import_from"
      ],
      "title": "model.v0_5.ArchitectureFromLibraryDescr",
      "type": "object"
    },
    "Author": {
      "additionalProperties": false,
      "properties": {
        "affiliation": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Affiliation",
          "title": "Affiliation"
        },
        "email": {
          "anyOf": [
            {
              "format": "email",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Email",
          "title": "Email"
        },
        "orcid": {
          "anyOf": [
            {
              "description": "An ORCID identifier, see https://orcid.org/",
              "title": "OrcidId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "An [ORCID iD](https://support.orcid.org/hc/en-us/sections/360001495313-What-is-ORCID\n) in hyphenated groups of 4 digits, (and [valid](\nhttps://support.orcid.org/hc/en-us/articles/360006897674-Structure-of-the-ORCID-Identifier\n) as per ISO 7064 11,2.)",
          "examples": [
            "0000-0001-2345-6789"
          ],
          "title": "Orcid"
        },
        "name": {
          "title": "Name",
          "type": "string"
        },
        "github_user": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Github User"
        }
      },
      "required": [
        "name"
      ],
      "title": "generic.v0_3.Author",
      "type": "object"
    },
    "FileDescr": {
      "additionalProperties": false,
      "description": "A file description",
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "File source",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        }
      },
      "required": [
        "source"
      ],
      "title": "_internal.io.FileDescr",
      "type": "object"
    },
    "RelativeFilePath": {
      "description": "A path relative to the `rdf.yaml` file (also if the RDF source is a URL).",
      "format": "path",
      "title": "RelativeFilePath",
      "type": "string"
    },
    "Version": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        }
      ],
      "description": "wraps a packaging.version.Version instance for validation in pydantic models",
      "title": "Version"
    },
    "YamlValue": {
      "anyOf": [
        {
          "type": "boolean"
        },
        {
          "format": "date",
          "type": "string"
        },
        {
          "format": "date-time",
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        },
        {
          "type": "string"
        },
        {
          "items": {
            "$ref": "#/$defs/YamlValue"
          },
          "type": "array"
        },
        {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "type": "object"
        },
        {
          "type": "null"
        }
      ]
    }
  },
  "additionalProperties": false,
  "properties": {
    "source": {
      "anyOf": [
        {
          "description": "A URL with the HTTP or HTTPS scheme.",
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "title": "HttpUrl",
          "type": "string"
        },
        {
          "$ref": "#/$defs/RelativeFilePath"
        },
        {
          "format": "file-path",
          "title": "FilePath",
          "type": "string"
        }
      ],
      "description": "Source of the weights file.",
      "title": "FileSource"
    },
    "sha256": {
      "anyOf": [
        {
          "description": "A SHA-256 hash value",
          "maxLength": 64,
          "minLength": 64,
          "title": "Sha256",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "SHA256 hash value of the **source** file.",
      "title": "Sha256"
    },
    "authors": {
      "anyOf": [
        {
          "items": {
            "$ref": "#/$defs/Author"
          },
          "type": "array"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
      "title": "Authors"
    },
    "parent": {
      "anyOf": [
        {
          "enum": [
            "keras_hdf5",
            "keras_v3",
            "onnx",
            "pytorch_state_dict",
            "tensorflow_js",
            "tensorflow_saved_model_bundle",
            "torchscript"
          ],
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
      "examples": [
        "pytorch_state_dict"
      ],
      "title": "Parent"
    },
    "comment": {
      "default": "",
      "description": "A comment about this weights entry, for example how these weights were created.",
      "title": "Comment",
      "type": "string"
    },
    "architecture": {
      "anyOf": [
        {
          "$ref": "#/$defs/ArchitectureFromFileDescr"
        },
        {
          "$ref": "#/$defs/ArchitectureFromLibraryDescr"
        }
      ],
      "title": "Architecture"
    },
    "pytorch_version": {
      "$ref": "#/$defs/Version",
      "description": "Version of the PyTorch library used.\nIf `architecture.depencencies` is specified it has to include pytorch and any version pinning has to be compatible."
    },
    "dependencies": {
      "anyOf": [
        {
          "$ref": "#/$defs/FileDescr",
          "examples": [
            {
              "source": "environment.yaml"
            }
          ]
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Custom depencies beyond pytorch described in a Conda environment file.\nAllows to specify custom dependencies, see conda docs:\n- [Exporting an environment file across platforms](https://conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html#exporting-an-environment-file-across-platforms)\n- [Creating an environment file manually](https://conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html#creating-an-environment-file-manually)\n\nThe conda environment file should include pytorch and any version pinning has to be compatible with\n**pytorch_version**."
    },
    "strict": {
      "default": true,
      "description": "Whether to allow missing or unexpected keys or to be strict about the architecture matching the state dict weights.",
      "title": "Strict",
      "type": "boolean"
    }
  },
  "required": [
    "source",
    "architecture",
    "pytorch_version"
  ],
  "title": "model.v0_5.PytorchStateDictWeightsDescr",
  "type": "object"
}

Fields:

• source (Annotated[FileSource, AfterValidator(wo_special_file_name)])
• sha256 (Sha256 | None)
• authors (list[Author] | None)
• parent (Annotated[WeightsFormat | None, Field(examples=['pytorch_state_dict'])])
• comment (str)
• architecture (ArchitectureFromFileDescr | ArchitectureFromLibraryDescr)
• pytorch_version (Version)
• dependencies (FileDescr_dependencies | None)
• strict (bool)

Validators:

• _validate

architecture pydantic-field

architecture: (
    ArchitectureFromFileDescr | ArchitectureFromLibraryDescr
)

authors pydantic-field

authors: list[Author] | None = None

Authors Either the person(s) that have trained this model resulting in the original weights file. (If this is the initial weights entry, i.e. it does not have a parent) Or the person(s) who have converted the weights to this weights format. (If this is a child weight, i.e. it has a parent field)

comment pydantic-field

comment: str = ''

A comment about this weights entry, for example how these weights were created.

dependencies pydantic-field

dependencies: FileDescr_dependencies | None = None

Custom depencies beyond pytorch described in a Conda environment file. Allows to specify custom dependencies, see conda docs: - Exporting an environment file across platforms - Creating an environment file manually

The conda environment file should include pytorch and any version pinning has to be compatible with pytorch_version.

parent pydantic-field

parent: Annotated[
    WeightsFormat | None,
    Field(examples=["pytorch_state_dict"]),
] = None

The source weights these weights were converted from. For example, if a model's weights were converted from the pytorch_state_dict format to torchscript, The pytorch_state_dict weights entry has no parent and is the parent of the torchscript weights. All weight entries except one (the initial set of weights resulting from training the model), need to have this field.

pytorch_version pydantic-field

pytorch_version: Version

Version of the PyTorch library used. If architecture.depencencies is specified it has to include pytorch and any version pinning has to be compatible.

sha256 pydantic-field

sha256: Sha256 | None = None

SHA256 hash value of the source file.

source pydantic-field

source: Annotated[
    FileSource, AfterValidator(wo_special_file_name)
]

Source of the weights file.

strict pydantic-field

strict: bool = True

Whether to allow missing or unexpected keys or to be strict about the architecture matching the state dict weights.

suffix property

suffix: str

type class-attribute

type: WeightsFormat = 'pytorch_state_dict'

weights_format_name class-attribute

weights_format_name: str = 'Pytorch State Dict'

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

download

download(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

alias for .get_reader

Source code in src/bioimageio/spec/_internal/io.py

def download(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """alias for `.get_reader`"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

get_reader

get_reader(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

open the file source (download if needed)

Source code in src/bioimageio/spec/_internal/io.py

def get_reader(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """open the file source (download if needed)"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

validate_sha256

validate_sha256(force_recompute: bool = False) -> None

validate the sha256 hash value of the source file

Source code in src/bioimageio/spec/_internal/io.py

def validate_sha256(self, force_recompute: bool = False) -> None:
    """validate the sha256 hash value of the **source** file"""
    context = get_validation_context()
    src_str = str(self.source)
    if force_recompute:
        actual_sha = None
    else:
        actual_sha = context.known_files.get(src_str)

    if actual_sha is None:
        if context.perform_io_checks or force_recompute:
            reader = get_reader(self.source, sha256=self.sha256)
            if force_recompute:
                actual_sha = get_sha256(reader)
            else:
                actual_sha = reader.sha256

            context.known_files[src_str] = actual_sha
        elif context.known_files and src_str not in context.known_files:
            # perform_io_checks is False, but known files were given,
            # so we expect all file references to be in there
            raise ValueError(f"File {src_str} not found in `known_files`.")

    if actual_sha is None or self.sha256 == actual_sha:
        return
    elif self.sha256 is None or context.update_hashes:
        self.sha256 = actual_sha
    elif self.sha256 != actual_sha:
        raise ValueError(
            f"Sha256 mismatch for {self.source}. Expected {self.sha256}, got "
            + f"{actual_sha}. Update expected `sha256` or point to the matching "
            + "file."
        )

ReflectPadding pydantic-model

Bases: Node

Show JSON schema:

{
  "additionalProperties": false,
  "properties": {
    "mode": {
      "const": "reflect",
      "default": "reflect",
      "title": "Mode",
      "type": "string"
    }
  },
  "title": "model.v0_5.ReflectPadding",
  "type": "object"
}

Fields:

• mode (Literal['reflect'])

mode pydantic-field

mode: Literal['reflect'] = 'reflect'

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

RelativeFilePath

Bases: RelativePathBase[Union[AbsoluteFilePath, HttpUrl, ZipPath]]

              flowchart TD
              bioimageio.spec.model.v0_5.RelativeFilePath[RelativeFilePath]
              bioimageio.spec._internal.io.RelativePathBase[RelativePathBase]

                              bioimageio.spec._internal.io.RelativePathBase --> bioimageio.spec.model.v0_5.RelativeFilePath
                


              click bioimageio.spec.model.v0_5.RelativeFilePath href "" "bioimageio.spec.model.v0_5.RelativeFilePath"
              click bioimageio.spec._internal.io.RelativePathBase href "" "bioimageio.spec._internal.io.RelativePathBase"
            

A path relative to the rdf.yaml file (also if the RDF source is a URL).

Used by:

•  v0_2  NotebookSource
•  v0_3  NotebookSource
• API Reference  common  FileSource

Methods:

Name	Description
__repr__
__str__
absolute	get the absolute path/url
format
get_absolute
model_post_init	add validation @private

Attributes:

Name	Type	Description
parent	RelativeDirectory
path	PurePath
suffix

parent property

parent: RelativeDirectory

path property

path: PurePath

suffix property

suffix

__repr__

__repr__() -> str

Source code in src/bioimageio/spec/_internal/io.py

def __repr__(self) -> str:
    return f"RelativePath('{self}')"

__str__

__str__() -> str

Source code in src/bioimageio/spec/_internal/io.py

def __str__(self) -> str:
    return self.root.as_posix()

absolute

absolute() -> AbsolutePathT

get the absolute path/url

(resolved at time of initialization with the root of the ValidationContext)

Source code in src/bioimageio/spec/_internal/io.py

def absolute(  # method not property analog to `pathlib.Path.absolute()`
    self,
) -> AbsolutePathT:
    """get the absolute path/url

    (resolved at time of initialization with the root of the ValidationContext)
    """
    return self._absolute

format

format() -> str

Source code in src/bioimageio/spec/_internal/io.py

@model_serializer()
def format(self) -> str:
    return str(self)

get_absolute

get_absolute(
    root: RootHttpUrl | Path | pydantic.AnyUrl | ZipFile,
) -> AbsoluteFilePath | HttpUrl | ZipPath

Source code in src/bioimageio/spec/_internal/io.py

def get_absolute(
    self, root: RootHttpUrl | Path | pydantic.AnyUrl | ZipFile
) -> AbsoluteFilePath | HttpUrl | ZipPath:
    absolute = self._get_absolute_impl(root)
    if (
        isinstance(absolute, Path)
        and (context := get_validation_context()).perform_io_checks
        and str(self.root) not in context.known_files
        and not absolute.is_file()
    ):
        raise ValueError(f"{absolute} does not point to an existing file")

    return absolute

model_post_init

model_post_init(__context: Any) -> None

add validation @private

Source code in src/bioimageio/spec/_internal/io.py

def model_post_init(self, __context: Any, /) -> None:
    """add validation @private"""
    if not self.root.parts:  # an empty path can only be a directory
        raise ValueError(f"{self.root} is not a valid file path.")

    super().model_post_init(__context)

ReproducibilityTolerance pydantic-model

Bases: _ReproducibilityToleranceBase[RelativeTolerance, AbsoluteTolerance]

Show JSON schema:

{
  "additionalProperties": true,
  "properties": {
    "mismatched_elements_per_million": {
      "default": 100,
      "description": "Maximum number of mismatched elements/pixels per million to tolerate.",
      "maximum": 30000,
      "minimum": 0,
      "title": "Mismatched Elements Per Million",
      "type": "integer"
    },
    "weights_formats": {
      "default": [],
      "description": "Limits the weights formats these details apply to.",
      "items": {
        "enum": [
          "keras_hdf5",
          "keras_v3",
          "onnx",
          "pytorch_state_dict",
          "tensorflow_js",
          "tensorflow_saved_model_bundle",
          "torchscript"
        ],
        "type": "string"
      },
      "title": "Weights Formats",
      "type": "array"
    },
    "relative_tolerance": {
      "default": 0.001,
      "description": "Maximum relative tolerance of reproduced test tensor.\nNeeds to be 0 for discrete (integer and boolean) outputs.",
      "maximum": 0.01,
      "minimum": 0,
      "title": "Relative Tolerance",
      "type": "number"
    },
    "absolute_tolerance": {
      "default": 0.001,
      "description": "Maximum absolute tolerance of reproduced test tensor.\n\nThe `absolute_tolerance` may not be greater than 1% of the maximum absolute value of the test tensor\n(this serves to prevent false positives in the presence of very small numbers).\nNeeds to be 0 for discrete (integer and boolean) outputs.",
      "minimum": 0,
      "title": "Absolute Tolerance",
      "type": "number"
    },
    "output_ids": {
      "default": [],
      "description": "Limits the output tensor IDs these reproducibility details apply to.\n\nIf empty, the reproducibility details apply to all outputs.",
      "items": {
        "maxLength": 32,
        "minLength": 1,
        "title": "TensorId",
        "type": "string"
      },
      "title": "Output Ids",
      "type": "array"
    }
  },
  "title": "model.v0_5.ReproducibilityTolerance",
  "type": "object"
}

Fields:

• mismatched_elements_per_million (MismatchedElementsPerMillion)
• weights_formats (Sequence[WeightsFormat])
• output_ids (Sequence[TensorId])
• relative_tolerance (RelativeTolerance)
• absolute_tolerance (AbsoluteTolerance)

absolute_tolerance pydantic-field

absolute_tolerance: AbsoluteTolerance = 0.001

Maximum absolute tolerance of reproduced test tensor.

The absolute_tolerance may not be greater than 1% of the maximum absolute value of the test tensor (this serves to prevent false positives in the presence of very small numbers). Needs to be 0 for discrete (integer and boolean) outputs.

mismatched_elements_per_million pydantic-field

mismatched_elements_per_million: MismatchedElementsPerMillion = 100

Maximum number of mismatched elements/pixels per million to tolerate.

output_ids pydantic-field

output_ids: Sequence[TensorId] = ()

Limits the output tensor IDs these reproducibility details apply to.

If empty, the reproducibility details apply to all outputs.

relative_tolerance pydantic-field

relative_tolerance: RelativeTolerance = 0.001

Maximum relative tolerance of reproduced test tensor. Needs to be 0 for discrete (integer and boolean) outputs.

weights_formats pydantic-field

weights_formats: Sequence[WeightsFormat] = ()

Limits the weights formats these details apply to.

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

serialize

serialize(
    handler: SerializerFunctionWrapHandler,
) -> dict[str, Any]

explicitly serialize tolerance values

(even if they are default values and exclude_defaults=True).

Source code in src/bioimageio/spec/model/v0_5.py

@model_serializer(mode="wrap")
def serialize(
    self,
    handler: SerializerFunctionWrapHandler,
) -> dict[str, Any]:
    """explicitly serialize tolerance values

    (even if they are default values and `exclude_defaults=True`).
    """
    data = handler(self)
    data["relative_tolerance"] = self.relative_tolerance
    data["absolute_tolerance"] = self.absolute_tolerance
    data["mismatched_elements_per_million"] = self.mismatched_elements_per_million
    return data

ResourceId

Bases: ValidatedString

              flowchart TD
              bioimageio.spec.model.v0_5.ResourceId[ResourceId]
              bioimageio.spec._internal.validated_string.ValidatedString[ValidatedString]

                              bioimageio.spec._internal.validated_string.ValidatedString --> bioimageio.spec.model.v0_5.ResourceId
                


              click bioimageio.spec.model.v0_5.ResourceId href "" "bioimageio.spec.model.v0_5.ResourceId"
              click bioimageio.spec._internal.validated_string.ValidatedString href "" "bioimageio.spec._internal.validated_string.ValidatedString"
            

Subclassed by:

•  v0_3  ApplicationId
•  v0_3  DatasetId
•  v0_3  NotebookId
•  v0_5
  •  DatasetId
  •  ModelId

Methods:

Name	Description
__get_pydantic_core_schema__
__get_pydantic_json_schema__
__new__

Attributes:

Name	Type	Description
root_model	type[RootModel[Any]]	the pydantic root model to validate the string

root_model class-attribute

root_model: type[RootModel[Any]] = RootModel[
    Annotated[
        NotEmpty[str],
        RestrictCharacters(
            string.ascii_lowercase + string.digits + "_-/."
        ),
        annotated_types.Predicate(
            lambda s: (
                not (
                    s.startswith("/") or s.endswith("/")
                )
            )
        ),
    ]
]

the pydantic root model to validate the string

__get_pydantic_core_schema__ classmethod

__get_pydantic_core_schema__(
    source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_core_schema__(
    cls, source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema:
    return no_info_after_validator_function(cls, handler(str))

__get_pydantic_json_schema__ classmethod

__get_pydantic_json_schema__(
    core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_json_schema__(
    cls, core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue:
    json_schema = cls.root_model.model_json_schema(mode=handler.mode)
    json_schema["title"] = cls.__name__.strip("_")
    if cls.__doc__:
        json_schema["description"] = cls.__doc__

    return json_schema

__new__

__new__(object: object)

Source code in src/bioimageio/spec/_internal/validated_string.py

def __new__(cls, object: object):
    _validated = cls.root_model.model_validate(object).root
    self = super().__new__(cls, _validated)
    self._validated = _validated
    return self._after_validator()

RunMode pydantic-model

Bases: Node

Show JSON schema:

{
  "additionalProperties": false,
  "properties": {
    "name": {
      "anyOf": [
        {
          "const": "deepimagej",
          "type": "string"
        },
        {
          "type": "string"
        }
      ],
      "description": "Run mode name",
      "title": "Name"
    },
    "kwargs": {
      "additionalProperties": true,
      "description": "Run mode specific key word arguments",
      "title": "Kwargs",
      "type": "object"
    }
  },
  "required": [
    "name"
  ],
  "title": "model.v0_4.RunMode",
  "type": "object"
}

Fields:

• name (Annotated[KnownRunMode | str, warn(KnownRunMode, "Unknown run mode '{value}'.")])
• kwargs (dict[str, Any])

kwargs pydantic-field

kwargs: dict[str, Any]

Run mode specific key word arguments

name pydantic-field

name: Annotated[
    KnownRunMode | str,
    warn(KnownRunMode, "Unknown run mode '{value}'."),
]

Run mode name

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

ScaleLinearAlongAxisKwargs pydantic-model

Bases: KwargsNode

Key word arguments for ScaleLinearDescr

Show JSON schema:

{
  "additionalProperties": false,
  "description": "Key word arguments for [ScaleLinearDescr][]",
  "properties": {
    "axis": {
      "description": "The axis of gain and offset values.",
      "examples": [
        "channel"
      ],
      "maxLength": 16,
      "minLength": 1,
      "title": "AxisId",
      "type": "string"
    },
    "gain": {
      "anyOf": [
        {
          "type": "number"
        },
        {
          "items": {
            "type": "number"
          },
          "minItems": 1,
          "type": "array"
        }
      ],
      "default": 1.0,
      "description": "multiplicative factor",
      "title": "Gain"
    },
    "offset": {
      "anyOf": [
        {
          "type": "number"
        },
        {
          "items": {
            "type": "number"
          },
          "minItems": 1,
          "type": "array"
        }
      ],
      "default": 0.0,
      "description": "additive term",
      "title": "Offset"
    }
  },
  "required": [
    "axis"
  ],
  "title": "model.v0_5.ScaleLinearAlongAxisKwargs",
  "type": "object"
}

Fields:

• axis (Annotated[NonBatchAxisId, Field(examples=['channel'])])
• gain (float | NotEmpty[list[float]])
• offset (float | NotEmpty[list[float]])

Validators:

• _validate

axis pydantic-field

axis: Annotated[NonBatchAxisId, Field(examples=["channel"])]

The axis of gain and offset values.

gain pydantic-field

gain: float | NotEmpty[list[float]] = 1.0

multiplicative factor

offset pydantic-field

offset: float | NotEmpty[list[float]] = 0.0

additive term

__contains__

__contains__(item: str) -> bool

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __contains__(self, item: str) -> bool:
    return item in self.__class__.model_fields

__getitem__

__getitem__(item: str) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __getitem__(self, item: str) -> Any:
    if item in self.__class__.model_fields:
        return getattr(self, item)
    else:
        raise KeyError(item)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get

get(item: str, default: Any = None) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def get(self, item: str, default: Any = None) -> Any:
    return self[item] if item in self else default  # ruff: ignore[SIM401]

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

ScaleLinearDescr pydantic-model

Bases: NodeWithExplicitlySetFields

Fixed linear scaling.

Examples:

1. Scale with scalar gain and offset
2. in YAML

   preprocessing:
     - id: scale_linear
       kwargs:
         gain: 2.0
         offset: 3.0
   

3. in Python:

preprocessing = [ ... ScaleLinearDescr(kwargs=ScaleLinearKwargs(gain= 2.0, offset=3.0)) ... ]

1. Independent scaling along an axis
2. in YAML

   preprocessing:
     - id: scale_linear
       kwargs:
         axis: 'channel'
         gain: [1.0, 2.0, 3.0]
   

3. in Python:

preprocessing = [ ... ScaleLinearDescr( ... kwargs=ScaleLinearAlongAxisKwargs( ... axis=AxisId("channel"), ... gain=[1.0, 2.0, 3.0], ... ) ... ) ... ]

Show JSON schema:

{
  "$defs": {
    "ScaleLinearAlongAxisKwargs": {
      "additionalProperties": false,
      "description": "Key word arguments for [ScaleLinearDescr][]",
      "properties": {
        "axis": {
          "description": "The axis of gain and offset values.",
          "examples": [
            "channel"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "gain": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "items": {
                "type": "number"
              },
              "minItems": 1,
              "type": "array"
            }
          ],
          "default": 1.0,
          "description": "multiplicative factor",
          "title": "Gain"
        },
        "offset": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "items": {
                "type": "number"
              },
              "minItems": 1,
              "type": "array"
            }
          ],
          "default": 0.0,
          "description": "additive term",
          "title": "Offset"
        }
      },
      "required": [
        "axis"
      ],
      "title": "model.v0_5.ScaleLinearAlongAxisKwargs",
      "type": "object"
    },
    "ScaleLinearKwargs": {
      "additionalProperties": false,
      "description": "Key word arguments for [ScaleLinearDescr][]",
      "properties": {
        "gain": {
          "default": 1.0,
          "description": "multiplicative factor",
          "title": "Gain",
          "type": "number"
        },
        "offset": {
          "default": 0.0,
          "description": "additive term",
          "title": "Offset",
          "type": "number"
        }
      },
      "title": "model.v0_5.ScaleLinearKwargs",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "description": "Fixed linear scaling.\n\nExamples:\n  1. Scale with scalar gain and offset\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: scale_linear\n        kwargs:\n          gain: 2.0\n          offset: 3.0\n    ```\n    - in Python:\n\n    >>> preprocessing = [\n    ...     ScaleLinearDescr(kwargs=ScaleLinearKwargs(gain= 2.0, offset=3.0))\n    ... ]\n\n  2. Independent scaling along an axis\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: scale_linear\n        kwargs:\n          axis: 'channel'\n          gain: [1.0, 2.0, 3.0]\n    ```\n    - in Python:\n\n    >>> preprocessing = [\n    ...     ScaleLinearDescr(\n    ...         kwargs=ScaleLinearAlongAxisKwargs(\n    ...             axis=AxisId(\"channel\"),\n    ...             gain=[1.0, 2.0, 3.0],\n    ...         )\n    ...     )\n    ... ]",
  "properties": {
    "id": {
      "const": "scale_linear",
      "title": "Id",
      "type": "string"
    },
    "kwargs": {
      "anyOf": [
        {
          "$ref": "#/$defs/ScaleLinearKwargs"
        },
        {
          "$ref": "#/$defs/ScaleLinearAlongAxisKwargs"
        }
      ],
      "title": "Kwargs"
    }
  },
  "required": [
    "id",
    "kwargs"
  ],
  "title": "model.v0_5.ScaleLinearDescr",
  "type": "object"
}

Fields:

• id (Literal['scale_linear'])
• kwargs (ScaleLinearKwargs | ScaleLinearAlongAxisKwargs)

id pydantic-field

id: Literal['scale_linear'] = 'scale_linear'

implemented_id class-attribute

implemented_id: Literal['scale_linear'] = 'scale_linear'

kwargs pydantic-field

kwargs: ScaleLinearKwargs | ScaleLinearAlongAxisKwargs

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

ScaleLinearKwargs pydantic-model

Bases: KwargsNode

Key word arguments for ScaleLinearDescr

Show JSON schema:

{
  "additionalProperties": false,
  "description": "Key word arguments for [ScaleLinearDescr][]",
  "properties": {
    "gain": {
      "default": 1.0,
      "description": "multiplicative factor",
      "title": "Gain",
      "type": "number"
    },
    "offset": {
      "default": 0.0,
      "description": "additive term",
      "title": "Offset",
      "type": "number"
    }
  },
  "title": "model.v0_5.ScaleLinearKwargs",
  "type": "object"
}

Fields:

• gain (float)
• offset (float)

Validators:

• _validate

gain pydantic-field

gain: float = 1.0

multiplicative factor

offset pydantic-field

offset: float = 0.0

additive term

__contains__

__contains__(item: str) -> bool

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __contains__(self, item: str) -> bool:
    return item in self.__class__.model_fields

__getitem__

__getitem__(item: str) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __getitem__(self, item: str) -> Any:
    if item in self.__class__.model_fields:
        return getattr(self, item)
    else:
        raise KeyError(item)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get

get(item: str, default: Any = None) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def get(self, item: str, default: Any = None) -> Any:
    return self[item] if item in self else default  # ruff: ignore[SIM401]

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

ScaleMeanVarianceDescr pydantic-model

Bases: NodeWithExplicitlySetFields

Scale a tensor's data distribution to match another tensor's mean/std. out = (tensor - mean) / (std + eps) * (ref_std + eps) + ref_mean.

Show JSON schema:

{
  "$defs": {
    "ScaleMeanVarianceKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [ScaleMeanVarianceKwargs][]",
      "properties": {
        "reference_tensor": {
          "description": "ID of unprocessed input tensor to match.",
          "maxLength": 32,
          "minLength": 1,
          "title": "TensorId",
          "type": "string"
        },
        "axes": {
          "anyOf": [
            {
              "items": {
                "maxLength": 16,
                "minLength": 1,
                "title": "AxisId",
                "type": "string"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The subset of axes to normalize jointly, i.e. axes to reduce to compute mean/std.\nFor example to normalize 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x')\nresulting in a tensor of equal shape normalized per channel, specify `axes=('batch', 'x', 'y')`.\nTo normalize samples independently, leave out the 'batch' axis.\nDefault: Scale all axes jointly.",
          "examples": [
            [
              "batch",
              "x",
              "y"
            ]
          ],
          "title": "Axes"
        },
        "eps": {
          "default": 1e-06,
          "description": "Epsilon for numeric stability:\n`out  = (tensor - mean) / (std + eps) * (ref_std + eps) + ref_mean.`",
          "exclusiveMinimum": 0,
          "maximum": 0.1,
          "title": "Eps",
          "type": "number"
        }
      },
      "required": [
        "reference_tensor"
      ],
      "title": "model.v0_5.ScaleMeanVarianceKwargs",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "description": "Scale a tensor's data distribution to match another tensor's mean/std.\n`out  = (tensor - mean) / (std + eps) * (ref_std + eps) + ref_mean.`",
  "properties": {
    "id": {
      "const": "scale_mean_variance",
      "title": "Id",
      "type": "string"
    },
    "kwargs": {
      "$ref": "#/$defs/ScaleMeanVarianceKwargs"
    }
  },
  "required": [
    "id",
    "kwargs"
  ],
  "title": "model.v0_5.ScaleMeanVarianceDescr",
  "type": "object"
}

Fields:

• id (Literal['scale_mean_variance'])
• kwargs (ScaleMeanVarianceKwargs)

id pydantic-field

id: Literal['scale_mean_variance'] = 'scale_mean_variance'

implemented_id class-attribute

implemented_id: Literal["scale_mean_variance"] = (
    "scale_mean_variance"
)

kwargs pydantic-field

kwargs: ScaleMeanVarianceKwargs

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

ScaleMeanVarianceKwargs pydantic-model

Bases: KwargsNode

key word arguments for ScaleMeanVarianceKwargs

Show JSON schema:

{
  "additionalProperties": false,
  "description": "key word arguments for [ScaleMeanVarianceKwargs][]",
  "properties": {
    "reference_tensor": {
      "description": "ID of unprocessed input tensor to match.",
      "maxLength": 32,
      "minLength": 1,
      "title": "TensorId",
      "type": "string"
    },
    "axes": {
      "anyOf": [
        {
          "items": {
            "maxLength": 16,
            "minLength": 1,
            "title": "AxisId",
            "type": "string"
          },
          "type": "array"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The subset of axes to normalize jointly, i.e. axes to reduce to compute mean/std.\nFor example to normalize 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x')\nresulting in a tensor of equal shape normalized per channel, specify `axes=('batch', 'x', 'y')`.\nTo normalize samples independently, leave out the 'batch' axis.\nDefault: Scale all axes jointly.",
      "examples": [
        [
          "batch",
          "x",
          "y"
        ]
      ],
      "title": "Axes"
    },
    "eps": {
      "default": 1e-06,
      "description": "Epsilon for numeric stability:\n`out  = (tensor - mean) / (std + eps) * (ref_std + eps) + ref_mean.`",
      "exclusiveMinimum": 0,
      "maximum": 0.1,
      "title": "Eps",
      "type": "number"
    }
  },
  "required": [
    "reference_tensor"
  ],
  "title": "model.v0_5.ScaleMeanVarianceKwargs",
  "type": "object"
}

Fields:

• reference_tensor (TensorId)
• axes (Annotated[Sequence[AxisId] | None, Field(examples=[('batch', 'x', 'y')])])
• eps (Annotated[float, Interval(gt=0, le=0.1)])

axes pydantic-field

axes: Annotated[
    Sequence[AxisId] | None,
    Field(examples=[("batch", "x", "y")]),
] = None

The subset of axes to normalize jointly, i.e. axes to reduce to compute mean/std. For example to normalize 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x') resulting in a tensor of equal shape normalized per channel, specify axes=('batch', 'x', 'y'). To normalize samples independently, leave out the 'batch' axis. Default: Scale all axes jointly.

eps pydantic-field

eps: Annotated[float, Interval(gt=0, le=0.1)] = 1e-06

Epsilon for numeric stability: out = (tensor - mean) / (std + eps) * (ref_std + eps) + ref_mean.

reference_tensor pydantic-field

reference_tensor: TensorId

ID of unprocessed input tensor to match.

__contains__

__contains__(item: str) -> bool

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __contains__(self, item: str) -> bool:
    return item in self.__class__.model_fields

__getitem__

__getitem__(item: str) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __getitem__(self, item: str) -> Any:
    if item in self.__class__.model_fields:
        return getattr(self, item)
    else:
        raise KeyError(item)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get

get(item: str, default: Any = None) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def get(self, item: str, default: Any = None) -> Any:
    return self[item] if item in self else default  # ruff: ignore[SIM401]

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

ScaleRangeDescr pydantic-model

Bases: NodeWithExplicitlySetFields

Scale with percentiles.

Examples:

1. Scale linearly to map 5th percentile to 0 and 99.8th percentile to 1.0

   • in YAML

     preprocessing:
       - id: scale_range
         kwargs:
           axes: ['y', 'x']
           max_percentile: 99.8
           min_percentile: 5.0
     

   • in Python

   >>> preprocessing = [
   ...     ScaleRangeDescr(
   ...         kwargs=ScaleRangeKwargs(
   ...           axes= (AxisId('y'), AxisId('x')),
   ...           max_percentile= 99.8,
   ...           min_percentile= 5.0,
   ...         )
   ...     )
   ... ]
   

2. Combine the above scaling with additional clipping to clip values outside the range given by the percentiles.

   • in YAML

     preprocessing:
       - id: scale_range
         kwargs:
           axes: ['y', 'x']
           max_percentile: 99.8
           min_percentile: 5.0
        - id: clip
          kwargs:
           min: 0.0
           max: 1.0
     

   • in Python

   >>> preprocessing = [
   ...   ScaleRangeDescr(
   ...     kwargs=ScaleRangeKwargs(
   ...       axes= (AxisId('y'), AxisId('x')),
   ...       max_percentile= 99.8,
   ...       min_percentile= 5.0,
   ...     )
   ...   ),
   ...   ClipDescr(
   ...     kwargs=ClipKwargs(
   ...       min=0.0,
   ...       max=1.0,
   ...     )
   ...   ),
   ... ]
   

Show JSON schema:

{
  "$defs": {
    "ScaleRangeKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [ScaleRangeDescr][]\n\nFor `min_percentile`=0.0 (the default) and `max_percentile`=100 (the default)\nthis processing step normalizes data to the [0, 1] intervall.\nFor other percentiles the normalized values will partially be outside the [0, 1]\nintervall. Use `ScaleRange` followed by `ClipDescr` if you want to limit the\nnormalized values to a range.",
      "properties": {
        "axes": {
          "anyOf": [
            {
              "items": {
                "maxLength": 16,
                "minLength": 1,
                "title": "AxisId",
                "type": "string"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The subset of axes to normalize jointly, i.e. axes to reduce to compute the min/max percentile value.\nFor example to normalize 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x')\nresulting in a tensor of equal shape normalized per channel, specify `axes=('batch', 'x', 'y')`.\nTo normalize samples independently, leave out the \"batch\" axis.\nDefault: Scale all axes jointly.",
          "examples": [
            [
              "batch",
              "x",
              "y"
            ]
          ],
          "title": "Axes"
        },
        "min_percentile": {
          "default": 0.0,
          "description": "The lower percentile used to determine the value to align with zero.",
          "exclusiveMaximum": 100,
          "minimum": 0,
          "title": "Min Percentile",
          "type": "number"
        },
        "max_percentile": {
          "default": 100.0,
          "description": "The upper percentile used to determine the value to align with one.\nHas to be bigger than `min_percentile`.\nThe range is 1 to 100 instead of 0 to 100 to avoid mistakenly\naccepting percentiles specified in the range 0.0 to 1.0.",
          "exclusiveMinimum": 1,
          "maximum": 100,
          "title": "Max Percentile",
          "type": "number"
        },
        "eps": {
          "default": 1e-06,
          "description": "Epsilon for numeric stability.\n`out = (tensor - v_lower) / (v_upper - v_lower + eps)`;\nwith `v_lower,v_upper` values at the respective percentiles.",
          "exclusiveMinimum": 0,
          "maximum": 0.1,
          "title": "Eps",
          "type": "number"
        },
        "reference_tensor": {
          "anyOf": [
            {
              "maxLength": 32,
              "minLength": 1,
              "title": "TensorId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "ID of the unprocessed input tensor to compute the percentiles from.\nDefault: The tensor itself.",
          "title": "Reference Tensor"
        }
      },
      "title": "model.v0_5.ScaleRangeKwargs",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "description": "Scale with percentiles.\n\nExamples:\n1. Scale linearly to map 5th percentile to 0 and 99.8th percentile to 1.0\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: scale_range\n        kwargs:\n          axes: ['y', 'x']\n          max_percentile: 99.8\n          min_percentile: 5.0\n    ```\n    - in Python\n\n    >>> preprocessing = [\n    ...     ScaleRangeDescr(\n    ...         kwargs=ScaleRangeKwargs(\n    ...           axes= (AxisId('y'), AxisId('x')),\n    ...           max_percentile= 99.8,\n    ...           min_percentile= 5.0,\n    ...         )\n    ...     )\n    ... ]\n\n  2. Combine the above scaling with additional clipping to clip values outside the range given by the percentiles.\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: scale_range\n        kwargs:\n          axes: ['y', 'x']\n          max_percentile: 99.8\n          min_percentile: 5.0\n       - id: clip\n         kwargs:\n          min: 0.0\n          max: 1.0\n    ```\n    - in Python\n\n    >>> preprocessing = [\n    ...   ScaleRangeDescr(\n    ...     kwargs=ScaleRangeKwargs(\n    ...       axes= (AxisId('y'), AxisId('x')),\n    ...       max_percentile= 99.8,\n    ...       min_percentile= 5.0,\n    ...     )\n    ...   ),\n    ...   ClipDescr(\n    ...     kwargs=ClipKwargs(\n    ...       min=0.0,\n    ...       max=1.0,\n    ...     )\n    ...   ),\n    ... ]",
  "properties": {
    "id": {
      "const": "scale_range",
      "title": "Id",
      "type": "string"
    },
    "kwargs": {
      "$ref": "#/$defs/ScaleRangeKwargs"
    }
  },
  "required": [
    "id"
  ],
  "title": "model.v0_5.ScaleRangeDescr",
  "type": "object"
}

Fields:

• id (Literal['scale_range'])
• kwargs (ScaleRangeKwargs)

id pydantic-field

id: Literal['scale_range'] = 'scale_range'

implemented_id class-attribute

implemented_id: Literal['scale_range'] = 'scale_range'

kwargs pydantic-field

kwargs: ScaleRangeKwargs

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

ScaleRangeKwargs pydantic-model

Bases: KwargsNode

key word arguments for ScaleRangeDescr

For min_percentile=0.0 (the default) and max_percentile=100 (the default) this processing step normalizes data to the [0, 1] intervall. For other percentiles the normalized values will partially be outside the [0, 1] intervall. Use ScaleRange followed by ClipDescr if you want to limit the normalized values to a range.

Show JSON schema:

{
  "additionalProperties": false,
  "description": "key word arguments for [ScaleRangeDescr][]\n\nFor `min_percentile`=0.0 (the default) and `max_percentile`=100 (the default)\nthis processing step normalizes data to the [0, 1] intervall.\nFor other percentiles the normalized values will partially be outside the [0, 1]\nintervall. Use `ScaleRange` followed by `ClipDescr` if you want to limit the\nnormalized values to a range.",
  "properties": {
    "axes": {
      "anyOf": [
        {
          "items": {
            "maxLength": 16,
            "minLength": 1,
            "title": "AxisId",
            "type": "string"
          },
          "type": "array"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The subset of axes to normalize jointly, i.e. axes to reduce to compute the min/max percentile value.\nFor example to normalize 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x')\nresulting in a tensor of equal shape normalized per channel, specify `axes=('batch', 'x', 'y')`.\nTo normalize samples independently, leave out the \"batch\" axis.\nDefault: Scale all axes jointly.",
      "examples": [
        [
          "batch",
          "x",
          "y"
        ]
      ],
      "title": "Axes"
    },
    "min_percentile": {
      "default": 0.0,
      "description": "The lower percentile used to determine the value to align with zero.",
      "exclusiveMaximum": 100,
      "minimum": 0,
      "title": "Min Percentile",
      "type": "number"
    },
    "max_percentile": {
      "default": 100.0,
      "description": "The upper percentile used to determine the value to align with one.\nHas to be bigger than `min_percentile`.\nThe range is 1 to 100 instead of 0 to 100 to avoid mistakenly\naccepting percentiles specified in the range 0.0 to 1.0.",
      "exclusiveMinimum": 1,
      "maximum": 100,
      "title": "Max Percentile",
      "type": "number"
    },
    "eps": {
      "default": 1e-06,
      "description": "Epsilon for numeric stability.\n`out = (tensor - v_lower) / (v_upper - v_lower + eps)`;\nwith `v_lower,v_upper` values at the respective percentiles.",
      "exclusiveMinimum": 0,
      "maximum": 0.1,
      "title": "Eps",
      "type": "number"
    },
    "reference_tensor": {
      "anyOf": [
        {
          "maxLength": 32,
          "minLength": 1,
          "title": "TensorId",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "ID of the unprocessed input tensor to compute the percentiles from.\nDefault: The tensor itself.",
      "title": "Reference Tensor"
    }
  },
  "title": "model.v0_5.ScaleRangeKwargs",
  "type": "object"
}

Fields:

• axes (Annotated[Sequence[AxisId] | None, Field(examples=[('batch', 'x', 'y')])])
• min_percentile (Annotated[float, Interval(ge=0, lt=100)])
• max_percentile (Annotated[float, Interval(gt=1, le=100)])
• eps (Annotated[float, Interval(gt=0, le=0.1)])
• reference_tensor (TensorId | None)

Validators:

• min_smaller_max → max_percentile

axes pydantic-field

axes: Annotated[
    Sequence[AxisId] | None,
    Field(examples=[("batch", "x", "y")]),
] = None

The subset of axes to normalize jointly, i.e. axes to reduce to compute the min/max percentile value. For example to normalize 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x') resulting in a tensor of equal shape normalized per channel, specify axes=('batch', 'x', 'y'). To normalize samples independently, leave out the "batch" axis. Default: Scale all axes jointly.

eps pydantic-field

eps: Annotated[float, Interval(gt=0, le=0.1)] = 1e-06

Epsilon for numeric stability. out = (tensor - v_lower) / (v_upper - v_lower + eps); with v_lower,v_upper values at the respective percentiles.

max_percentile pydantic-field

max_percentile: Annotated[
    float, Interval(gt=1, le=100)
] = 100.0

The upper percentile used to determine the value to align with one. Has to be bigger than min_percentile. The range is 1 to 100 instead of 0 to 100 to avoid mistakenly accepting percentiles specified in the range 0.0 to 1.0.

min_percentile pydantic-field

min_percentile: Annotated[
    float, Interval(ge=0, lt=100)
] = 0.0

The lower percentile used to determine the value to align with zero.

reference_tensor pydantic-field

reference_tensor: TensorId | None = None

ID of the unprocessed input tensor to compute the percentiles from. Default: The tensor itself.

__contains__

__contains__(item: str) -> bool

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __contains__(self, item: str) -> bool:
    return item in self.__class__.model_fields

__getitem__

__getitem__(item: str) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __getitem__(self, item: str) -> Any:
    if item in self.__class__.model_fields:
        return getattr(self, item)
    else:
        raise KeyError(item)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get

get(item: str, default: Any = None) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def get(self, item: str, default: Any = None) -> Any:
    return self[item] if item in self else default  # ruff: ignore[SIM401]

min_smaller_max pydantic-validator

min_smaller_max(
    value: float, info: ValidationInfo
) -> float

Source code in src/bioimageio/spec/model/v0_5.py

@field_validator("max_percentile", mode="after")
@classmethod
def min_smaller_max(cls, value: float, info: ValidationInfo) -> float:
    if (min_p := info.data["min_percentile"]) >= value:
        raise ValueError(f"min_percentile {min_p} >= max_percentile {value}")

    return value

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

Sha256

Bases: ValidatedString

              flowchart TD
              bioimageio.spec.model.v0_5.Sha256[Sha256]
              bioimageio.spec._internal.validated_string.ValidatedString[ValidatedString]

                              bioimageio.spec._internal.validated_string.ValidatedString --> bioimageio.spec.model.v0_5.Sha256
                


              click bioimageio.spec.model.v0_5.Sha256 href "" "bioimageio.spec.model.v0_5.Sha256"
              click bioimageio.spec._internal.validated_string.ValidatedString href "" "bioimageio.spec._internal.validated_string.ValidatedString"
            

A SHA-256 hash value

Returned by:

• API Reference  utils  get_sha256

Used by:

• API Reference
  •  common  BytesReader
  •  spec
    •  load_dataset_description
    •  load_description
    •  load_description_and_validate_format_only
    •  load_model_description

Methods:

Name	Description
__get_pydantic_core_schema__
__get_pydantic_json_schema__
__new__

Attributes:

Name	Type	Description
root_model	type[RootModel[Any]]	the pydantic root model to validate the string

root_model class-attribute

root_model: type[RootModel[Any]] = RootModel[
    Annotated[
        str,
        StringConstraints(
            strip_whitespace=True,
            to_lower=True,
            min_length=64,
            max_length=64,
        ),
    ]
]

the pydantic root model to validate the string

__get_pydantic_core_schema__ classmethod

__get_pydantic_core_schema__(
    source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_core_schema__(
    cls, source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema:
    return no_info_after_validator_function(cls, handler(str))

__get_pydantic_json_schema__ classmethod

__get_pydantic_json_schema__(
    core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_json_schema__(
    cls, core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue:
    json_schema = cls.root_model.model_json_schema(mode=handler.mode)
    json_schema["title"] = cls.__name__.strip("_")
    if cls.__doc__:
        json_schema["description"] = cls.__doc__

    return json_schema

__new__

__new__(object: object)

Source code in src/bioimageio/spec/_internal/validated_string.py

def __new__(cls, object: object):
    _validated = cls.root_model.model_validate(object).root
    self = super().__new__(cls, _validated)
    self._validated = _validated
    return self._after_validator()

SiUnit

Bases: ValidatedString

              flowchart TD
              bioimageio.spec.model.v0_5.SiUnit[SiUnit]
              bioimageio.spec._internal.validated_string.ValidatedString[ValidatedString]

                              bioimageio.spec._internal.validated_string.ValidatedString --> bioimageio.spec.model.v0_5.SiUnit
                


              click bioimageio.spec.model.v0_5.SiUnit href "" "bioimageio.spec.model.v0_5.SiUnit"
              click bioimageio.spec._internal.validated_string.ValidatedString href "" "bioimageio.spec._internal.validated_string.ValidatedString"
            

An SI unit

Methods:

Name	Description
__get_pydantic_core_schema__
__get_pydantic_json_schema__
__new__

Attributes:

Name	Type	Description
root_model	type[RootModel[Any]]	the pydantic root model to validate the string

root_model class-attribute

root_model: type[RootModel[Any]] = RootModel[
    Annotated[
        str,
        StringConstraints(
            min_length=1, pattern=SI_UNIT_REGEX
        ),
        BeforeValidator(_normalize_multiplication),
    ]
]

the pydantic root model to validate the string

__get_pydantic_core_schema__ classmethod

__get_pydantic_core_schema__(
    source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_core_schema__(
    cls, source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema:
    return no_info_after_validator_function(cls, handler(str))

__get_pydantic_json_schema__ classmethod

__get_pydantic_json_schema__(
    core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_json_schema__(
    cls, core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue:
    json_schema = cls.root_model.model_json_schema(mode=handler.mode)
    json_schema["title"] = cls.__name__.strip("_")
    if cls.__doc__:
        json_schema["description"] = cls.__doc__

    return json_schema

__new__

__new__(object: object)

Source code in src/bioimageio/spec/_internal/validated_string.py

def __new__(cls, object: object):
    _validated = cls.root_model.model_validate(object).root
    self = super().__new__(cls, _validated)
    self._validated = _validated
    return self._after_validator()

SigmoidDescr pydantic-model

Bases: NodeWithExplicitlySetFields

The logistic sigmoid function, a.k.a. expit function.

Examples:

• in YAML

  postprocessing:
    - id: sigmoid
  

• in Python:

  >>> postprocessing = [SigmoidDescr()]
  

Show JSON schema:

{
  "additionalProperties": false,
  "description": "The logistic sigmoid function, a.k.a. expit function.\n\nExamples:\n- in YAML\n    ```yaml\n    postprocessing:\n      - id: sigmoid\n    ```\n- in Python:\n\n    >>> postprocessing = [SigmoidDescr()]",
  "properties": {
    "id": {
      "const": "sigmoid",
      "title": "Id",
      "type": "string"
    }
  },
  "required": [
    "id"
  ],
  "title": "model.v0_5.SigmoidDescr",
  "type": "object"
}

Fields:

• id (Literal['sigmoid'])

id pydantic-field

id: Literal['sigmoid'] = 'sigmoid'

implemented_id class-attribute

implemented_id: Literal['sigmoid'] = 'sigmoid'

kwargs property

kwargs: KwargsNode

empty kwargs

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

SizeReference pydantic-model

Bases: Node

A tensor axis size (extent in pixels/frames) defined in relation to a reference axis.

axis.size = reference.size * reference.scale / axis.scale + offset

Note: 1. The axis and the referenced axis need to have the same unit (or no unit). 2. Batch axes may not be referenced. 3. Fractions are rounded down. 4. If the reference axis is concatenable the referencing axis is assumed to be concatenable as well with the same block order.

Example: An unisotropic input image of wh=10049 pixels depicts a phsical space of 200196mm². Let's assume that we want to express the image height h in relation to its width w instead of only accepting input images of exactly 10049 pixels (for example to express a range of valid image shapes by parametrizing w, see ParameterizedSize).

>>> w = SpaceInputAxis(id=AxisId("w"), size=100, unit="millimeter", scale=2)
>>> h = SpaceInputAxis(
...     id=AxisId("h"),
...     size=SizeReference(tensor_id=TensorId("input"), axis_id=AxisId("w"), offset=-1),
...     unit="millimeter",
...     scale=4,
... )
>>> print(h.size.get_size(h, w))
49

⇒ h = w * w.scale / h.scale + offset = 100 * 2mm / 4mm - 1 = 49

Show JSON schema:

{
  "additionalProperties": false,
  "description": "A tensor axis size (extent in pixels/frames) defined in relation to a reference axis.\n\n`axis.size = reference.size * reference.scale / axis.scale + offset`\n\nNote:\n1. The axis and the referenced axis need to have the same unit (or no unit).\n2. Batch axes may not be referenced.\n3. Fractions are rounded down.\n4. If the reference axis is `concatenable` the referencing axis is assumed to be\n    `concatenable` as well with the same block order.\n\nExample:\nAn unisotropic input image of w*h=100*49 pixels depicts a phsical space of 200*196mm\u00b2.\nLet's assume that we want to express the image height h in relation to its width w\ninstead of only accepting input images of exactly 100*49 pixels\n(for example to express a range of valid image shapes by parametrizing w, see `ParameterizedSize`).\n\n>>> w = SpaceInputAxis(id=AxisId(\"w\"), size=100, unit=\"millimeter\", scale=2)\n>>> h = SpaceInputAxis(\n...     id=AxisId(\"h\"),\n...     size=SizeReference(tensor_id=TensorId(\"input\"), axis_id=AxisId(\"w\"), offset=-1),\n...     unit=\"millimeter\",\n...     scale=4,\n... )\n>>> print(h.size.get_size(h, w))\n49\n\n\u21d2 h = w * w.scale / h.scale + offset = 100 * 2mm / 4mm - 1 = 49",
  "properties": {
    "tensor_id": {
      "description": "tensor id of the reference axis",
      "maxLength": 32,
      "minLength": 1,
      "title": "TensorId",
      "type": "string"
    },
    "axis_id": {
      "description": "axis id of the reference axis",
      "maxLength": 16,
      "minLength": 1,
      "title": "AxisId",
      "type": "string"
    },
    "offset": {
      "default": 0,
      "title": "Offset",
      "type": "integer"
    }
  },
  "required": [
    "tensor_id",
    "axis_id"
  ],
  "title": "model.v0_5.SizeReference",
  "type": "object"
}

Fields:

• tensor_id (TensorId)
• axis_id (AxisId)
• offset (StrictInt)

axis_id pydantic-field

axis_id: AxisId

axis id of the reference axis

offset pydantic-field

offset: StrictInt = 0

tensor_id pydantic-field

tensor_id: TensorId

tensor id of the reference axis

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get_size

get_size(
    axis: ChannelAxis
    | IndexInputAxis
    | IndexOutputAxis
    | TimeInputAxis
    | SpaceInputAxis
    | TimeOutputAxis
    | TimeOutputAxisWithHalo
    | SpaceOutputAxis
    | SpaceOutputAxisWithHalo,
    ref_axis: ChannelAxis
    | IndexInputAxis
    | IndexOutputAxis
    | TimeInputAxis
    | SpaceInputAxis
    | TimeOutputAxis
    | TimeOutputAxisWithHalo
    | SpaceOutputAxis
    | SpaceOutputAxisWithHalo,
    n: ParameterizedSize_N = 0,
    ref_size: int | None = None,
)

Compute the concrete size for a given axis and its reference axis.

Parameters:

Name	Type	Description	Default
axis	ChannelAxis | IndexInputAxis | IndexOutputAxis | TimeInputAxis | SpaceInputAxis | TimeOutputAxis | TimeOutputAxisWithHalo | SpaceOutputAxis | SpaceOutputAxisWithHalo	The axis this SizeReference is the size of.	required
ref_axis	ChannelAxis | IndexInputAxis | IndexOutputAxis | TimeInputAxis | SpaceInputAxis | TimeOutputAxis | TimeOutputAxisWithHalo | SpaceOutputAxis | SpaceOutputAxisWithHalo	The reference axis to compute the size from.	required
n	ParameterizedSize_N	If the ref_axis is parameterized (of type ParameterizedSize) and no fixed ref_size is given, n is used to compute the size of the parameterized ref_axis.	0
ref_size	int | None	Overwrite the reference size instead of deriving it from ref_axis (ref_axis.scale is still used; any given n is ignored).	None

Source code in src/bioimageio/spec/model/v0_5.py

def get_size(
    self,
    axis: ChannelAxis
    | IndexInputAxis
    | IndexOutputAxis
    | TimeInputAxis
    | SpaceInputAxis
    | TimeOutputAxis
    | TimeOutputAxisWithHalo
    | SpaceOutputAxis
    | SpaceOutputAxisWithHalo,
    ref_axis: ChannelAxis
    | IndexInputAxis
    | IndexOutputAxis
    | TimeInputAxis
    | SpaceInputAxis
    | TimeOutputAxis
    | TimeOutputAxisWithHalo
    | SpaceOutputAxis
    | SpaceOutputAxisWithHalo,
    n: ParameterizedSize_N = 0,
    ref_size: int | None = None,
):
    """Compute the concrete size for a given axis and its reference axis.

    Args:
        axis: The axis this [SizeReference][] is the size of.
        ref_axis: The reference axis to compute the size from.
        n: If the **ref_axis** is parameterized (of type `ParameterizedSize`)
            and no fixed **ref_size** is given,
            **n** is used to compute the size of the parameterized **ref_axis**.
        ref_size: Overwrite the reference size instead of deriving it from
            **ref_axis**
            (**ref_axis.scale** is still used; any given **n** is ignored).
    """
    assert axis.size == self, (
        "Given `axis.size` is not defined by this `SizeReference`"
    )

    assert ref_axis.id == self.axis_id, (
        f"Expected `ref_axis.id` to be {self.axis_id}, but got {ref_axis.id}."
    )

    assert axis.unit == ref_axis.unit, (
        "`SizeReference` requires `axis` and `ref_axis` to have the same `unit`,"
        f" but {axis.unit}!={ref_axis.unit}"
    )
    if ref_size is None:
        if isinstance(ref_axis.size, (int, float)):
            ref_size = ref_axis.size
        elif isinstance(ref_axis.size, ParameterizedSize):
            ref_size = ref_axis.size.get_size(n)
        elif isinstance(ref_axis.size, DataDependentSize):
            raise ValueError(
                "Reference axis referenced in `SizeReference` may not be a `DataDependentSize`."
            )
        elif isinstance(ref_axis.size, SizeReference):
            raise ValueError(
                "Reference axis referenced in `SizeReference` may not be sized by a"
                + " `SizeReference` itself."
            )
        else:
            assert_never(ref_axis.size)

    return int(ref_size * ref_axis.scale / axis.scale + self.offset)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

SoftmaxDescr pydantic-model

Bases: NodeWithExplicitlySetFields

The softmax function.

Examples:

• in YAML

  postprocessing:
    - id: softmax
      kwargs:
        axis: channel
  

• in Python:

  >>> postprocessing = [SoftmaxDescr(kwargs=SoftmaxKwargs(axis=AxisId("channel")))]
  

Show JSON schema:

{
  "$defs": {
    "SoftmaxKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [SoftmaxDescr][]",
      "properties": {
        "axis": {
          "default": "channel",
          "description": "The axis to apply the softmax function along.\nNote:\n    Defaults to 'channel' axis\n    (which may not exist, in which case\n    a different axis id has to be specified).",
          "examples": [
            "channel"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        }
      },
      "title": "model.v0_5.SoftmaxKwargs",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "description": "The softmax function.\n\nExamples:\n- in YAML\n    ```yaml\n    postprocessing:\n      - id: softmax\n        kwargs:\n          axis: channel\n    ```\n- in Python:\n\n    >>> postprocessing = [SoftmaxDescr(kwargs=SoftmaxKwargs(axis=AxisId(\"channel\")))]",
  "properties": {
    "id": {
      "const": "softmax",
      "title": "Id",
      "type": "string"
    },
    "kwargs": {
      "$ref": "#/$defs/SoftmaxKwargs"
    }
  },
  "required": [
    "id"
  ],
  "title": "model.v0_5.SoftmaxDescr",
  "type": "object"
}

Fields:

• id (Literal['softmax'])
• kwargs (SoftmaxKwargs)

id pydantic-field

id: Literal['softmax'] = 'softmax'

implemented_id class-attribute

implemented_id: Literal['softmax'] = 'softmax'

kwargs pydantic-field

kwargs: SoftmaxKwargs

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

SoftmaxKwargs pydantic-model

Bases: KwargsNode

key word arguments for SoftmaxDescr

Show JSON schema:

{
  "additionalProperties": false,
  "description": "key word arguments for [SoftmaxDescr][]",
  "properties": {
    "axis": {
      "default": "channel",
      "description": "The axis to apply the softmax function along.\nNote:\n    Defaults to 'channel' axis\n    (which may not exist, in which case\n    a different axis id has to be specified).",
      "examples": [
        "channel"
      ],
      "maxLength": 16,
      "minLength": 1,
      "title": "AxisId",
      "type": "string"
    }
  },
  "title": "model.v0_5.SoftmaxKwargs",
  "type": "object"
}

Fields:

• axis (Annotated[NonBatchAxisId, Field(examples=['channel'])])

axis pydantic-field

axis: Annotated[
    NonBatchAxisId, Field(examples=["channel"])
] = CHANNEL_AXIS_ID

The axis to apply the softmax function along. Note: Defaults to 'channel' axis (which may not exist, in which case a different axis id has to be specified).

__contains__

__contains__(item: str) -> bool

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __contains__(self, item: str) -> bool:
    return item in self.__class__.model_fields

__getitem__

__getitem__(item: str) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __getitem__(self, item: str) -> Any:
    if item in self.__class__.model_fields:
        return getattr(self, item)
    else:
        raise KeyError(item)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get

get(item: str, default: Any = None) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def get(self, item: str, default: Any = None) -> Any:
    return self[item] if item in self else default  # ruff: ignore[SIM401]

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

SpaceAxisBase pydantic-model

Bases: AxisBase

Show JSON schema:

{
  "additionalProperties": false,
  "properties": {
    "id": {
      "default": "x",
      "examples": [
        "x",
        "y",
        "z"
      ],
      "maxLength": 16,
      "minLength": 1,
      "title": "AxisId",
      "type": "string"
    },
    "description": {
      "default": "",
      "description": "A short description of this axis beyond its type and id.",
      "maxLength": 128,
      "title": "Description",
      "type": "string"
    },
    "type": {
      "const": "space",
      "title": "Type",
      "type": "string"
    },
    "unit": {
      "anyOf": [
        {
          "enum": [
            "attometer",
            "angstrom",
            "centimeter",
            "decimeter",
            "exameter",
            "femtometer",
            "foot",
            "gigameter",
            "hectometer",
            "inch",
            "kilometer",
            "megameter",
            "meter",
            "micrometer",
            "mile",
            "millimeter",
            "nanometer",
            "parsec",
            "petameter",
            "picometer",
            "terameter",
            "yard",
            "yoctometer",
            "yottameter",
            "zeptometer",
            "zettameter"
          ],
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Unit"
    },
    "scale": {
      "default": 1.0,
      "exclusiveMinimum": 0,
      "title": "Scale",
      "type": "number"
    }
  },
  "required": [
    "type"
  ],
  "title": "model.v0_5.SpaceAxisBase",
  "type": "object"
}

Fields:

• description (Annotated[str, MaxLen(128)])
• type (Literal['space'])
• id (Annotated[NonBatchAxisId, Field(examples=['x', 'y', 'z'])])
• unit (SpaceUnit | None)
• scale (Annotated[float, Gt(0)])

description pydantic-field

description: Annotated[str, MaxLen(128)] = ''

A short description of this axis beyond its type and id.

id pydantic-field

id: Annotated[
    NonBatchAxisId, Field(examples=["x", "y", "z"])
] = DEFAULT_SPACE_AXIS_ID

An axis id unique across all axes of one tensor.

implemented_type class-attribute

implemented_type: Literal['space'] = 'space'

scale pydantic-field

scale: Annotated[float, Gt(0)] = 1.0

type pydantic-field

type: Literal['space'] = 'space'

unit pydantic-field

unit: SpaceUnit | None = None

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

SpaceInputAxis pydantic-model

Bases: SpaceAxisBase, _WithInputAxisSize

Show JSON schema:

{
  "$defs": {
    "ParameterizedSize": {
      "additionalProperties": false,
      "description": "Describes a range of valid tensor axis sizes as `size = min + n*step`.\n\n- **min** and **step** are given by the model description.\n- All blocksize paramters n = 0,1,2,... yield a valid `size`.\n- A greater blocksize paramter n = 0,1,2,... results in a greater **size**.\n  This allows to adjust the axis size more generically.",
      "properties": {
        "min": {
          "exclusiveMinimum": 0,
          "title": "Min",
          "type": "integer"
        },
        "step": {
          "exclusiveMinimum": 0,
          "title": "Step",
          "type": "integer"
        }
      },
      "required": [
        "min",
        "step"
      ],
      "title": "model.v0_5.ParameterizedSize",
      "type": "object"
    },
    "SizeReference": {
      "additionalProperties": false,
      "description": "A tensor axis size (extent in pixels/frames) defined in relation to a reference axis.\n\n`axis.size = reference.size * reference.scale / axis.scale + offset`\n\nNote:\n1. The axis and the referenced axis need to have the same unit (or no unit).\n2. Batch axes may not be referenced.\n3. Fractions are rounded down.\n4. If the reference axis is `concatenable` the referencing axis is assumed to be\n    `concatenable` as well with the same block order.\n\nExample:\nAn unisotropic input image of w*h=100*49 pixels depicts a phsical space of 200*196mm\u00b2.\nLet's assume that we want to express the image height h in relation to its width w\ninstead of only accepting input images of exactly 100*49 pixels\n(for example to express a range of valid image shapes by parametrizing w, see `ParameterizedSize`).\n\n>>> w = SpaceInputAxis(id=AxisId(\"w\"), size=100, unit=\"millimeter\", scale=2)\n>>> h = SpaceInputAxis(\n...     id=AxisId(\"h\"),\n...     size=SizeReference(tensor_id=TensorId(\"input\"), axis_id=AxisId(\"w\"), offset=-1),\n...     unit=\"millimeter\",\n...     scale=4,\n... )\n>>> print(h.size.get_size(h, w))\n49\n\n\u21d2 h = w * w.scale / h.scale + offset = 100 * 2mm / 4mm - 1 = 49",
      "properties": {
        "tensor_id": {
          "description": "tensor id of the reference axis",
          "maxLength": 32,
          "minLength": 1,
          "title": "TensorId",
          "type": "string"
        },
        "axis_id": {
          "description": "axis id of the reference axis",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "offset": {
          "default": 0,
          "title": "Offset",
          "type": "integer"
        }
      },
      "required": [
        "tensor_id",
        "axis_id"
      ],
      "title": "model.v0_5.SizeReference",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "properties": {
    "size": {
      "anyOf": [
        {
          "exclusiveMinimum": 0,
          "type": "integer"
        },
        {
          "$ref": "#/$defs/ParameterizedSize"
        },
        {
          "$ref": "#/$defs/SizeReference"
        }
      ],
      "description": "The size/length of this axis can be specified as\n- fixed integer\n- parameterized series of valid sizes ([ParameterizedSize][])\n- reference to another axis with an optional offset ([SizeReference][])",
      "examples": [
        10,
        {
          "min": 32,
          "step": 16
        },
        {
          "axis_id": "a",
          "offset": 5,
          "tensor_id": "t"
        }
      ],
      "title": "Size"
    },
    "id": {
      "default": "x",
      "examples": [
        "x",
        "y",
        "z"
      ],
      "maxLength": 16,
      "minLength": 1,
      "title": "AxisId",
      "type": "string"
    },
    "description": {
      "default": "",
      "description": "A short description of this axis beyond its type and id.",
      "maxLength": 128,
      "title": "Description",
      "type": "string"
    },
    "type": {
      "const": "space",
      "title": "Type",
      "type": "string"
    },
    "unit": {
      "anyOf": [
        {
          "enum": [
            "attometer",
            "angstrom",
            "centimeter",
            "decimeter",
            "exameter",
            "femtometer",
            "foot",
            "gigameter",
            "hectometer",
            "inch",
            "kilometer",
            "megameter",
            "meter",
            "micrometer",
            "mile",
            "millimeter",
            "nanometer",
            "parsec",
            "petameter",
            "picometer",
            "terameter",
            "yard",
            "yoctometer",
            "yottameter",
            "zeptometer",
            "zettameter"
          ],
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Unit"
    },
    "scale": {
      "default": 1.0,
      "exclusiveMinimum": 0,
      "title": "Scale",
      "type": "number"
    },
    "concatenable": {
      "default": false,
      "description": "If a model has a `concatenable` input axis, it can be processed blockwise,\nsplitting a longer sample axis into blocks matching its input tensor description.\nOutput axes are concatenable if they have a [SizeReference][] to a concatenable\ninput axis.",
      "title": "Concatenable",
      "type": "boolean"
    }
  },
  "required": [
    "size",
    "type"
  ],
  "title": "model.v0_5.SpaceInputAxis",
  "type": "object"
}

Fields:

• size (Annotated[Annotated[int, Gt(0)] | ParameterizedSize | SizeReference, Field(examples=[10, ParameterizedSize(min=32, step=16).model_dump(mode='json'), {'tensor_id': 't', 'axis_id': 'a', 'offset': 5}])])
• id (Annotated[NonBatchAxisId, Field(examples=['x', 'y', 'z'])])
• description (Annotated[str, MaxLen(128)])
• type (Literal['space'])
• unit (SpaceUnit | None)
• scale (Annotated[float, Gt(0)])
• concatenable (bool)

concatenable pydantic-field

concatenable: bool = False

If a model has a concatenable input axis, it can be processed blockwise, splitting a longer sample axis into blocks matching its input tensor description. Output axes are concatenable if they have a SizeReference to a concatenable input axis.

description pydantic-field

description: Annotated[str, MaxLen(128)] = ''

A short description of this axis beyond its type and id.

id pydantic-field

id: Annotated[
    NonBatchAxisId, Field(examples=["x", "y", "z"])
] = DEFAULT_SPACE_AXIS_ID

An axis id unique across all axes of one tensor.

implemented_type class-attribute

implemented_type: Literal['space'] = 'space'

scale pydantic-field

scale: Annotated[float, Gt(0)] = 1.0

size pydantic-field

size: Annotated[
    Annotated[int, Gt(0)]
    | ParameterizedSize
    | SizeReference,
    Field(
        examples=[
            10,
            ParameterizedSize(min=32, step=16).model_dump(
                mode="json"
            ),
            {"tensor_id": "t", "axis_id": "a", "offset": 5},
        ]
    ),
]

The size/length of this axis can be specified as - fixed integer - parameterized series of valid sizes (ParameterizedSize) - reference to another axis with an optional offset (SizeReference)

type pydantic-field

type: Literal['space'] = 'space'

unit pydantic-field

unit: SpaceUnit | None = None

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

SpaceOutputAxis pydantic-model

Bases: SpaceAxisBase, _WithOutputAxisSize

Show JSON schema:

{
  "$defs": {
    "SizeReference": {
      "additionalProperties": false,
      "description": "A tensor axis size (extent in pixels/frames) defined in relation to a reference axis.\n\n`axis.size = reference.size * reference.scale / axis.scale + offset`\n\nNote:\n1. The axis and the referenced axis need to have the same unit (or no unit).\n2. Batch axes may not be referenced.\n3. Fractions are rounded down.\n4. If the reference axis is `concatenable` the referencing axis is assumed to be\n    `concatenable` as well with the same block order.\n\nExample:\nAn unisotropic input image of w*h=100*49 pixels depicts a phsical space of 200*196mm\u00b2.\nLet's assume that we want to express the image height h in relation to its width w\ninstead of only accepting input images of exactly 100*49 pixels\n(for example to express a range of valid image shapes by parametrizing w, see `ParameterizedSize`).\n\n>>> w = SpaceInputAxis(id=AxisId(\"w\"), size=100, unit=\"millimeter\", scale=2)\n>>> h = SpaceInputAxis(\n...     id=AxisId(\"h\"),\n...     size=SizeReference(tensor_id=TensorId(\"input\"), axis_id=AxisId(\"w\"), offset=-1),\n...     unit=\"millimeter\",\n...     scale=4,\n... )\n>>> print(h.size.get_size(h, w))\n49\n\n\u21d2 h = w * w.scale / h.scale + offset = 100 * 2mm / 4mm - 1 = 49",
      "properties": {
        "tensor_id": {
          "description": "tensor id of the reference axis",
          "maxLength": 32,
          "minLength": 1,
          "title": "TensorId",
          "type": "string"
        },
        "axis_id": {
          "description": "axis id of the reference axis",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "offset": {
          "default": 0,
          "title": "Offset",
          "type": "integer"
        }
      },
      "required": [
        "tensor_id",
        "axis_id"
      ],
      "title": "model.v0_5.SizeReference",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "properties": {
    "size": {
      "anyOf": [
        {
          "exclusiveMinimum": 0,
          "type": "integer"
        },
        {
          "$ref": "#/$defs/SizeReference"
        }
      ],
      "description": "The size/length of this axis can be specified as\n- fixed integer\n- reference to another axis with an optional offset (see [SizeReference][])",
      "examples": [
        10,
        {
          "axis_id": "a",
          "offset": 5,
          "tensor_id": "t"
        }
      ],
      "title": "Size"
    },
    "id": {
      "default": "x",
      "examples": [
        "x",
        "y",
        "z"
      ],
      "maxLength": 16,
      "minLength": 1,
      "title": "AxisId",
      "type": "string"
    },
    "description": {
      "default": "",
      "description": "A short description of this axis beyond its type and id.",
      "maxLength": 128,
      "title": "Description",
      "type": "string"
    },
    "type": {
      "const": "space",
      "title": "Type",
      "type": "string"
    },
    "unit": {
      "anyOf": [
        {
          "enum": [
            "attometer",
            "angstrom",
            "centimeter",
            "decimeter",
            "exameter",
            "femtometer",
            "foot",
            "gigameter",
            "hectometer",
            "inch",
            "kilometer",
            "megameter",
            "meter",
            "micrometer",
            "mile",
            "millimeter",
            "nanometer",
            "parsec",
            "petameter",
            "picometer",
            "terameter",
            "yard",
            "yoctometer",
            "yottameter",
            "zeptometer",
            "zettameter"
          ],
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Unit"
    },
    "scale": {
      "default": 1.0,
      "exclusiveMinimum": 0,
      "title": "Scale",
      "type": "number"
    }
  },
  "required": [
    "size",
    "type"
  ],
  "title": "model.v0_5.SpaceOutputAxis",
  "type": "object"
}

Fields:

• size (Annotated[Annotated[int, Gt(0)] | SizeReference, Field(examples=[10, {'tensor_id': 't', 'axis_id': 'a', 'offset': 5}])])
• id (Annotated[NonBatchAxisId, Field(examples=['x', 'y', 'z'])])
• description (Annotated[str, MaxLen(128)])
• type (Literal['space'])
• unit (SpaceUnit | None)
• scale (Annotated[float, Gt(0)])

description pydantic-field

description: Annotated[str, MaxLen(128)] = ''

A short description of this axis beyond its type and id.

id pydantic-field

id: Annotated[
    NonBatchAxisId, Field(examples=["x", "y", "z"])
] = DEFAULT_SPACE_AXIS_ID

An axis id unique across all axes of one tensor.

implemented_type class-attribute

implemented_type: Literal['space'] = 'space'

scale pydantic-field

scale: Annotated[float, Gt(0)] = 1.0

size pydantic-field

size: Annotated[
    Annotated[int, Gt(0)] | SizeReference,
    Field(
        examples=[
            10,
            {"tensor_id": "t", "axis_id": "a", "offset": 5},
        ]
    ),
]

The size/length of this axis can be specified as - fixed integer - reference to another axis with an optional offset (see SizeReference)

type pydantic-field

type: Literal['space'] = 'space'

unit pydantic-field

unit: SpaceUnit | None = None

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

SpaceOutputAxisWithHalo pydantic-model

Bases: SpaceAxisBase, WithHalo

Show JSON schema:

{
  "$defs": {
    "SizeReference": {
      "additionalProperties": false,
      "description": "A tensor axis size (extent in pixels/frames) defined in relation to a reference axis.\n\n`axis.size = reference.size * reference.scale / axis.scale + offset`\n\nNote:\n1. The axis and the referenced axis need to have the same unit (or no unit).\n2. Batch axes may not be referenced.\n3. Fractions are rounded down.\n4. If the reference axis is `concatenable` the referencing axis is assumed to be\n    `concatenable` as well with the same block order.\n\nExample:\nAn unisotropic input image of w*h=100*49 pixels depicts a phsical space of 200*196mm\u00b2.\nLet's assume that we want to express the image height h in relation to its width w\ninstead of only accepting input images of exactly 100*49 pixels\n(for example to express a range of valid image shapes by parametrizing w, see `ParameterizedSize`).\n\n>>> w = SpaceInputAxis(id=AxisId(\"w\"), size=100, unit=\"millimeter\", scale=2)\n>>> h = SpaceInputAxis(\n...     id=AxisId(\"h\"),\n...     size=SizeReference(tensor_id=TensorId(\"input\"), axis_id=AxisId(\"w\"), offset=-1),\n...     unit=\"millimeter\",\n...     scale=4,\n... )\n>>> print(h.size.get_size(h, w))\n49\n\n\u21d2 h = w * w.scale / h.scale + offset = 100 * 2mm / 4mm - 1 = 49",
      "properties": {
        "tensor_id": {
          "description": "tensor id of the reference axis",
          "maxLength": 32,
          "minLength": 1,
          "title": "TensorId",
          "type": "string"
        },
        "axis_id": {
          "description": "axis id of the reference axis",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "offset": {
          "default": 0,
          "title": "Offset",
          "type": "integer"
        }
      },
      "required": [
        "tensor_id",
        "axis_id"
      ],
      "title": "model.v0_5.SizeReference",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "properties": {
    "halo": {
      "description": "The halo should be cropped from the output tensor to avoid boundary effects.\nIt is to be cropped from both sides, i.e. `size_after_crop = size - 2 * halo`.\nTo document a halo that is already cropped by the model use `size.offset` instead.",
      "minimum": 1,
      "title": "Halo",
      "type": "integer"
    },
    "size": {
      "$ref": "#/$defs/SizeReference",
      "description": "reference to another axis with an optional offset (see [SizeReference][])",
      "examples": [
        {
          "axis_id": "a",
          "offset": 5,
          "tensor_id": "t"
        }
      ]
    },
    "id": {
      "default": "x",
      "examples": [
        "x",
        "y",
        "z"
      ],
      "maxLength": 16,
      "minLength": 1,
      "title": "AxisId",
      "type": "string"
    },
    "description": {
      "default": "",
      "description": "A short description of this axis beyond its type and id.",
      "maxLength": 128,
      "title": "Description",
      "type": "string"
    },
    "type": {
      "const": "space",
      "title": "Type",
      "type": "string"
    },
    "unit": {
      "anyOf": [
        {
          "enum": [
            "attometer",
            "angstrom",
            "centimeter",
            "decimeter",
            "exameter",
            "femtometer",
            "foot",
            "gigameter",
            "hectometer",
            "inch",
            "kilometer",
            "megameter",
            "meter",
            "micrometer",
            "mile",
            "millimeter",
            "nanometer",
            "parsec",
            "petameter",
            "picometer",
            "terameter",
            "yard",
            "yoctometer",
            "yottameter",
            "zeptometer",
            "zettameter"
          ],
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Unit"
    },
    "scale": {
      "default": 1.0,
      "exclusiveMinimum": 0,
      "title": "Scale",
      "type": "number"
    }
  },
  "required": [
    "halo",
    "size",
    "type"
  ],
  "title": "model.v0_5.SpaceOutputAxisWithHalo",
  "type": "object"
}

Fields:

• halo (Annotated[int, Ge(1)])
• size (Annotated[SizeReference, Field(examples=[{'tensor_id': 't', 'axis_id': 'a', 'offset': 5}])])
• id (Annotated[NonBatchAxisId, Field(examples=['x', 'y', 'z'])])
• description (Annotated[str, MaxLen(128)])
• type (Literal['space'])
• unit (SpaceUnit | None)
• scale (Annotated[float, Gt(0)])

description pydantic-field

description: Annotated[str, MaxLen(128)] = ''

A short description of this axis beyond its type and id.

halo pydantic-field

halo: Annotated[int, Ge(1)]

The halo should be cropped from the output tensor to avoid boundary effects. It is to be cropped from both sides, i.e. size_after_crop = size - 2 * halo. To document a halo that is already cropped by the model use size.offset instead.

id pydantic-field

id: Annotated[
    NonBatchAxisId, Field(examples=["x", "y", "z"])
] = DEFAULT_SPACE_AXIS_ID

An axis id unique across all axes of one tensor.

implemented_type class-attribute

implemented_type: Literal['space'] = 'space'

scale pydantic-field

scale: Annotated[float, Gt(0)] = 1.0

size pydantic-field

size: Annotated[
    SizeReference,
    Field(
        examples=[
            {"tensor_id": "t", "axis_id": "a", "offset": 5}
        ]
    ),
]

reference to another axis with an optional offset (see SizeReference)

type pydantic-field

type: Literal['space'] = 'space'

unit pydantic-field

unit: SpaceUnit | None = None

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

StardistPostprocessingDescr pydantic-model

Bases: NodeWithExplicitlySetFields

Stardist postprocessing including non-maximum suppression and converting polygon representations to instance labels

as described in: - Uwe Schmidt, Martin Weigert, Coleman Broaddus, and Gene Myers. Cell Detection with Star-convex Polygons. International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI), Granada, Spain, September 2018. - Martin Weigert, Uwe Schmidt, Robert Haase, Ko Sugawara, and Gene Myers. Star-convex Polyhedra for 3D Object Detection and Segmentation in Microscopy. The IEEE Winter Conference on Applications of Computer Vision (WACV), Snowmass Village, Colorado, March 2020.

Note: Only available if the stardist package is installed.

Show JSON schema:

{
  "$defs": {
    "StardistPostprocessingKwargs2D": {
      "additionalProperties": false,
      "properties": {
        "prob_threshold": {
          "description": "The probability threshold for object candidate selection.",
          "title": "Prob Threshold",
          "type": "number"
        },
        "nms_threshold": {
          "description": "The IoU threshold for non-maximum suppression.",
          "title": "Nms Threshold",
          "type": "number"
        },
        "n_rays": {
          "description": "Number of radial lines (rays) cast from the center of an object to its boundary.",
          "title": "N Rays",
          "type": "integer"
        },
        "grid": {
          "description": "Grid size of network predictions.",
          "maxItems": 2,
          "minItems": 2,
          "prefixItems": [
            {
              "type": "integer"
            },
            {
              "type": "integer"
            }
          ],
          "title": "Grid",
          "type": "array"
        },
        "b": {
          "anyOf": [
            {
              "type": "integer"
            },
            {
              "maxItems": 2,
              "minItems": 2,
              "prefixItems": [
                {
                  "maxItems": 2,
                  "minItems": 2,
                  "prefixItems": [
                    {
                      "type": "integer"
                    },
                    {
                      "type": "integer"
                    }
                  ],
                  "type": "array"
                },
                {
                  "maxItems": 2,
                  "minItems": 2,
                  "prefixItems": [
                    {
                      "type": "integer"
                    },
                    {
                      "type": "integer"
                    }
                  ],
                  "type": "array"
                }
              ],
              "type": "array"
            }
          ],
          "description": "Border region in which object probability is set to zero.",
          "title": "B"
        }
      },
      "required": [
        "prob_threshold",
        "nms_threshold",
        "n_rays",
        "grid",
        "b"
      ],
      "title": "model.v0_5.StardistPostprocessingKwargs2D",
      "type": "object"
    },
    "StardistPostprocessingKwargs3D": {
      "additionalProperties": false,
      "properties": {
        "prob_threshold": {
          "description": "The probability threshold for object candidate selection.",
          "title": "Prob Threshold",
          "type": "number"
        },
        "nms_threshold": {
          "description": "The IoU threshold for non-maximum suppression.",
          "title": "Nms Threshold",
          "type": "number"
        },
        "n_rays": {
          "description": "Number of radial lines (rays) cast from the center of an object to its boundary.",
          "title": "N Rays",
          "type": "integer"
        },
        "grid": {
          "description": "Grid size of network predictions.",
          "maxItems": 3,
          "minItems": 3,
          "prefixItems": [
            {
              "type": "integer"
            },
            {
              "type": "integer"
            },
            {
              "type": "integer"
            }
          ],
          "title": "Grid",
          "type": "array"
        },
        "b": {
          "anyOf": [
            {
              "type": "integer"
            },
            {
              "maxItems": 3,
              "minItems": 3,
              "prefixItems": [
                {
                  "maxItems": 2,
                  "minItems": 2,
                  "prefixItems": [
                    {
                      "type": "integer"
                    },
                    {
                      "type": "integer"
                    }
                  ],
                  "type": "array"
                },
                {
                  "maxItems": 2,
                  "minItems": 2,
                  "prefixItems": [
                    {
                      "type": "integer"
                    },
                    {
                      "type": "integer"
                    }
                  ],
                  "type": "array"
                },
                {
                  "maxItems": 2,
                  "minItems": 2,
                  "prefixItems": [
                    {
                      "type": "integer"
                    },
                    {
                      "type": "integer"
                    }
                  ],
                  "type": "array"
                }
              ],
              "type": "array"
            }
          ],
          "description": "Border region in which object probability is set to zero.",
          "title": "B"
        },
        "anisotropy": {
          "description": "Anisotropy factors for 3D star-convex polyhedra, i.e. the physical pixel size along each spatial axis.",
          "maxItems": 3,
          "minItems": 3,
          "prefixItems": [
            {
              "type": "number"
            },
            {
              "type": "number"
            },
            {
              "type": "number"
            }
          ],
          "title": "Anisotropy",
          "type": "array"
        },
        "overlap_label": {
          "anyOf": [
            {
              "type": "integer"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Optional label to apply to any area of overlapping predicted objects.",
          "title": "Overlap Label"
        }
      },
      "required": [
        "prob_threshold",
        "nms_threshold",
        "n_rays",
        "grid",
        "b",
        "anisotropy"
      ],
      "title": "model.v0_5.StardistPostprocessingKwargs3D",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "description": "Stardist postprocessing including non-maximum suppression and converting polygon representations to instance labels\n\nas described in:\n- Uwe Schmidt, Martin Weigert, Coleman Broaddus, and Gene Myers.\n[*Cell Detection with Star-convex Polygons*](https://arxiv.org/abs/1806.03535).\nInternational Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI), Granada, Spain, September 2018.\n- Martin Weigert, Uwe Schmidt, Robert Haase, Ko Sugawara, and Gene Myers.\n[*Star-convex Polyhedra for 3D Object Detection and Segmentation in Microscopy*](http://openaccess.thecvf.com/content_WACV_2020/papers/Weigert_Star-convex_Polyhedra_for_3D_Object_Detection_and_Segmentation_in_Microscopy_WACV_2020_paper.pdf).\nThe IEEE Winter Conference on Applications of Computer Vision (WACV), Snowmass Village, Colorado, March 2020.\n\nNote: Only available if the `stardist` package is installed.",
  "properties": {
    "id": {
      "const": "stardist_postprocessing",
      "title": "Id",
      "type": "string"
    },
    "kwargs": {
      "anyOf": [
        {
          "$ref": "#/$defs/StardistPostprocessingKwargs2D"
        },
        {
          "$ref": "#/$defs/StardistPostprocessingKwargs3D"
        }
      ],
      "title": "Kwargs"
    }
  },
  "required": [
    "id",
    "kwargs"
  ],
  "title": "model.v0_5.StardistPostprocessingDescr",
  "type": "object"
}

Fields:

• id (Literal['stardist_postprocessing'])
• kwargs (StardistPostprocessingKwargs2D | StardistPostprocessingKwargs3D)

id pydantic-field

id: Literal["stardist_postprocessing"] = (
    "stardist_postprocessing"
)

implemented_id class-attribute

implemented_id: Literal["stardist_postprocessing"] = (
    "stardist_postprocessing"
)

kwargs pydantic-field

kwargs: (
    StardistPostprocessingKwargs2D
    | StardistPostprocessingKwargs3D
)

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

StardistPostprocessingKwargs2D pydantic-model

Bases: _StardistPostprocessingKwargsBase

Show JSON schema:

{
  "additionalProperties": false,
  "properties": {
    "prob_threshold": {
      "description": "The probability threshold for object candidate selection.",
      "title": "Prob Threshold",
      "type": "number"
    },
    "nms_threshold": {
      "description": "The IoU threshold for non-maximum suppression.",
      "title": "Nms Threshold",
      "type": "number"
    },
    "n_rays": {
      "description": "Number of radial lines (rays) cast from the center of an object to its boundary.",
      "title": "N Rays",
      "type": "integer"
    },
    "grid": {
      "description": "Grid size of network predictions.",
      "maxItems": 2,
      "minItems": 2,
      "prefixItems": [
        {
          "type": "integer"
        },
        {
          "type": "integer"
        }
      ],
      "title": "Grid",
      "type": "array"
    },
    "b": {
      "anyOf": [
        {
          "type": "integer"
        },
        {
          "maxItems": 2,
          "minItems": 2,
          "prefixItems": [
            {
              "maxItems": 2,
              "minItems": 2,
              "prefixItems": [
                {
                  "type": "integer"
                },
                {
                  "type": "integer"
                }
              ],
              "type": "array"
            },
            {
              "maxItems": 2,
              "minItems": 2,
              "prefixItems": [
                {
                  "type": "integer"
                },
                {
                  "type": "integer"
                }
              ],
              "type": "array"
            }
          ],
          "type": "array"
        }
      ],
      "description": "Border region in which object probability is set to zero.",
      "title": "B"
    }
  },
  "required": [
    "prob_threshold",
    "nms_threshold",
    "n_rays",
    "grid",
    "b"
  ],
  "title": "model.v0_5.StardistPostprocessingKwargs2D",
  "type": "object"
}

Fields:

• prob_threshold (float)
• nms_threshold (float)
• n_rays (int)
• grid (tuple[int, int])
• b (int | tuple[tuple[int, int], tuple[int, int]])

b pydantic-field

b: int | tuple[tuple[int, int], tuple[int, int]]

Border region in which object probability is set to zero.

grid pydantic-field

grid: tuple[int, int]

Grid size of network predictions.

n_rays pydantic-field

n_rays: int

Number of radial lines (rays) cast from the center of an object to its boundary.

nms_threshold pydantic-field

nms_threshold: float

The IoU threshold for non-maximum suppression.

prob_threshold pydantic-field

prob_threshold: float

The probability threshold for object candidate selection.

__contains__

__contains__(item: str) -> bool

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __contains__(self, item: str) -> bool:
    return item in self.__class__.model_fields

__getitem__

__getitem__(item: str) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __getitem__(self, item: str) -> Any:
    if item in self.__class__.model_fields:
        return getattr(self, item)
    else:
        raise KeyError(item)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get

get(item: str, default: Any = None) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def get(self, item: str, default: Any = None) -> Any:
    return self[item] if item in self else default  # ruff: ignore[SIM401]

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

StardistPostprocessingKwargs3D pydantic-model

Bases: _StardistPostprocessingKwargsBase

Show JSON schema:

{
  "additionalProperties": false,
  "properties": {
    "prob_threshold": {
      "description": "The probability threshold for object candidate selection.",
      "title": "Prob Threshold",
      "type": "number"
    },
    "nms_threshold": {
      "description": "The IoU threshold for non-maximum suppression.",
      "title": "Nms Threshold",
      "type": "number"
    },
    "n_rays": {
      "description": "Number of radial lines (rays) cast from the center of an object to its boundary.",
      "title": "N Rays",
      "type": "integer"
    },
    "grid": {
      "description": "Grid size of network predictions.",
      "maxItems": 3,
      "minItems": 3,
      "prefixItems": [
        {
          "type": "integer"
        },
        {
          "type": "integer"
        },
        {
          "type": "integer"
        }
      ],
      "title": "Grid",
      "type": "array"
    },
    "b": {
      "anyOf": [
        {
          "type": "integer"
        },
        {
          "maxItems": 3,
          "minItems": 3,
          "prefixItems": [
            {
              "maxItems": 2,
              "minItems": 2,
              "prefixItems": [
                {
                  "type": "integer"
                },
                {
                  "type": "integer"
                }
              ],
              "type": "array"
            },
            {
              "maxItems": 2,
              "minItems": 2,
              "prefixItems": [
                {
                  "type": "integer"
                },
                {
                  "type": "integer"
                }
              ],
              "type": "array"
            },
            {
              "maxItems": 2,
              "minItems": 2,
              "prefixItems": [
                {
                  "type": "integer"
                },
                {
                  "type": "integer"
                }
              ],
              "type": "array"
            }
          ],
          "type": "array"
        }
      ],
      "description": "Border region in which object probability is set to zero.",
      "title": "B"
    },
    "anisotropy": {
      "description": "Anisotropy factors for 3D star-convex polyhedra, i.e. the physical pixel size along each spatial axis.",
      "maxItems": 3,
      "minItems": 3,
      "prefixItems": [
        {
          "type": "number"
        },
        {
          "type": "number"
        },
        {
          "type": "number"
        }
      ],
      "title": "Anisotropy",
      "type": "array"
    },
    "overlap_label": {
      "anyOf": [
        {
          "type": "integer"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Optional label to apply to any area of overlapping predicted objects.",
      "title": "Overlap Label"
    }
  },
  "required": [
    "prob_threshold",
    "nms_threshold",
    "n_rays",
    "grid",
    "b",
    "anisotropy"
  ],
  "title": "model.v0_5.StardistPostprocessingKwargs3D",
  "type": "object"
}

Fields:

• prob_threshold (float)
• nms_threshold (float)
• n_rays (int)
• grid (tuple[int, int, int])
• b (int | tuple[tuple[int, int], tuple[int, int], tuple[int, int]])
• anisotropy (tuple[float, float, float])
• overlap_label (int | None)

anisotropy pydantic-field

anisotropy: tuple[float, float, float]

Anisotropy factors for 3D star-convex polyhedra, i.e. the physical pixel size along each spatial axis.

b pydantic-field

b: (
    int
    | tuple[
        tuple[int, int], tuple[int, int], tuple[int, int]
    ]
)

Border region in which object probability is set to zero.

grid pydantic-field

grid: tuple[int, int, int]

Grid size of network predictions.

n_rays pydantic-field

n_rays: int

Number of radial lines (rays) cast from the center of an object to its boundary.

nms_threshold pydantic-field

nms_threshold: float

The IoU threshold for non-maximum suppression.

overlap_label pydantic-field

overlap_label: int | None = None

Optional label to apply to any area of overlapping predicted objects.

prob_threshold pydantic-field

prob_threshold: float

The probability threshold for object candidate selection.

__contains__

__contains__(item: str) -> bool

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __contains__(self, item: str) -> bool:
    return item in self.__class__.model_fields

__getitem__

__getitem__(item: str) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __getitem__(self, item: str) -> Any:
    if item in self.__class__.model_fields:
        return getattr(self, item)
    else:
        raise KeyError(item)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get

get(item: str, default: Any = None) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def get(self, item: str, default: Any = None) -> Any:
    return self[item] if item in self else default  # ruff: ignore[SIM401]

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

SymmetricPadding pydantic-model

Bases: Node

Show JSON schema:

{
  "additionalProperties": false,
  "properties": {
    "mode": {
      "const": "symmetric",
      "default": "symmetric",
      "title": "Mode",
      "type": "string"
    }
  },
  "title": "model.v0_5.SymmetricPadding",
  "type": "object"
}

Fields:

• mode (Literal['symmetric'])

mode pydantic-field

mode: Literal['symmetric'] = 'symmetric'

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

TensorDescrBase pydantic-model

Bases: Node, Generic[IO_AxisT]

Show JSON schema:

{
  "$defs": {
    "BatchAxis": {
      "additionalProperties": false,
      "properties": {
        "id": {
          "default": "batch",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "batch",
          "title": "Type",
          "type": "string"
        },
        "size": {
          "anyOf": [
            {
              "const": 1,
              "type": "integer"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The batch size may be fixed to 1,\notherwise (the default) it may be chosen arbitrarily depending on available memory",
          "title": "Size"
        }
      },
      "required": [
        "type"
      ],
      "title": "model.v0_5.BatchAxis",
      "type": "object"
    },
    "ChannelAxis": {
      "additionalProperties": false,
      "properties": {
        "id": {
          "default": "channel",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "channel",
          "title": "Type",
          "type": "string"
        },
        "channel_names": {
          "description": "Name/label for each channel. The number of channels is given by `len(channel_names)`.",
          "items": {
            "type": "string"
          },
          "minItems": 1,
          "title": "Channel Names",
          "type": "array"
        }
      },
      "required": [
        "type",
        "channel_names"
      ],
      "title": "model.v0_5.ChannelAxis",
      "type": "object"
    },
    "DataDependentSize": {
      "additionalProperties": false,
      "properties": {
        "min": {
          "default": 1,
          "exclusiveMinimum": 0,
          "title": "Min",
          "type": "integer"
        },
        "max": {
          "anyOf": [
            {
              "exclusiveMinimum": 1,
              "type": "integer"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Max"
        }
      },
      "title": "model.v0_5.DataDependentSize",
      "type": "object"
    },
    "FileDescr": {
      "additionalProperties": false,
      "description": "A file description",
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "File source",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        }
      },
      "required": [
        "source"
      ],
      "title": "_internal.io.FileDescr",
      "type": "object"
    },
    "IndexInputAxis": {
      "additionalProperties": false,
      "properties": {
        "size": {
          "anyOf": [
            {
              "exclusiveMinimum": 0,
              "type": "integer"
            },
            {
              "$ref": "#/$defs/ParameterizedSize"
            },
            {
              "$ref": "#/$defs/SizeReference"
            }
          ],
          "description": "The size/length of this axis can be specified as\n- fixed integer\n- parameterized series of valid sizes ([ParameterizedSize][])\n- reference to another axis with an optional offset ([SizeReference][])",
          "examples": [
            10,
            {
              "min": 32,
              "step": 16
            },
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ],
          "title": "Size"
        },
        "id": {
          "default": "index",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "index",
          "title": "Type",
          "type": "string"
        },
        "concatenable": {
          "default": false,
          "description": "If a model has a `concatenable` input axis, it can be processed blockwise,\nsplitting a longer sample axis into blocks matching its input tensor description.\nOutput axes are concatenable if they have a [SizeReference][] to a concatenable\ninput axis.",
          "title": "Concatenable",
          "type": "boolean"
        }
      },
      "required": [
        "size",
        "type"
      ],
      "title": "model.v0_5.IndexInputAxis",
      "type": "object"
    },
    "IndexOutputAxis": {
      "additionalProperties": false,
      "properties": {
        "id": {
          "default": "index",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "index",
          "title": "Type",
          "type": "string"
        },
        "size": {
          "anyOf": [
            {
              "exclusiveMinimum": 0,
              "type": "integer"
            },
            {
              "$ref": "#/$defs/SizeReference"
            },
            {
              "$ref": "#/$defs/DataDependentSize"
            }
          ],
          "description": "The size/length of this axis can be specified as\n- fixed integer\n- reference to another axis with an optional offset ([SizeReference][])\n- data dependent size using [DataDependentSize][] (size is only known after model inference)",
          "examples": [
            10,
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ],
          "title": "Size"
        }
      },
      "required": [
        "type",
        "size"
      ],
      "title": "model.v0_5.IndexOutputAxis",
      "type": "object"
    },
    "IntervalOrRatioDataDescr": {
      "additionalProperties": false,
      "properties": {
        "type": {
          "default": "float32",
          "enum": [
            "float32",
            "float64",
            "uint8",
            "int8",
            "uint16",
            "int16",
            "uint32",
            "int32",
            "uint64",
            "int64"
          ],
          "examples": [
            "float32",
            "float64",
            "uint8",
            "uint16"
          ],
          "title": "Type",
          "type": "string"
        },
        "range": {
          "default": [
            null,
            null
          ],
          "description": "Tuple `(minimum, maximum)` specifying the allowed range of the data in this tensor.\n`None` corresponds to min/max of what can be expressed by **type**.",
          "maxItems": 2,
          "minItems": 2,
          "prefixItems": [
            {
              "anyOf": [
                {
                  "type": "number"
                },
                {
                  "type": "null"
                }
              ]
            },
            {
              "anyOf": [
                {
                  "type": "number"
                },
                {
                  "type": "null"
                }
              ]
            }
          ],
          "title": "Range",
          "type": "array"
        },
        "unit": {
          "anyOf": [
            {
              "const": "arbitrary unit",
              "type": "string"
            },
            {
              "description": "An SI unit",
              "minLength": 1,
              "pattern": "^(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^[+-]?[1-9]\\d*)?((\u00b7(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^[+-]?[1-9]\\d*)?)|(/(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^+?[1-9]\\d*)?))*$",
              "title": "SiUnit",
              "type": "string"
            }
          ],
          "default": "arbitrary unit",
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "description": "Scale for data on an interval (or ratio) scale.",
          "title": "Scale",
          "type": "number"
        },
        "offset": {
          "anyOf": [
            {
              "type": "number"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Offset for data on a ratio scale.",
          "title": "Offset"
        }
      },
      "title": "model.v0_5.IntervalOrRatioDataDescr",
      "type": "object"
    },
    "NominalOrOrdinalDataDescr": {
      "additionalProperties": false,
      "properties": {
        "values": {
          "anyOf": [
            {
              "items": {
                "type": "integer"
              },
              "minItems": 1,
              "type": "array"
            },
            {
              "items": {
                "type": "number"
              },
              "minItems": 1,
              "type": "array"
            },
            {
              "items": {
                "type": "boolean"
              },
              "minItems": 1,
              "type": "array"
            },
            {
              "items": {
                "type": "string"
              },
              "minItems": 1,
              "type": "array"
            }
          ],
          "description": "A fixed set of nominal or an ascending sequence of ordinal values.\nIn this case `data.type` is required to be an unsigend integer type, e.g. 'uint8'.\nString `values` are interpreted as labels for tensor values 0, ..., N.\nNote: as YAML 1.2 does not natively support a \"set\" datatype,\nnominal values should be given as a sequence (aka list/array) as well.",
          "title": "Values"
        },
        "type": {
          "default": "uint8",
          "enum": [
            "float32",
            "float64",
            "uint8",
            "int8",
            "uint16",
            "int16",
            "uint32",
            "int32",
            "uint64",
            "int64",
            "bool"
          ],
          "examples": [
            "float32",
            "uint8",
            "uint16",
            "int64",
            "bool"
          ],
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "const": "arbitrary unit",
              "type": "string"
            },
            {
              "description": "An SI unit",
              "minLength": 1,
              "pattern": "^(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^[+-]?[1-9]\\d*)?((\u00b7(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^[+-]?[1-9]\\d*)?)|(/(Q|R|Y|Z|E|P|T|G|M|k|h|da|d|c|m|\u00b5|n|p|f|a|z|y|r|q)?(m|g|s|A|K|mol|cd|Hz|N|Pa|J|W|C|V|F|\u03a9|S|Wb|T|H|lm|lx|Bq|Gy|Sv|kat|l|L)(\\^+?[1-9]\\d*)?))*$",
              "title": "SiUnit",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        }
      },
      "required": [
        "values"
      ],
      "title": "model.v0_5.NominalOrOrdinalDataDescr",
      "type": "object"
    },
    "ParameterizedSize": {
      "additionalProperties": false,
      "description": "Describes a range of valid tensor axis sizes as `size = min + n*step`.\n\n- **min** and **step** are given by the model description.\n- All blocksize paramters n = 0,1,2,... yield a valid `size`.\n- A greater blocksize paramter n = 0,1,2,... results in a greater **size**.\n  This allows to adjust the axis size more generically.",
      "properties": {
        "min": {
          "exclusiveMinimum": 0,
          "title": "Min",
          "type": "integer"
        },
        "step": {
          "exclusiveMinimum": 0,
          "title": "Step",
          "type": "integer"
        }
      },
      "required": [
        "min",
        "step"
      ],
      "title": "model.v0_5.ParameterizedSize",
      "type": "object"
    },
    "RelativeFilePath": {
      "description": "A path relative to the `rdf.yaml` file (also if the RDF source is a URL).",
      "format": "path",
      "title": "RelativeFilePath",
      "type": "string"
    },
    "SizeReference": {
      "additionalProperties": false,
      "description": "A tensor axis size (extent in pixels/frames) defined in relation to a reference axis.\n\n`axis.size = reference.size * reference.scale / axis.scale + offset`\n\nNote:\n1. The axis and the referenced axis need to have the same unit (or no unit).\n2. Batch axes may not be referenced.\n3. Fractions are rounded down.\n4. If the reference axis is `concatenable` the referencing axis is assumed to be\n    `concatenable` as well with the same block order.\n\nExample:\nAn unisotropic input image of w*h=100*49 pixels depicts a phsical space of 200*196mm\u00b2.\nLet's assume that we want to express the image height h in relation to its width w\ninstead of only accepting input images of exactly 100*49 pixels\n(for example to express a range of valid image shapes by parametrizing w, see `ParameterizedSize`).\n\n>>> w = SpaceInputAxis(id=AxisId(\"w\"), size=100, unit=\"millimeter\", scale=2)\n>>> h = SpaceInputAxis(\n...     id=AxisId(\"h\"),\n...     size=SizeReference(tensor_id=TensorId(\"input\"), axis_id=AxisId(\"w\"), offset=-1),\n...     unit=\"millimeter\",\n...     scale=4,\n... )\n>>> print(h.size.get_size(h, w))\n49\n\n\u21d2 h = w * w.scale / h.scale + offset = 100 * 2mm / 4mm - 1 = 49",
      "properties": {
        "tensor_id": {
          "description": "tensor id of the reference axis",
          "maxLength": 32,
          "minLength": 1,
          "title": "TensorId",
          "type": "string"
        },
        "axis_id": {
          "description": "axis id of the reference axis",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "offset": {
          "default": 0,
          "title": "Offset",
          "type": "integer"
        }
      },
      "required": [
        "tensor_id",
        "axis_id"
      ],
      "title": "model.v0_5.SizeReference",
      "type": "object"
    },
    "SpaceInputAxis": {
      "additionalProperties": false,
      "properties": {
        "size": {
          "anyOf": [
            {
              "exclusiveMinimum": 0,
              "type": "integer"
            },
            {
              "$ref": "#/$defs/ParameterizedSize"
            },
            {
              "$ref": "#/$defs/SizeReference"
            }
          ],
          "description": "The size/length of this axis can be specified as\n- fixed integer\n- parameterized series of valid sizes ([ParameterizedSize][])\n- reference to another axis with an optional offset ([SizeReference][])",
          "examples": [
            10,
            {
              "min": 32,
              "step": 16
            },
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ],
          "title": "Size"
        },
        "id": {
          "default": "x",
          "examples": [
            "x",
            "y",
            "z"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "space",
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "enum": [
                "attometer",
                "angstrom",
                "centimeter",
                "decimeter",
                "exameter",
                "femtometer",
                "foot",
                "gigameter",
                "hectometer",
                "inch",
                "kilometer",
                "megameter",
                "meter",
                "micrometer",
                "mile",
                "millimeter",
                "nanometer",
                "parsec",
                "petameter",
                "picometer",
                "terameter",
                "yard",
                "yoctometer",
                "yottameter",
                "zeptometer",
                "zettameter"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "exclusiveMinimum": 0,
          "title": "Scale",
          "type": "number"
        },
        "concatenable": {
          "default": false,
          "description": "If a model has a `concatenable` input axis, it can be processed blockwise,\nsplitting a longer sample axis into blocks matching its input tensor description.\nOutput axes are concatenable if they have a [SizeReference][] to a concatenable\ninput axis.",
          "title": "Concatenable",
          "type": "boolean"
        }
      },
      "required": [
        "size",
        "type"
      ],
      "title": "model.v0_5.SpaceInputAxis",
      "type": "object"
    },
    "SpaceOutputAxis": {
      "additionalProperties": false,
      "properties": {
        "size": {
          "anyOf": [
            {
              "exclusiveMinimum": 0,
              "type": "integer"
            },
            {
              "$ref": "#/$defs/SizeReference"
            }
          ],
          "description": "The size/length of this axis can be specified as\n- fixed integer\n- reference to another axis with an optional offset (see [SizeReference][])",
          "examples": [
            10,
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ],
          "title": "Size"
        },
        "id": {
          "default": "x",
          "examples": [
            "x",
            "y",
            "z"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "space",
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "enum": [
                "attometer",
                "angstrom",
                "centimeter",
                "decimeter",
                "exameter",
                "femtometer",
                "foot",
                "gigameter",
                "hectometer",
                "inch",
                "kilometer",
                "megameter",
                "meter",
                "micrometer",
                "mile",
                "millimeter",
                "nanometer",
                "parsec",
                "petameter",
                "picometer",
                "terameter",
                "yard",
                "yoctometer",
                "yottameter",
                "zeptometer",
                "zettameter"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "exclusiveMinimum": 0,
          "title": "Scale",
          "type": "number"
        }
      },
      "required": [
        "size",
        "type"
      ],
      "title": "model.v0_5.SpaceOutputAxis",
      "type": "object"
    },
    "SpaceOutputAxisWithHalo": {
      "additionalProperties": false,
      "properties": {
        "halo": {
          "description": "The halo should be cropped from the output tensor to avoid boundary effects.\nIt is to be cropped from both sides, i.e. `size_after_crop = size - 2 * halo`.\nTo document a halo that is already cropped by the model use `size.offset` instead.",
          "minimum": 1,
          "title": "Halo",
          "type": "integer"
        },
        "size": {
          "$ref": "#/$defs/SizeReference",
          "description": "reference to another axis with an optional offset (see [SizeReference][])",
          "examples": [
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ]
        },
        "id": {
          "default": "x",
          "examples": [
            "x",
            "y",
            "z"
          ],
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "space",
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "enum": [
                "attometer",
                "angstrom",
                "centimeter",
                "decimeter",
                "exameter",
                "femtometer",
                "foot",
                "gigameter",
                "hectometer",
                "inch",
                "kilometer",
                "megameter",
                "meter",
                "micrometer",
                "mile",
                "millimeter",
                "nanometer",
                "parsec",
                "petameter",
                "picometer",
                "terameter",
                "yard",
                "yoctometer",
                "yottameter",
                "zeptometer",
                "zettameter"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "exclusiveMinimum": 0,
          "title": "Scale",
          "type": "number"
        }
      },
      "required": [
        "halo",
        "size",
        "type"
      ],
      "title": "model.v0_5.SpaceOutputAxisWithHalo",
      "type": "object"
    },
    "TimeInputAxis": {
      "additionalProperties": false,
      "properties": {
        "size": {
          "anyOf": [
            {
              "exclusiveMinimum": 0,
              "type": "integer"
            },
            {
              "$ref": "#/$defs/ParameterizedSize"
            },
            {
              "$ref": "#/$defs/SizeReference"
            }
          ],
          "description": "The size/length of this axis can be specified as\n- fixed integer\n- parameterized series of valid sizes ([ParameterizedSize][])\n- reference to another axis with an optional offset ([SizeReference][])",
          "examples": [
            10,
            {
              "min": 32,
              "step": 16
            },
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ],
          "title": "Size"
        },
        "id": {
          "default": "time",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "time",
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "enum": [
                "attosecond",
                "centisecond",
                "day",
                "decisecond",
                "exasecond",
                "femtosecond",
                "gigasecond",
                "hectosecond",
                "hour",
                "kilosecond",
                "megasecond",
                "microsecond",
                "millisecond",
                "minute",
                "nanosecond",
                "petasecond",
                "picosecond",
                "second",
                "terasecond",
                "yoctosecond",
                "yottasecond",
                "zeptosecond",
                "zettasecond"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "exclusiveMinimum": 0,
          "title": "Scale",
          "type": "number"
        },
        "concatenable": {
          "default": false,
          "description": "If a model has a `concatenable` input axis, it can be processed blockwise,\nsplitting a longer sample axis into blocks matching its input tensor description.\nOutput axes are concatenable if they have a [SizeReference][] to a concatenable\ninput axis.",
          "title": "Concatenable",
          "type": "boolean"
        }
      },
      "required": [
        "size",
        "type"
      ],
      "title": "model.v0_5.TimeInputAxis",
      "type": "object"
    },
    "TimeOutputAxis": {
      "additionalProperties": false,
      "properties": {
        "size": {
          "anyOf": [
            {
              "exclusiveMinimum": 0,
              "type": "integer"
            },
            {
              "$ref": "#/$defs/SizeReference"
            }
          ],
          "description": "The size/length of this axis can be specified as\n- fixed integer\n- reference to another axis with an optional offset (see [SizeReference][])",
          "examples": [
            10,
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ],
          "title": "Size"
        },
        "id": {
          "default": "time",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "time",
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "enum": [
                "attosecond",
                "centisecond",
                "day",
                "decisecond",
                "exasecond",
                "femtosecond",
                "gigasecond",
                "hectosecond",
                "hour",
                "kilosecond",
                "megasecond",
                "microsecond",
                "millisecond",
                "minute",
                "nanosecond",
                "petasecond",
                "picosecond",
                "second",
                "terasecond",
                "yoctosecond",
                "yottasecond",
                "zeptosecond",
                "zettasecond"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "exclusiveMinimum": 0,
          "title": "Scale",
          "type": "number"
        }
      },
      "required": [
        "size",
        "type"
      ],
      "title": "model.v0_5.TimeOutputAxis",
      "type": "object"
    },
    "TimeOutputAxisWithHalo": {
      "additionalProperties": false,
      "properties": {
        "halo": {
          "description": "The halo should be cropped from the output tensor to avoid boundary effects.\nIt is to be cropped from both sides, i.e. `size_after_crop = size - 2 * halo`.\nTo document a halo that is already cropped by the model use `size.offset` instead.",
          "minimum": 1,
          "title": "Halo",
          "type": "integer"
        },
        "size": {
          "$ref": "#/$defs/SizeReference",
          "description": "reference to another axis with an optional offset (see [SizeReference][])",
          "examples": [
            {
              "axis_id": "a",
              "offset": 5,
              "tensor_id": "t"
            }
          ]
        },
        "id": {
          "default": "time",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "description": {
          "default": "",
          "description": "A short description of this axis beyond its type and id.",
          "maxLength": 128,
          "title": "Description",
          "type": "string"
        },
        "type": {
          "const": "time",
          "title": "Type",
          "type": "string"
        },
        "unit": {
          "anyOf": [
            {
              "enum": [
                "attosecond",
                "centisecond",
                "day",
                "decisecond",
                "exasecond",
                "femtosecond",
                "gigasecond",
                "hectosecond",
                "hour",
                "kilosecond",
                "megasecond",
                "microsecond",
                "millisecond",
                "minute",
                "nanosecond",
                "petasecond",
                "picosecond",
                "second",
                "terasecond",
                "yoctosecond",
                "yottasecond",
                "zeptosecond",
                "zettasecond"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Unit"
        },
        "scale": {
          "default": 1.0,
          "exclusiveMinimum": 0,
          "title": "Scale",
          "type": "number"
        }
      },
      "required": [
        "halo",
        "size",
        "type"
      ],
      "title": "model.v0_5.TimeOutputAxisWithHalo",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "properties": {
    "id": {
      "description": "Tensor id. No duplicates are allowed.",
      "maxLength": 32,
      "minLength": 1,
      "title": "TensorId",
      "type": "string"
    },
    "description": {
      "default": "",
      "description": "free text description",
      "maxLength": 128,
      "title": "Description",
      "type": "string"
    },
    "axes": {
      "description": "tensor axes",
      "items": {
        "anyOf": [
          {
            "discriminator": {
              "mapping": {
                "batch": "#/$defs/BatchAxis",
                "channel": "#/$defs/ChannelAxis",
                "index": "#/$defs/IndexInputAxis",
                "space": "#/$defs/SpaceInputAxis",
                "time": "#/$defs/TimeInputAxis"
              },
              "propertyName": "type"
            },
            "oneOf": [
              {
                "$ref": "#/$defs/BatchAxis"
              },
              {
                "$ref": "#/$defs/ChannelAxis"
              },
              {
                "$ref": "#/$defs/IndexInputAxis"
              },
              {
                "$ref": "#/$defs/TimeInputAxis"
              },
              {
                "$ref": "#/$defs/SpaceInputAxis"
              }
            ]
          },
          {
            "discriminator": {
              "mapping": {
                "batch": "#/$defs/BatchAxis",
                "channel": "#/$defs/ChannelAxis",
                "index": "#/$defs/IndexOutputAxis",
                "space": {
                  "oneOf": [
                    {
                      "$ref": "#/$defs/SpaceOutputAxis"
                    },
                    {
                      "$ref": "#/$defs/SpaceOutputAxisWithHalo"
                    }
                  ]
                },
                "time": {
                  "oneOf": [
                    {
                      "$ref": "#/$defs/TimeOutputAxis"
                    },
                    {
                      "$ref": "#/$defs/TimeOutputAxisWithHalo"
                    }
                  ]
                }
              },
              "propertyName": "type"
            },
            "oneOf": [
              {
                "$ref": "#/$defs/BatchAxis"
              },
              {
                "$ref": "#/$defs/ChannelAxis"
              },
              {
                "$ref": "#/$defs/IndexOutputAxis"
              },
              {
                "oneOf": [
                  {
                    "$ref": "#/$defs/TimeOutputAxis"
                  },
                  {
                    "$ref": "#/$defs/TimeOutputAxisWithHalo"
                  }
                ]
              },
              {
                "oneOf": [
                  {
                    "$ref": "#/$defs/SpaceOutputAxis"
                  },
                  {
                    "$ref": "#/$defs/SpaceOutputAxisWithHalo"
                  }
                ]
              }
            ]
          }
        ]
      },
      "minItems": 1,
      "title": "Axes",
      "type": "array"
    },
    "test_tensor": {
      "anyOf": [
        {
          "$ref": "#/$defs/FileDescr"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "An example tensor to use for testing.\nUsing the model with the test input tensors is expected to yield the test output tensors.\nEach test tensor has be a an ndarray in the\n[numpy.lib file format](https://numpy.org/doc/stable/reference/generated/numpy.lib.format.html#module-numpy.lib.format).\nThe file extension must be '.npy'."
    },
    "sample_tensor": {
      "anyOf": [
        {
          "$ref": "#/$defs/FileDescr"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "A sample tensor to illustrate a possible input/output for the model,\nThe sample image primarily serves to inform a human user about an example use case\nand is typically stored as .hdf5, .png or .tiff.\nIt has to be readable by the [imageio library](https://imageio.readthedocs.io/en/stable/formats/index.html#supported-formats)\n(numpy's `.npy` format is not supported).\nThe image dimensionality has to match the number of axes specified in this tensor description."
    },
    "data": {
      "anyOf": [
        {
          "$ref": "#/$defs/NominalOrOrdinalDataDescr"
        },
        {
          "$ref": "#/$defs/IntervalOrRatioDataDescr"
        },
        {
          "items": {
            "anyOf": [
              {
                "$ref": "#/$defs/NominalOrOrdinalDataDescr"
              },
              {
                "$ref": "#/$defs/IntervalOrRatioDataDescr"
              }
            ]
          },
          "minItems": 1,
          "type": "array"
        }
      ],
      "default": {
        "type": "float32",
        "range": [
          null,
          null
        ],
        "unit": "arbitrary unit",
        "scale": 1.0,
        "offset": null
      },
      "description": "Description of the tensor's data values, optionally per channel.\nIf specified per channel, the data `type` needs to match across channels.",
      "title": "Data"
    }
  },
  "required": [
    "id",
    "axes"
  ],
  "title": "model.v0_5.TensorDescrBase",
  "type": "object"
}

Fields:

• id (TensorId)
• description (Annotated[str, MaxLen(128)])
• axes (NotEmpty[Sequence[IO_AxisT]])
• test_tensor (FAIR[FileDescr_package | None])
• sample_tensor (FAIR[FileDescr_package | None])
• data (TensorDataDescr | NotEmpty[Sequence[TensorDataDescr]])

Validators:

• _validate_axes → axes
• _validate_sample_tensor
• _check_data_type_across_channels → data
• _check_data_matches_channelaxis

axes pydantic-field

axes: NotEmpty[Sequence[IO_AxisT]]

tensor axes

data pydantic-field

data: TensorDataDescr | NotEmpty[Sequence[TensorDataDescr]]

Description of the tensor's data values, optionally per channel. If specified per channel, the data type needs to match across channels.

description pydantic-field

description: Annotated[str, MaxLen(128)] = ''

free text description

dtype property

dtype: Literal[
    "float32",
    "float64",
    "uint8",
    "int8",
    "uint16",
    "int16",
    "uint32",
    "int32",
    "uint64",
    "int64",
    "bool",
]

dtype as specified under data.type or data[i].type

id pydantic-field

id: TensorId

Tensor id. No duplicates are allowed.

sample_tensor pydantic-field

sample_tensor: FAIR[FileDescr_package | None] = None

A sample tensor to illustrate a possible input/output for the model, The sample image primarily serves to inform a human user about an example use case and is typically stored as .hdf5, .png or .tiff. It has to be readable by the imageio library (numpy's .npy format is not supported). The image dimensionality has to match the number of axes specified in this tensor description.

shape property

shape

test_tensor pydantic-field

test_tensor: FAIR[FileDescr_package | None] = None

An example tensor to use for testing. Using the model with the test input tensors is expected to yield the test output tensors. Each test tensor has be a an ndarray in the numpy.lib file format. The file extension must be '.npy'.

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get_axis_sizes_for_array

get_axis_sizes_for_array(
    array: NDArray[Any],
) -> dict[AxisId, int]

Source code in src/bioimageio/spec/model/v0_5.py

def get_axis_sizes_for_array(self, array: NDArray[Any]) -> dict[AxisId, int]:
    if len(array.shape) != len(self.axes):
        raise ValueError(
            f"Dimension mismatch: array shape {array.shape} (#{len(array.shape)})"
            + f" incompatible with {len(self.axes)} axes."
        )
    return {a.id: array.shape[i] for i, a in enumerate(self.axes)}

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

TensorId

Bases: LowerCaseIdentifier

              flowchart TD
              bioimageio.spec.model.v0_5.TensorId[TensorId]
              bioimageio.spec._internal.types.LowerCaseIdentifier[LowerCaseIdentifier]
              bioimageio.spec._internal.validated_string.ValidatedString[ValidatedString]

                              bioimageio.spec._internal.types.LowerCaseIdentifier --> bioimageio.spec.model.v0_5.TensorId
                                bioimageio.spec._internal.validated_string.ValidatedString --> bioimageio.spec._internal.types.LowerCaseIdentifier
                



              click bioimageio.spec.model.v0_5.TensorId href "" "bioimageio.spec.model.v0_5.TensorId"
              click bioimageio.spec._internal.types.LowerCaseIdentifier href "" "bioimageio.spec._internal.types.LowerCaseIdentifier"
              click bioimageio.spec._internal.validated_string.ValidatedString href "" "bioimageio.spec._internal.validated_string.ValidatedString"
            

Returned by:

•  v0_5  get_halos

Used by:

•  v0_5
  •  get_halos
  •  validate_tensors

Methods:

Name	Description
__get_pydantic_core_schema__
__get_pydantic_json_schema__
__new__

Attributes:

Name	Type	Description
root_model	type[RootModel[Any]]	the pydantic root model to validate the string

root_model class-attribute

root_model: type[RootModel[Any]] = RootModel[
    Annotated[LowerCaseIdentifierAnno, MaxLen(32)]
]

the pydantic root model to validate the string

__get_pydantic_core_schema__ classmethod

__get_pydantic_core_schema__(
    source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_core_schema__(
    cls, source_type: Any, handler: GetCoreSchemaHandler
) -> CoreSchema:
    return no_info_after_validator_function(cls, handler(str))

__get_pydantic_json_schema__ classmethod

__get_pydantic_json_schema__(
    core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue

Source code in src/bioimageio/spec/_internal/validated_string.py

@classmethod
def __get_pydantic_json_schema__(
    cls, core_schema: CoreSchema, handler: GetJsonSchemaHandler
) -> JsonSchemaValue:
    json_schema = cls.root_model.model_json_schema(mode=handler.mode)
    json_schema["title"] = cls.__name__.strip("_")
    if cls.__doc__:
        json_schema["description"] = cls.__doc__

    return json_schema

__new__

__new__(object: object)

Source code in src/bioimageio/spec/_internal/validated_string.py

def __new__(cls, object: object):
    _validated = cls.root_model.model_validate(object).root
    self = super().__new__(cls, _validated)
    self._validated = _validated
    return self._after_validator()

TensorflowJsWeightsDescr pydantic-model

Bases: WeightsEntryDescrBase

Show JSON schema:

{
  "$defs": {
    "Author": {
      "additionalProperties": false,
      "properties": {
        "affiliation": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Affiliation",
          "title": "Affiliation"
        },
        "email": {
          "anyOf": [
            {
              "format": "email",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Email",
          "title": "Email"
        },
        "orcid": {
          "anyOf": [
            {
              "description": "An ORCID identifier, see https://orcid.org/",
              "title": "OrcidId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "An [ORCID iD](https://support.orcid.org/hc/en-us/sections/360001495313-What-is-ORCID\n) in hyphenated groups of 4 digits, (and [valid](\nhttps://support.orcid.org/hc/en-us/articles/360006897674-Structure-of-the-ORCID-Identifier\n) as per ISO 7064 11,2.)",
          "examples": [
            "0000-0001-2345-6789"
          ],
          "title": "Orcid"
        },
        "name": {
          "title": "Name",
          "type": "string"
        },
        "github_user": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Github User"
        }
      },
      "required": [
        "name"
      ],
      "title": "generic.v0_3.Author",
      "type": "object"
    },
    "RelativeFilePath": {
      "description": "A path relative to the `rdf.yaml` file (also if the RDF source is a URL).",
      "format": "path",
      "title": "RelativeFilePath",
      "type": "string"
    },
    "Version": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        }
      ],
      "description": "wraps a packaging.version.Version instance for validation in pydantic models",
      "title": "Version"
    }
  },
  "additionalProperties": false,
  "properties": {
    "source": {
      "anyOf": [
        {
          "description": "A URL with the HTTP or HTTPS scheme.",
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "title": "HttpUrl",
          "type": "string"
        },
        {
          "$ref": "#/$defs/RelativeFilePath"
        },
        {
          "format": "file-path",
          "title": "FilePath",
          "type": "string"
        }
      ],
      "description": "The multi-file weights.\nAll required files/folders should be a zip archive.",
      "title": "FileSource"
    },
    "sha256": {
      "anyOf": [
        {
          "description": "A SHA-256 hash value",
          "maxLength": 64,
          "minLength": 64,
          "title": "Sha256",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "SHA256 hash value of the **source** file.",
      "title": "Sha256"
    },
    "authors": {
      "anyOf": [
        {
          "items": {
            "$ref": "#/$defs/Author"
          },
          "type": "array"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
      "title": "Authors"
    },
    "parent": {
      "anyOf": [
        {
          "enum": [
            "keras_hdf5",
            "keras_v3",
            "onnx",
            "pytorch_state_dict",
            "tensorflow_js",
            "tensorflow_saved_model_bundle",
            "torchscript"
          ],
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
      "examples": [
        "pytorch_state_dict"
      ],
      "title": "Parent"
    },
    "comment": {
      "default": "",
      "description": "A comment about this weights entry, for example how these weights were created.",
      "title": "Comment",
      "type": "string"
    },
    "tensorflow_version": {
      "$ref": "#/$defs/Version",
      "description": "Version of the TensorFlow library used."
    }
  },
  "required": [
    "source",
    "tensorflow_version"
  ],
  "title": "model.v0_5.TensorflowJsWeightsDescr",
  "type": "object"
}

Fields:

• sha256 (Sha256 | None)
• authors (list[Author] | None)
• parent (Annotated[WeightsFormat | None, Field(examples=['pytorch_state_dict'])])
• comment (str)
• tensorflow_version (Version)
• source (Annotated[FileSource, AfterValidator(wo_special_file_name)])

Validators:

• _validate

authors pydantic-field

authors: list[Author] | None = None

Authors Either the person(s) that have trained this model resulting in the original weights file. (If this is the initial weights entry, i.e. it does not have a parent) Or the person(s) who have converted the weights to this weights format. (If this is a child weight, i.e. it has a parent field)

comment pydantic-field

comment: str = ''

A comment about this weights entry, for example how these weights were created.

parent pydantic-field

parent: Annotated[
    WeightsFormat | None,
    Field(examples=["pytorch_state_dict"]),
] = None

The source weights these weights were converted from. For example, if a model's weights were converted from the pytorch_state_dict format to torchscript, The pytorch_state_dict weights entry has no parent and is the parent of the torchscript weights. All weight entries except one (the initial set of weights resulting from training the model), need to have this field.

sha256 pydantic-field

sha256: Sha256 | None = None

SHA256 hash value of the source file.

source pydantic-field

source: Annotated[
    FileSource, AfterValidator(wo_special_file_name)
]

The multi-file weights. All required files/folders should be a zip archive.

suffix property

suffix: str

tensorflow_version pydantic-field

tensorflow_version: Version

Version of the TensorFlow library used.

type class-attribute

type: WeightsFormat = 'tensorflow_js'

weights_format_name class-attribute

weights_format_name: str = 'Tensorflow.js'

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

download

download(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

alias for .get_reader

Source code in src/bioimageio/spec/_internal/io.py

def download(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """alias for `.get_reader`"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

get_reader

get_reader(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

open the file source (download if needed)

Source code in src/bioimageio/spec/_internal/io.py

def get_reader(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """open the file source (download if needed)"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

validate_sha256

validate_sha256(force_recompute: bool = False) -> None

validate the sha256 hash value of the source file

Source code in src/bioimageio/spec/_internal/io.py

def validate_sha256(self, force_recompute: bool = False) -> None:
    """validate the sha256 hash value of the **source** file"""
    context = get_validation_context()
    src_str = str(self.source)
    if force_recompute:
        actual_sha = None
    else:
        actual_sha = context.known_files.get(src_str)

    if actual_sha is None:
        if context.perform_io_checks or force_recompute:
            reader = get_reader(self.source, sha256=self.sha256)
            if force_recompute:
                actual_sha = get_sha256(reader)
            else:
                actual_sha = reader.sha256

            context.known_files[src_str] = actual_sha
        elif context.known_files and src_str not in context.known_files:
            # perform_io_checks is False, but known files were given,
            # so we expect all file references to be in there
            raise ValueError(f"File {src_str} not found in `known_files`.")

    if actual_sha is None or self.sha256 == actual_sha:
        return
    elif self.sha256 is None or context.update_hashes:
        self.sha256 = actual_sha
    elif self.sha256 != actual_sha:
        raise ValueError(
            f"Sha256 mismatch for {self.source}. Expected {self.sha256}, got "
            + f"{actual_sha}. Update expected `sha256` or point to the matching "
            + "file."
        )

TensorflowSavedModelBundleWeightsDescr pydantic-model

Bases: WeightsEntryDescrBase

Show JSON schema:

{
  "$defs": {
    "Author": {
      "additionalProperties": false,
      "properties": {
        "affiliation": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Affiliation",
          "title": "Affiliation"
        },
        "email": {
          "anyOf": [
            {
              "format": "email",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Email",
          "title": "Email"
        },
        "orcid": {
          "anyOf": [
            {
              "description": "An ORCID identifier, see https://orcid.org/",
              "title": "OrcidId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "An [ORCID iD](https://support.orcid.org/hc/en-us/sections/360001495313-What-is-ORCID\n) in hyphenated groups of 4 digits, (and [valid](\nhttps://support.orcid.org/hc/en-us/articles/360006897674-Structure-of-the-ORCID-Identifier\n) as per ISO 7064 11,2.)",
          "examples": [
            "0000-0001-2345-6789"
          ],
          "title": "Orcid"
        },
        "name": {
          "title": "Name",
          "type": "string"
        },
        "github_user": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Github User"
        }
      },
      "required": [
        "name"
      ],
      "title": "generic.v0_3.Author",
      "type": "object"
    },
    "FileDescr": {
      "additionalProperties": false,
      "description": "A file description",
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "File source",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        }
      },
      "required": [
        "source"
      ],
      "title": "_internal.io.FileDescr",
      "type": "object"
    },
    "RelativeFilePath": {
      "description": "A path relative to the `rdf.yaml` file (also if the RDF source is a URL).",
      "format": "path",
      "title": "RelativeFilePath",
      "type": "string"
    },
    "Version": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        }
      ],
      "description": "wraps a packaging.version.Version instance for validation in pydantic models",
      "title": "Version"
    }
  },
  "additionalProperties": false,
  "properties": {
    "source": {
      "anyOf": [
        {
          "description": "A URL with the HTTP or HTTPS scheme.",
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "title": "HttpUrl",
          "type": "string"
        },
        {
          "$ref": "#/$defs/RelativeFilePath"
        },
        {
          "format": "file-path",
          "title": "FilePath",
          "type": "string"
        }
      ],
      "description": "The multi-file weights.\nAll required files/folders should be a zip archive.",
      "title": "FileSource"
    },
    "sha256": {
      "anyOf": [
        {
          "description": "A SHA-256 hash value",
          "maxLength": 64,
          "minLength": 64,
          "title": "Sha256",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "SHA256 hash value of the **source** file.",
      "title": "Sha256"
    },
    "authors": {
      "anyOf": [
        {
          "items": {
            "$ref": "#/$defs/Author"
          },
          "type": "array"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
      "title": "Authors"
    },
    "parent": {
      "anyOf": [
        {
          "enum": [
            "keras_hdf5",
            "keras_v3",
            "onnx",
            "pytorch_state_dict",
            "tensorflow_js",
            "tensorflow_saved_model_bundle",
            "torchscript"
          ],
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
      "examples": [
        "pytorch_state_dict"
      ],
      "title": "Parent"
    },
    "comment": {
      "default": "",
      "description": "A comment about this weights entry, for example how these weights were created.",
      "title": "Comment",
      "type": "string"
    },
    "tensorflow_version": {
      "$ref": "#/$defs/Version",
      "description": "Version of the TensorFlow library used."
    },
    "dependencies": {
      "anyOf": [
        {
          "$ref": "#/$defs/FileDescr",
          "examples": [
            {
              "source": "environment.yaml"
            }
          ]
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Custom dependencies beyond tensorflow.\nShould include tensorflow and any version pinning has to be compatible with **tensorflow_version**."
    }
  },
  "required": [
    "source",
    "tensorflow_version"
  ],
  "title": "model.v0_5.TensorflowSavedModelBundleWeightsDescr",
  "type": "object"
}

Fields:

• sha256 (Sha256 | None)
• authors (list[Author] | None)
• parent (Annotated[WeightsFormat | None, Field(examples=['pytorch_state_dict'])])
• comment (str)
• tensorflow_version (Version)
• dependencies (FileDescr_dependencies | None)
• source (Annotated[FileSource, AfterValidator(wo_special_file_name)])

Validators:

• _validate

authors pydantic-field

authors: list[Author] | None = None

Authors Either the person(s) that have trained this model resulting in the original weights file. (If this is the initial weights entry, i.e. it does not have a parent) Or the person(s) who have converted the weights to this weights format. (If this is a child weight, i.e. it has a parent field)

comment pydantic-field

comment: str = ''

A comment about this weights entry, for example how these weights were created.

dependencies pydantic-field

dependencies: FileDescr_dependencies | None = None

Custom dependencies beyond tensorflow. Should include tensorflow and any version pinning has to be compatible with tensorflow_version.

parent pydantic-field

parent: Annotated[
    WeightsFormat | None,
    Field(examples=["pytorch_state_dict"]),
] = None

The source weights these weights were converted from. For example, if a model's weights were converted from the pytorch_state_dict format to torchscript, The pytorch_state_dict weights entry has no parent and is the parent of the torchscript weights. All weight entries except one (the initial set of weights resulting from training the model), need to have this field.

sha256 pydantic-field

sha256: Sha256 | None = None

SHA256 hash value of the source file.

source pydantic-field

source: Annotated[
    FileSource, AfterValidator(wo_special_file_name)
]

The multi-file weights. All required files/folders should be a zip archive.

suffix property

suffix: str

tensorflow_version pydantic-field

tensorflow_version: Version

Version of the TensorFlow library used.

type class-attribute

type: WeightsFormat = 'tensorflow_saved_model_bundle'

weights_format_name class-attribute

weights_format_name: str = 'Tensorflow Saved Model'

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

download

download(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

alias for .get_reader

Source code in src/bioimageio/spec/_internal/io.py

def download(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """alias for `.get_reader`"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

get_reader

get_reader(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

open the file source (download if needed)

Source code in src/bioimageio/spec/_internal/io.py

def get_reader(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """open the file source (download if needed)"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

validate_sha256

validate_sha256(force_recompute: bool = False) -> None

validate the sha256 hash value of the source file

Source code in src/bioimageio/spec/_internal/io.py

def validate_sha256(self, force_recompute: bool = False) -> None:
    """validate the sha256 hash value of the **source** file"""
    context = get_validation_context()
    src_str = str(self.source)
    if force_recompute:
        actual_sha = None
    else:
        actual_sha = context.known_files.get(src_str)

    if actual_sha is None:
        if context.perform_io_checks or force_recompute:
            reader = get_reader(self.source, sha256=self.sha256)
            if force_recompute:
                actual_sha = get_sha256(reader)
            else:
                actual_sha = reader.sha256

            context.known_files[src_str] = actual_sha
        elif context.known_files and src_str not in context.known_files:
            # perform_io_checks is False, but known files were given,
            # so we expect all file references to be in there
            raise ValueError(f"File {src_str} not found in `known_files`.")

    if actual_sha is None or self.sha256 == actual_sha:
        return
    elif self.sha256 is None or context.update_hashes:
        self.sha256 = actual_sha
    elif self.sha256 != actual_sha:
        raise ValueError(
            f"Sha256 mismatch for {self.source}. Expected {self.sha256}, got "
            + f"{actual_sha}. Update expected `sha256` or point to the matching "
            + "file."
        )

TimeAxisBase pydantic-model

Bases: AxisBase

Show JSON schema:

{
  "additionalProperties": false,
  "properties": {
    "id": {
      "default": "time",
      "maxLength": 16,
      "minLength": 1,
      "title": "AxisId",
      "type": "string"
    },
    "description": {
      "default": "",
      "description": "A short description of this axis beyond its type and id.",
      "maxLength": 128,
      "title": "Description",
      "type": "string"
    },
    "type": {
      "const": "time",
      "title": "Type",
      "type": "string"
    },
    "unit": {
      "anyOf": [
        {
          "enum": [
            "attosecond",
            "centisecond",
            "day",
            "decisecond",
            "exasecond",
            "femtosecond",
            "gigasecond",
            "hectosecond",
            "hour",
            "kilosecond",
            "megasecond",
            "microsecond",
            "millisecond",
            "minute",
            "nanosecond",
            "petasecond",
            "picosecond",
            "second",
            "terasecond",
            "yoctosecond",
            "yottasecond",
            "zeptosecond",
            "zettasecond"
          ],
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Unit"
    },
    "scale": {
      "default": 1.0,
      "exclusiveMinimum": 0,
      "title": "Scale",
      "type": "number"
    }
  },
  "required": [
    "type"
  ],
  "title": "model.v0_5.TimeAxisBase",
  "type": "object"
}

Fields:

• description (Annotated[str, MaxLen(128)])
• type (Literal['time'])
• id (NonBatchAxisId)
• unit (TimeUnit | None)
• scale (Annotated[float, Gt(0)])

description pydantic-field

description: Annotated[str, MaxLen(128)] = ''

A short description of this axis beyond its type and id.

id pydantic-field

id: NonBatchAxisId = DEFAULT_TIME_AXIS_ID

An axis id unique across all axes of one tensor.

implemented_type class-attribute

implemented_type: Literal['time'] = 'time'

scale pydantic-field

scale: Annotated[float, Gt(0)] = 1.0

type pydantic-field

type: Literal['time'] = 'time'

unit pydantic-field

unit: TimeUnit | None = None

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

TimeInputAxis pydantic-model

Bases: TimeAxisBase, _WithInputAxisSize

Show JSON schema:

{
  "$defs": {
    "ParameterizedSize": {
      "additionalProperties": false,
      "description": "Describes a range of valid tensor axis sizes as `size = min + n*step`.\n\n- **min** and **step** are given by the model description.\n- All blocksize paramters n = 0,1,2,... yield a valid `size`.\n- A greater blocksize paramter n = 0,1,2,... results in a greater **size**.\n  This allows to adjust the axis size more generically.",
      "properties": {
        "min": {
          "exclusiveMinimum": 0,
          "title": "Min",
          "type": "integer"
        },
        "step": {
          "exclusiveMinimum": 0,
          "title": "Step",
          "type": "integer"
        }
      },
      "required": [
        "min",
        "step"
      ],
      "title": "model.v0_5.ParameterizedSize",
      "type": "object"
    },
    "SizeReference": {
      "additionalProperties": false,
      "description": "A tensor axis size (extent in pixels/frames) defined in relation to a reference axis.\n\n`axis.size = reference.size * reference.scale / axis.scale + offset`\n\nNote:\n1. The axis and the referenced axis need to have the same unit (or no unit).\n2. Batch axes may not be referenced.\n3. Fractions are rounded down.\n4. If the reference axis is `concatenable` the referencing axis is assumed to be\n    `concatenable` as well with the same block order.\n\nExample:\nAn unisotropic input image of w*h=100*49 pixels depicts a phsical space of 200*196mm\u00b2.\nLet's assume that we want to express the image height h in relation to its width w\ninstead of only accepting input images of exactly 100*49 pixels\n(for example to express a range of valid image shapes by parametrizing w, see `ParameterizedSize`).\n\n>>> w = SpaceInputAxis(id=AxisId(\"w\"), size=100, unit=\"millimeter\", scale=2)\n>>> h = SpaceInputAxis(\n...     id=AxisId(\"h\"),\n...     size=SizeReference(tensor_id=TensorId(\"input\"), axis_id=AxisId(\"w\"), offset=-1),\n...     unit=\"millimeter\",\n...     scale=4,\n... )\n>>> print(h.size.get_size(h, w))\n49\n\n\u21d2 h = w * w.scale / h.scale + offset = 100 * 2mm / 4mm - 1 = 49",
      "properties": {
        "tensor_id": {
          "description": "tensor id of the reference axis",
          "maxLength": 32,
          "minLength": 1,
          "title": "TensorId",
          "type": "string"
        },
        "axis_id": {
          "description": "axis id of the reference axis",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "offset": {
          "default": 0,
          "title": "Offset",
          "type": "integer"
        }
      },
      "required": [
        "tensor_id",
        "axis_id"
      ],
      "title": "model.v0_5.SizeReference",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "properties": {
    "size": {
      "anyOf": [
        {
          "exclusiveMinimum": 0,
          "type": "integer"
        },
        {
          "$ref": "#/$defs/ParameterizedSize"
        },
        {
          "$ref": "#/$defs/SizeReference"
        }
      ],
      "description": "The size/length of this axis can be specified as\n- fixed integer\n- parameterized series of valid sizes ([ParameterizedSize][])\n- reference to another axis with an optional offset ([SizeReference][])",
      "examples": [
        10,
        {
          "min": 32,
          "step": 16
        },
        {
          "axis_id": "a",
          "offset": 5,
          "tensor_id": "t"
        }
      ],
      "title": "Size"
    },
    "id": {
      "default": "time",
      "maxLength": 16,
      "minLength": 1,
      "title": "AxisId",
      "type": "string"
    },
    "description": {
      "default": "",
      "description": "A short description of this axis beyond its type and id.",
      "maxLength": 128,
      "title": "Description",
      "type": "string"
    },
    "type": {
      "const": "time",
      "title": "Type",
      "type": "string"
    },
    "unit": {
      "anyOf": [
        {
          "enum": [
            "attosecond",
            "centisecond",
            "day",
            "decisecond",
            "exasecond",
            "femtosecond",
            "gigasecond",
            "hectosecond",
            "hour",
            "kilosecond",
            "megasecond",
            "microsecond",
            "millisecond",
            "minute",
            "nanosecond",
            "petasecond",
            "picosecond",
            "second",
            "terasecond",
            "yoctosecond",
            "yottasecond",
            "zeptosecond",
            "zettasecond"
          ],
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Unit"
    },
    "scale": {
      "default": 1.0,
      "exclusiveMinimum": 0,
      "title": "Scale",
      "type": "number"
    },
    "concatenable": {
      "default": false,
      "description": "If a model has a `concatenable` input axis, it can be processed blockwise,\nsplitting a longer sample axis into blocks matching its input tensor description.\nOutput axes are concatenable if they have a [SizeReference][] to a concatenable\ninput axis.",
      "title": "Concatenable",
      "type": "boolean"
    }
  },
  "required": [
    "size",
    "type"
  ],
  "title": "model.v0_5.TimeInputAxis",
  "type": "object"
}

Fields:

• size (Annotated[Annotated[int, Gt(0)] | ParameterizedSize | SizeReference, Field(examples=[10, ParameterizedSize(min=32, step=16).model_dump(mode='json'), {'tensor_id': 't', 'axis_id': 'a', 'offset': 5}])])
• id (NonBatchAxisId)
• description (Annotated[str, MaxLen(128)])
• type (Literal['time'])
• unit (TimeUnit | None)
• scale (Annotated[float, Gt(0)])
• concatenable (bool)

concatenable pydantic-field

concatenable: bool = False

If a model has a concatenable input axis, it can be processed blockwise, splitting a longer sample axis into blocks matching its input tensor description. Output axes are concatenable if they have a SizeReference to a concatenable input axis.

description pydantic-field

description: Annotated[str, MaxLen(128)] = ''

A short description of this axis beyond its type and id.

id pydantic-field

id: NonBatchAxisId = DEFAULT_TIME_AXIS_ID

An axis id unique across all axes of one tensor.

implemented_type class-attribute

implemented_type: Literal['time'] = 'time'

scale pydantic-field

scale: Annotated[float, Gt(0)] = 1.0

size pydantic-field

size: Annotated[
    Annotated[int, Gt(0)]
    | ParameterizedSize
    | SizeReference,
    Field(
        examples=[
            10,
            ParameterizedSize(min=32, step=16).model_dump(
                mode="json"
            ),
            {"tensor_id": "t", "axis_id": "a", "offset": 5},
        ]
    ),
]

The size/length of this axis can be specified as - fixed integer - parameterized series of valid sizes (ParameterizedSize) - reference to another axis with an optional offset (SizeReference)

type pydantic-field

type: Literal['time'] = 'time'

unit pydantic-field

unit: TimeUnit | None = None

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

TimeOutputAxis pydantic-model

Bases: TimeAxisBase, _WithOutputAxisSize

Show JSON schema:

{
  "$defs": {
    "SizeReference": {
      "additionalProperties": false,
      "description": "A tensor axis size (extent in pixels/frames) defined in relation to a reference axis.\n\n`axis.size = reference.size * reference.scale / axis.scale + offset`\n\nNote:\n1. The axis and the referenced axis need to have the same unit (or no unit).\n2. Batch axes may not be referenced.\n3. Fractions are rounded down.\n4. If the reference axis is `concatenable` the referencing axis is assumed to be\n    `concatenable` as well with the same block order.\n\nExample:\nAn unisotropic input image of w*h=100*49 pixels depicts a phsical space of 200*196mm\u00b2.\nLet's assume that we want to express the image height h in relation to its width w\ninstead of only accepting input images of exactly 100*49 pixels\n(for example to express a range of valid image shapes by parametrizing w, see `ParameterizedSize`).\n\n>>> w = SpaceInputAxis(id=AxisId(\"w\"), size=100, unit=\"millimeter\", scale=2)\n>>> h = SpaceInputAxis(\n...     id=AxisId(\"h\"),\n...     size=SizeReference(tensor_id=TensorId(\"input\"), axis_id=AxisId(\"w\"), offset=-1),\n...     unit=\"millimeter\",\n...     scale=4,\n... )\n>>> print(h.size.get_size(h, w))\n49\n\n\u21d2 h = w * w.scale / h.scale + offset = 100 * 2mm / 4mm - 1 = 49",
      "properties": {
        "tensor_id": {
          "description": "tensor id of the reference axis",
          "maxLength": 32,
          "minLength": 1,
          "title": "TensorId",
          "type": "string"
        },
        "axis_id": {
          "description": "axis id of the reference axis",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "offset": {
          "default": 0,
          "title": "Offset",
          "type": "integer"
        }
      },
      "required": [
        "tensor_id",
        "axis_id"
      ],
      "title": "model.v0_5.SizeReference",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "properties": {
    "size": {
      "anyOf": [
        {
          "exclusiveMinimum": 0,
          "type": "integer"
        },
        {
          "$ref": "#/$defs/SizeReference"
        }
      ],
      "description": "The size/length of this axis can be specified as\n- fixed integer\n- reference to another axis with an optional offset (see [SizeReference][])",
      "examples": [
        10,
        {
          "axis_id": "a",
          "offset": 5,
          "tensor_id": "t"
        }
      ],
      "title": "Size"
    },
    "id": {
      "default": "time",
      "maxLength": 16,
      "minLength": 1,
      "title": "AxisId",
      "type": "string"
    },
    "description": {
      "default": "",
      "description": "A short description of this axis beyond its type and id.",
      "maxLength": 128,
      "title": "Description",
      "type": "string"
    },
    "type": {
      "const": "time",
      "title": "Type",
      "type": "string"
    },
    "unit": {
      "anyOf": [
        {
          "enum": [
            "attosecond",
            "centisecond",
            "day",
            "decisecond",
            "exasecond",
            "femtosecond",
            "gigasecond",
            "hectosecond",
            "hour",
            "kilosecond",
            "megasecond",
            "microsecond",
            "millisecond",
            "minute",
            "nanosecond",
            "petasecond",
            "picosecond",
            "second",
            "terasecond",
            "yoctosecond",
            "yottasecond",
            "zeptosecond",
            "zettasecond"
          ],
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Unit"
    },
    "scale": {
      "default": 1.0,
      "exclusiveMinimum": 0,
      "title": "Scale",
      "type": "number"
    }
  },
  "required": [
    "size",
    "type"
  ],
  "title": "model.v0_5.TimeOutputAxis",
  "type": "object"
}

Fields:

• size (Annotated[Annotated[int, Gt(0)] | SizeReference, Field(examples=[10, {'tensor_id': 't', 'axis_id': 'a', 'offset': 5}])])
• id (NonBatchAxisId)
• description (Annotated[str, MaxLen(128)])
• type (Literal['time'])
• unit (TimeUnit | None)
• scale (Annotated[float, Gt(0)])

description pydantic-field

description: Annotated[str, MaxLen(128)] = ''

A short description of this axis beyond its type and id.

id pydantic-field

id: NonBatchAxisId = DEFAULT_TIME_AXIS_ID

An axis id unique across all axes of one tensor.

implemented_type class-attribute

implemented_type: Literal['time'] = 'time'

scale pydantic-field

scale: Annotated[float, Gt(0)] = 1.0

size pydantic-field

size: Annotated[
    Annotated[int, Gt(0)] | SizeReference,
    Field(
        examples=[
            10,
            {"tensor_id": "t", "axis_id": "a", "offset": 5},
        ]
    ),
]

The size/length of this axis can be specified as - fixed integer - reference to another axis with an optional offset (see SizeReference)

type pydantic-field

type: Literal['time'] = 'time'

unit pydantic-field

unit: TimeUnit | None = None

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

TimeOutputAxisWithHalo pydantic-model

Bases: TimeAxisBase, WithHalo

Show JSON schema:

{
  "$defs": {
    "SizeReference": {
      "additionalProperties": false,
      "description": "A tensor axis size (extent in pixels/frames) defined in relation to a reference axis.\n\n`axis.size = reference.size * reference.scale / axis.scale + offset`\n\nNote:\n1. The axis and the referenced axis need to have the same unit (or no unit).\n2. Batch axes may not be referenced.\n3. Fractions are rounded down.\n4. If the reference axis is `concatenable` the referencing axis is assumed to be\n    `concatenable` as well with the same block order.\n\nExample:\nAn unisotropic input image of w*h=100*49 pixels depicts a phsical space of 200*196mm\u00b2.\nLet's assume that we want to express the image height h in relation to its width w\ninstead of only accepting input images of exactly 100*49 pixels\n(for example to express a range of valid image shapes by parametrizing w, see `ParameterizedSize`).\n\n>>> w = SpaceInputAxis(id=AxisId(\"w\"), size=100, unit=\"millimeter\", scale=2)\n>>> h = SpaceInputAxis(\n...     id=AxisId(\"h\"),\n...     size=SizeReference(tensor_id=TensorId(\"input\"), axis_id=AxisId(\"w\"), offset=-1),\n...     unit=\"millimeter\",\n...     scale=4,\n... )\n>>> print(h.size.get_size(h, w))\n49\n\n\u21d2 h = w * w.scale / h.scale + offset = 100 * 2mm / 4mm - 1 = 49",
      "properties": {
        "tensor_id": {
          "description": "tensor id of the reference axis",
          "maxLength": 32,
          "minLength": 1,
          "title": "TensorId",
          "type": "string"
        },
        "axis_id": {
          "description": "axis id of the reference axis",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "offset": {
          "default": 0,
          "title": "Offset",
          "type": "integer"
        }
      },
      "required": [
        "tensor_id",
        "axis_id"
      ],
      "title": "model.v0_5.SizeReference",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "properties": {
    "halo": {
      "description": "The halo should be cropped from the output tensor to avoid boundary effects.\nIt is to be cropped from both sides, i.e. `size_after_crop = size - 2 * halo`.\nTo document a halo that is already cropped by the model use `size.offset` instead.",
      "minimum": 1,
      "title": "Halo",
      "type": "integer"
    },
    "size": {
      "$ref": "#/$defs/SizeReference",
      "description": "reference to another axis with an optional offset (see [SizeReference][])",
      "examples": [
        {
          "axis_id": "a",
          "offset": 5,
          "tensor_id": "t"
        }
      ]
    },
    "id": {
      "default": "time",
      "maxLength": 16,
      "minLength": 1,
      "title": "AxisId",
      "type": "string"
    },
    "description": {
      "default": "",
      "description": "A short description of this axis beyond its type and id.",
      "maxLength": 128,
      "title": "Description",
      "type": "string"
    },
    "type": {
      "const": "time",
      "title": "Type",
      "type": "string"
    },
    "unit": {
      "anyOf": [
        {
          "enum": [
            "attosecond",
            "centisecond",
            "day",
            "decisecond",
            "exasecond",
            "femtosecond",
            "gigasecond",
            "hectosecond",
            "hour",
            "kilosecond",
            "megasecond",
            "microsecond",
            "millisecond",
            "minute",
            "nanosecond",
            "petasecond",
            "picosecond",
            "second",
            "terasecond",
            "yoctosecond",
            "yottasecond",
            "zeptosecond",
            "zettasecond"
          ],
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Unit"
    },
    "scale": {
      "default": 1.0,
      "exclusiveMinimum": 0,
      "title": "Scale",
      "type": "number"
    }
  },
  "required": [
    "halo",
    "size",
    "type"
  ],
  "title": "model.v0_5.TimeOutputAxisWithHalo",
  "type": "object"
}

Fields:

• halo (Annotated[int, Ge(1)])
• size (Annotated[SizeReference, Field(examples=[{'tensor_id': 't', 'axis_id': 'a', 'offset': 5}])])
• id (NonBatchAxisId)
• description (Annotated[str, MaxLen(128)])
• type (Literal['time'])
• unit (TimeUnit | None)
• scale (Annotated[float, Gt(0)])

description pydantic-field

description: Annotated[str, MaxLen(128)] = ''

A short description of this axis beyond its type and id.

halo pydantic-field

halo: Annotated[int, Ge(1)]

The halo should be cropped from the output tensor to avoid boundary effects. It is to be cropped from both sides, i.e. size_after_crop = size - 2 * halo. To document a halo that is already cropped by the model use size.offset instead.

id pydantic-field

id: NonBatchAxisId = DEFAULT_TIME_AXIS_ID

An axis id unique across all axes of one tensor.

implemented_type class-attribute

implemented_type: Literal['time'] = 'time'

scale pydantic-field

scale: Annotated[float, Gt(0)] = 1.0

size pydantic-field

size: Annotated[
    SizeReference,
    Field(
        examples=[
            {"tensor_id": "t", "axis_id": "a", "offset": 5}
        ]
    ),
]

reference to another axis with an optional offset (see SizeReference)

type pydantic-field

type: Literal['time'] = 'time'

unit pydantic-field

unit: TimeUnit | None = None

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

TorchscriptWeightsDescr pydantic-model

Bases: WeightsEntryDescrBase

Show JSON schema:

{
  "$defs": {
    "Author": {
      "additionalProperties": false,
      "properties": {
        "affiliation": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Affiliation",
          "title": "Affiliation"
        },
        "email": {
          "anyOf": [
            {
              "format": "email",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Email",
          "title": "Email"
        },
        "orcid": {
          "anyOf": [
            {
              "description": "An ORCID identifier, see https://orcid.org/",
              "title": "OrcidId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "An [ORCID iD](https://support.orcid.org/hc/en-us/sections/360001495313-What-is-ORCID\n) in hyphenated groups of 4 digits, (and [valid](\nhttps://support.orcid.org/hc/en-us/articles/360006897674-Structure-of-the-ORCID-Identifier\n) as per ISO 7064 11,2.)",
          "examples": [
            "0000-0001-2345-6789"
          ],
          "title": "Orcid"
        },
        "name": {
          "title": "Name",
          "type": "string"
        },
        "github_user": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Github User"
        }
      },
      "required": [
        "name"
      ],
      "title": "generic.v0_3.Author",
      "type": "object"
    },
    "RelativeFilePath": {
      "description": "A path relative to the `rdf.yaml` file (also if the RDF source is a URL).",
      "format": "path",
      "title": "RelativeFilePath",
      "type": "string"
    },
    "Version": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        }
      ],
      "description": "wraps a packaging.version.Version instance for validation in pydantic models",
      "title": "Version"
    }
  },
  "additionalProperties": false,
  "properties": {
    "source": {
      "anyOf": [
        {
          "description": "A URL with the HTTP or HTTPS scheme.",
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "title": "HttpUrl",
          "type": "string"
        },
        {
          "$ref": "#/$defs/RelativeFilePath"
        },
        {
          "format": "file-path",
          "title": "FilePath",
          "type": "string"
        }
      ],
      "description": "Source of the weights file.",
      "title": "FileSource"
    },
    "sha256": {
      "anyOf": [
        {
          "description": "A SHA-256 hash value",
          "maxLength": 64,
          "minLength": 64,
          "title": "Sha256",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "SHA256 hash value of the **source** file.",
      "title": "Sha256"
    },
    "authors": {
      "anyOf": [
        {
          "items": {
            "$ref": "#/$defs/Author"
          },
          "type": "array"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
      "title": "Authors"
    },
    "parent": {
      "anyOf": [
        {
          "enum": [
            "keras_hdf5",
            "keras_v3",
            "onnx",
            "pytorch_state_dict",
            "tensorflow_js",
            "tensorflow_saved_model_bundle",
            "torchscript"
          ],
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
      "examples": [
        "pytorch_state_dict"
      ],
      "title": "Parent"
    },
    "comment": {
      "default": "",
      "description": "A comment about this weights entry, for example how these weights were created.",
      "title": "Comment",
      "type": "string"
    },
    "pytorch_version": {
      "$ref": "#/$defs/Version",
      "description": "Version of the PyTorch library used."
    }
  },
  "required": [
    "source",
    "pytorch_version"
  ],
  "title": "model.v0_5.TorchscriptWeightsDescr",
  "type": "object"
}

Fields:

• source (Annotated[FileSource, AfterValidator(wo_special_file_name)])
• sha256 (Sha256 | None)
• authors (list[Author] | None)
• parent (Annotated[WeightsFormat | None, Field(examples=['pytorch_state_dict'])])
• comment (str)
• pytorch_version (Version)

Validators:

• _validate

authors pydantic-field

authors: list[Author] | None = None

Authors Either the person(s) that have trained this model resulting in the original weights file. (If this is the initial weights entry, i.e. it does not have a parent) Or the person(s) who have converted the weights to this weights format. (If this is a child weight, i.e. it has a parent field)

comment pydantic-field

comment: str = ''

A comment about this weights entry, for example how these weights were created.

parent pydantic-field

parent: Annotated[
    WeightsFormat | None,
    Field(examples=["pytorch_state_dict"]),
] = None

The source weights these weights were converted from. For example, if a model's weights were converted from the pytorch_state_dict format to torchscript, The pytorch_state_dict weights entry has no parent and is the parent of the torchscript weights. All weight entries except one (the initial set of weights resulting from training the model), need to have this field.

pytorch_version pydantic-field

pytorch_version: Version

Version of the PyTorch library used.

sha256 pydantic-field

sha256: Sha256 | None = None

SHA256 hash value of the source file.

source pydantic-field

source: Annotated[
    FileSource, AfterValidator(wo_special_file_name)
]

Source of the weights file.

suffix property

suffix: str

type class-attribute

type: WeightsFormat = 'torchscript'

weights_format_name class-attribute

weights_format_name: str = 'TorchScript'

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

download

download(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

alias for .get_reader

Source code in src/bioimageio/spec/_internal/io.py

def download(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """alias for `.get_reader`"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

get_reader

get_reader(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

open the file source (download if needed)

Source code in src/bioimageio/spec/_internal/io.py

def get_reader(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """open the file source (download if needed)"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

validate_sha256

validate_sha256(force_recompute: bool = False) -> None

validate the sha256 hash value of the source file

Source code in src/bioimageio/spec/_internal/io.py

def validate_sha256(self, force_recompute: bool = False) -> None:
    """validate the sha256 hash value of the **source** file"""
    context = get_validation_context()
    src_str = str(self.source)
    if force_recompute:
        actual_sha = None
    else:
        actual_sha = context.known_files.get(src_str)

    if actual_sha is None:
        if context.perform_io_checks or force_recompute:
            reader = get_reader(self.source, sha256=self.sha256)
            if force_recompute:
                actual_sha = get_sha256(reader)
            else:
                actual_sha = reader.sha256

            context.known_files[src_str] = actual_sha
        elif context.known_files and src_str not in context.known_files:
            # perform_io_checks is False, but known files were given,
            # so we expect all file references to be in there
            raise ValueError(f"File {src_str} not found in `known_files`.")

    if actual_sha is None or self.sha256 == actual_sha:
        return
    elif self.sha256 is None or context.update_hashes:
        self.sha256 = actual_sha
    elif self.sha256 != actual_sha:
        raise ValueError(
            f"Sha256 mismatch for {self.source}. Expected {self.sha256}, got "
            + f"{actual_sha}. Update expected `sha256` or point to the matching "
            + "file."
        )

TrainingDetails pydantic-model

Bases: Node

Show JSON schema:

{
  "$defs": {
    "YamlValue": {
      "anyOf": [
        {
          "type": "boolean"
        },
        {
          "format": "date",
          "type": "string"
        },
        {
          "format": "date-time",
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        },
        {
          "type": "string"
        },
        {
          "items": {
            "$ref": "#/$defs/YamlValue"
          },
          "type": "array"
        },
        {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "type": "object"
        },
        {
          "type": "null"
        }
      ]
    }
  },
  "additionalProperties": true,
  "properties": {
    "training_preprocessing": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Detailed image preprocessing steps during model training:\n\nMention:\n- *Normalization methods*\n- *Augmentation strategies*\n- *Resizing/resampling procedures*\n- *Artifact handling*",
      "title": "Training Preprocessing"
    },
    "training_epochs": {
      "anyOf": [
        {
          "type": "number"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Number of training epochs.",
      "title": "Training Epochs"
    },
    "training_batch_size": {
      "anyOf": [
        {
          "type": "number"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Batch size used in training.",
      "title": "Training Batch Size"
    },
    "initial_learning_rate": {
      "anyOf": [
        {
          "type": "number"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Initial learning rate used in training.",
      "title": "Initial Learning Rate"
    },
    "learning_rate_schedule": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Learning rate schedule used in training.",
      "title": "Learning Rate Schedule"
    },
    "loss_function": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Loss function used in training, e.g. nn.MSELoss.",
      "title": "Loss Function"
    },
    "loss_function_kwargs": {
      "additionalProperties": {
        "$ref": "#/$defs/YamlValue"
      },
      "description": "key word arguments for the `loss_function`",
      "title": "Loss Function Kwargs",
      "type": "object"
    },
    "optimizer": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "optimizer, e.g. torch.optim.Adam",
      "title": "Optimizer"
    },
    "optimizer_kwargs": {
      "additionalProperties": {
        "$ref": "#/$defs/YamlValue"
      },
      "description": "key word arguments for the `optimizer`",
      "title": "Optimizer Kwargs",
      "type": "object"
    },
    "regularization": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Regularization techniques used during training, e.g. drop-out or weight decay.",
      "title": "Regularization"
    },
    "training_duration": {
      "anyOf": [
        {
          "type": "number"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Total training duration in hours.",
      "title": "Training Duration"
    }
  },
  "title": "model.v0_5.TrainingDetails",
  "type": "object"
}

Fields:

• training_preprocessing (str | None)
• training_epochs (float | None)
• training_batch_size (float | None)
• initial_learning_rate (float | None)
• learning_rate_schedule (str | None)
• loss_function (str | None)
• loss_function_kwargs (dict[str, YamlValue])
• optimizer (str | None)
• optimizer_kwargs (dict[str, YamlValue])
• regularization (str | None)
• training_duration (float | None)

initial_learning_rate pydantic-field

initial_learning_rate: float | None = None

Initial learning rate used in training.

learning_rate_schedule pydantic-field

learning_rate_schedule: str | None = None

Learning rate schedule used in training.

loss_function pydantic-field

loss_function: str | None = None

Loss function used in training, e.g. nn.MSELoss.

loss_function_kwargs pydantic-field

loss_function_kwargs: dict[str, YamlValue]

key word arguments for the loss_function

optimizer pydantic-field

optimizer: str | None = None

optimizer, e.g. torch.optim.Adam

optimizer_kwargs pydantic-field

optimizer_kwargs: dict[str, YamlValue]

key word arguments for the optimizer

regularization pydantic-field

regularization: str | None = None

Regularization techniques used during training, e.g. drop-out or weight decay.

training_batch_size pydantic-field

training_batch_size: float | None = None

Batch size used in training.

training_duration pydantic-field

training_duration: float | None = None

Total training duration in hours.

training_epochs pydantic-field

training_epochs: float | None = None

Number of training epochs.

training_preprocessing pydantic-field

training_preprocessing: str | None = None

Detailed image preprocessing steps during model training:

Mention: - Normalization methods - Augmentation strategies - Resizing/resampling procedures - Artifact handling

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

Uploader pydantic-model

Bases: Node

Show JSON schema:

{
  "additionalProperties": false,
  "properties": {
    "email": {
      "description": "Email",
      "format": "email",
      "title": "Email",
      "type": "string"
    },
    "name": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "name",
      "title": "Name"
    }
  },
  "required": [
    "email"
  ],
  "title": "generic.v0_2.Uploader",
  "type": "object"
}

Fields:

• email (EmailStr)
• name (Annotated[str, AfterValidator(_remove_slashes)] | None)

email pydantic-field

email: EmailStr

Email

name pydantic-field

name: (
    Annotated[str, AfterValidator(_remove_slashes)] | None
) = None

name

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

Version

Bases: RootModel[Union[str, int, float]]

              flowchart TD
              bioimageio.spec.model.v0_5.Version[Version]

              

              click bioimageio.spec.model.v0_5.Version href "" "bioimageio.spec.model.v0_5.Version"
            

wraps a packaging.version.Version instance for validation in pydantic models

Methods:

Name	Description
__eq__
__ge__
__le__
__lt__
__str__
model_post_init	set _version attribute @private

Attributes:

Name	Type	Description
base_version	str	The "base version" of the version.
dev	int | None	The development number of the version.
epoch	int	The epoch of the version.
is_devrelease	bool	Whether this version is a development release.
is_postrelease	bool	Whether this version is a post-release.
is_prerelease	bool	Whether this version is a pre-release.
local	str | None	The local version segment of the version.
major	int	The first item of :attr:release or 0 if unavailable.
micro	int	The third item of :attr:release or 0 if unavailable.
minor	int	The second item of :attr:release or 0 if unavailable.
post	int | None	The post-release number of the version.
pre	tuple[str, int] | None	The pre-release segment of the version.
public	str	The public portion of the version.
release	tuple[int, ...]	The components of the "release" segment of the version.

base_version property

base_version: str

The "base version" of the version.

>>> Version("1.2.3").base_version
'1.2.3'
>>> Version("1.2.3+abc").base_version
'1.2.3'
>>> Version("1!1.2.3+abc.dev1").base_version
'1!1.2.3'

The "base version" is the public version of the project without any pre or post release markers.

dev property

dev: int | None

The development number of the version.

>>> print(Version("1.2.3").dev)
None
>>> Version("1.2.3.dev1").dev
1

epoch property

epoch: int

The epoch of the version.

>>> Version("2.0.0").epoch
0
>>> Version("1!2.0.0").epoch
1

is_devrelease property

is_devrelease: bool

Whether this version is a development release.

>>> Version("1.2.3").is_devrelease
False
>>> Version("1.2.3.dev1").is_devrelease
True

is_postrelease property

is_postrelease: bool

Whether this version is a post-release.

>>> Version("1.2.3").is_postrelease
False
>>> Version("1.2.3.post1").is_postrelease
True

is_prerelease property

is_prerelease: bool

Whether this version is a pre-release.

>>> Version("1.2.3").is_prerelease
False
>>> Version("1.2.3a1").is_prerelease
True
>>> Version("1.2.3b1").is_prerelease
True
>>> Version("1.2.3rc1").is_prerelease
True
>>> Version("1.2.3dev1").is_prerelease
True

local property

local: str | None

The local version segment of the version.

>>> print(Version("1.2.3").local)
None
>>> Version("1.2.3+abc").local
'abc'

major property

major: int

The first item of :attr:release or 0 if unavailable.

>>> Version("1.2.3").major
1

micro property

micro: int

The third item of :attr:release or 0 if unavailable.

>>> Version("1.2.3").micro
3
>>> Version("1").micro
0

minor property

minor: int

The second item of :attr:release or 0 if unavailable.

>>> Version("1.2.3").minor
2
>>> Version("1").minor
0

post property

post: int | None

The post-release number of the version.

>>> print(Version("1.2.3").post)
None
>>> Version("1.2.3.post1").post
1

pre property

pre: tuple[str, int] | None

The pre-release segment of the version.

>>> print(Version("1.2.3").pre)
None
>>> Version("1.2.3a1").pre
('a', 1)
>>> Version("1.2.3b1").pre
('b', 1)
>>> Version("1.2.3rc1").pre
('rc', 1)

public property

public: str

The public portion of the version.

>>> Version("1.2.3").public
'1.2.3'
>>> Version("1.2.3+abc").public
'1.2.3'
>>> Version("1.2.3+abc.dev1").public
'1.2.3'

release property

release: tuple[int, ...]

The components of the "release" segment of the version.

>>> Version("1.2.3").release
(1, 2, 3)
>>> Version("2.0.0").release
(2, 0, 0)
>>> Version("1!2.0.0.post0").release
(2, 0, 0)

Includes trailing zeroes but not the epoch or any pre-release / development / post-release suffixes.

__eq__

__eq__(other: object)

Source code in src/bioimageio/spec/_internal/version_type.py

def __eq__(self, other: object):
    if not isinstance(other, Version):
        return NotImplemented
    return self._version == other._version

__ge__

__ge__(other: Any)

Source code in src/bioimageio/spec/_internal/version_type.py

def __ge__(self, other: Any):
    if not isinstance(other, Version):
        return NotImplemented
    return self._version >= other._version

__le__

__le__(other: Any)

Source code in src/bioimageio/spec/_internal/version_type.py

def __le__(self, other: Any):
    if not isinstance(other, Version):
        return NotImplemented
    return self._version <= other._version

__lt__

__lt__(other: Any)

Source code in src/bioimageio/spec/_internal/version_type.py

def __lt__(self, other: Any):
    if not isinstance(other, Version):
        return NotImplemented

    return self._version < other._version

__str__

__str__()

Source code in src/bioimageio/spec/_internal/version_type.py

def __str__(self):
    return str(self._version)

model_post_init

model_post_init(__context: Any) -> None

set _version attribute @private

Source code in src/bioimageio/spec/_internal/version_type.py

def model_post_init(self, __context: Any, /) -> None:
    """set `_version` attribute @private"""
    self._version = packaging.version.Version(str(self.root))
    return super().model_post_init(__context)

WeightsDescr pydantic-model

Bases: Node

Show JSON schema:

{
  "$defs": {
    "ArchitectureFromFileDescr": {
      "additionalProperties": false,
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "Architecture source file",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        },
        "callable": {
          "description": "Identifier of the callable that returns a torch.nn.Module instance.",
          "examples": [
            "MyNetworkClass",
            "get_my_model"
          ],
          "minLength": 1,
          "title": "Identifier",
          "type": "string"
        },
        "kwargs": {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "description": "key word arguments for the `callable`",
          "title": "Kwargs",
          "type": "object"
        }
      },
      "required": [
        "source",
        "callable"
      ],
      "title": "model.v0_5.ArchitectureFromFileDescr",
      "type": "object"
    },
    "ArchitectureFromLibraryDescr": {
      "additionalProperties": false,
      "properties": {
        "callable": {
          "description": "Identifier of the callable that returns a torch.nn.Module instance.",
          "examples": [
            "MyNetworkClass",
            "get_my_model"
          ],
          "minLength": 1,
          "title": "Identifier",
          "type": "string"
        },
        "kwargs": {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "description": "key word arguments for the `callable`",
          "title": "Kwargs",
          "type": "object"
        },
        "import_from": {
          "description": "Where to import the callable from, i.e. `from <import_from> import <callable>`",
          "title": "Import From",
          "type": "string"
        }
      },
      "required": [
        "callable",
        "import_from"
      ],
      "title": "model.v0_5.ArchitectureFromLibraryDescr",
      "type": "object"
    },
    "Author": {
      "additionalProperties": false,
      "properties": {
        "affiliation": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Affiliation",
          "title": "Affiliation"
        },
        "email": {
          "anyOf": [
            {
              "format": "email",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Email",
          "title": "Email"
        },
        "orcid": {
          "anyOf": [
            {
              "description": "An ORCID identifier, see https://orcid.org/",
              "title": "OrcidId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "An [ORCID iD](https://support.orcid.org/hc/en-us/sections/360001495313-What-is-ORCID\n) in hyphenated groups of 4 digits, (and [valid](\nhttps://support.orcid.org/hc/en-us/articles/360006897674-Structure-of-the-ORCID-Identifier\n) as per ISO 7064 11,2.)",
          "examples": [
            "0000-0001-2345-6789"
          ],
          "title": "Orcid"
        },
        "name": {
          "title": "Name",
          "type": "string"
        },
        "github_user": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Github User"
        }
      },
      "required": [
        "name"
      ],
      "title": "generic.v0_3.Author",
      "type": "object"
    },
    "FileDescr": {
      "additionalProperties": false,
      "description": "A file description",
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "File source",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        }
      },
      "required": [
        "source"
      ],
      "title": "_internal.io.FileDescr",
      "type": "object"
    },
    "KerasHdf5WeightsDescr": {
      "additionalProperties": false,
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "Source of the weights file.",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        },
        "authors": {
          "anyOf": [
            {
              "items": {
                "$ref": "#/$defs/Author"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
          "title": "Authors"
        },
        "parent": {
          "anyOf": [
            {
              "enum": [
                "keras_hdf5",
                "keras_v3",
                "onnx",
                "pytorch_state_dict",
                "tensorflow_js",
                "tensorflow_saved_model_bundle",
                "torchscript"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
          "examples": [
            "pytorch_state_dict"
          ],
          "title": "Parent"
        },
        "comment": {
          "default": "",
          "description": "A comment about this weights entry, for example how these weights were created.",
          "title": "Comment",
          "type": "string"
        },
        "tensorflow_version": {
          "$ref": "#/$defs/Version",
          "description": "TensorFlow version used to create these weights."
        }
      },
      "required": [
        "source",
        "tensorflow_version"
      ],
      "title": "model.v0_5.KerasHdf5WeightsDescr",
      "type": "object"
    },
    "KerasV3WeightsDescr": {
      "additionalProperties": false,
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "Source of the .keras weights file.",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        },
        "authors": {
          "anyOf": [
            {
              "items": {
                "$ref": "#/$defs/Author"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
          "title": "Authors"
        },
        "parent": {
          "anyOf": [
            {
              "enum": [
                "keras_hdf5",
                "keras_v3",
                "onnx",
                "pytorch_state_dict",
                "tensorflow_js",
                "tensorflow_saved_model_bundle",
                "torchscript"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
          "examples": [
            "pytorch_state_dict"
          ],
          "title": "Parent"
        },
        "comment": {
          "default": "",
          "description": "A comment about this weights entry, for example how these weights were created.",
          "title": "Comment",
          "type": "string"
        },
        "keras_version": {
          "$ref": "#/$defs/Version",
          "description": "Keras version used to create these weights.",
          "ge": 3
        },
        "backend": {
          "description": "Keras backend used to create these weights.",
          "maxItems": 2,
          "minItems": 2,
          "prefixItems": [
            {
              "enum": [
                "tensorflow",
                "jax",
                "torch"
              ],
              "type": "string"
            },
            {
              "$ref": "#/$defs/Version"
            }
          ],
          "title": "Backend",
          "type": "array"
        }
      },
      "required": [
        "source",
        "keras_version",
        "backend"
      ],
      "title": "model.v0_5.KerasV3WeightsDescr",
      "type": "object"
    },
    "OnnxWeightsDescr": {
      "additionalProperties": false,
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "Source of the weights file.",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        },
        "authors": {
          "anyOf": [
            {
              "items": {
                "$ref": "#/$defs/Author"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
          "title": "Authors"
        },
        "parent": {
          "anyOf": [
            {
              "enum": [
                "keras_hdf5",
                "keras_v3",
                "onnx",
                "pytorch_state_dict",
                "tensorflow_js",
                "tensorflow_saved_model_bundle",
                "torchscript"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
          "examples": [
            "pytorch_state_dict"
          ],
          "title": "Parent"
        },
        "comment": {
          "default": "",
          "description": "A comment about this weights entry, for example how these weights were created.",
          "title": "Comment",
          "type": "string"
        },
        "opset_version": {
          "description": "ONNX opset version",
          "minimum": 7,
          "title": "Opset Version",
          "type": "integer"
        },
        "external_data": {
          "anyOf": [
            {
              "$ref": "#/$defs/FileDescr",
              "examples": [
                {
                  "source": "weights.onnx.data"
                }
              ]
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Source of the external ONNX data file holding the weights.\n(If present **source** holds the ONNX architecture without weights)."
        }
      },
      "required": [
        "source",
        "opset_version"
      ],
      "title": "model.v0_5.OnnxWeightsDescr",
      "type": "object"
    },
    "PytorchStateDictWeightsDescr": {
      "additionalProperties": false,
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "Source of the weights file.",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        },
        "authors": {
          "anyOf": [
            {
              "items": {
                "$ref": "#/$defs/Author"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
          "title": "Authors"
        },
        "parent": {
          "anyOf": [
            {
              "enum": [
                "keras_hdf5",
                "keras_v3",
                "onnx",
                "pytorch_state_dict",
                "tensorflow_js",
                "tensorflow_saved_model_bundle",
                "torchscript"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
          "examples": [
            "pytorch_state_dict"
          ],
          "title": "Parent"
        },
        "comment": {
          "default": "",
          "description": "A comment about this weights entry, for example how these weights were created.",
          "title": "Comment",
          "type": "string"
        },
        "architecture": {
          "anyOf": [
            {
              "$ref": "#/$defs/ArchitectureFromFileDescr"
            },
            {
              "$ref": "#/$defs/ArchitectureFromLibraryDescr"
            }
          ],
          "title": "Architecture"
        },
        "pytorch_version": {
          "$ref": "#/$defs/Version",
          "description": "Version of the PyTorch library used.\nIf `architecture.depencencies` is specified it has to include pytorch and any version pinning has to be compatible."
        },
        "dependencies": {
          "anyOf": [
            {
              "$ref": "#/$defs/FileDescr",
              "examples": [
                {
                  "source": "environment.yaml"
                }
              ]
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Custom depencies beyond pytorch described in a Conda environment file.\nAllows to specify custom dependencies, see conda docs:\n- [Exporting an environment file across platforms](https://conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html#exporting-an-environment-file-across-platforms)\n- [Creating an environment file manually](https://conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html#creating-an-environment-file-manually)\n\nThe conda environment file should include pytorch and any version pinning has to be compatible with\n**pytorch_version**."
        },
        "strict": {
          "default": true,
          "description": "Whether to allow missing or unexpected keys or to be strict about the architecture matching the state dict weights.",
          "title": "Strict",
          "type": "boolean"
        }
      },
      "required": [
        "source",
        "architecture",
        "pytorch_version"
      ],
      "title": "model.v0_5.PytorchStateDictWeightsDescr",
      "type": "object"
    },
    "RelativeFilePath": {
      "description": "A path relative to the `rdf.yaml` file (also if the RDF source is a URL).",
      "format": "path",
      "title": "RelativeFilePath",
      "type": "string"
    },
    "TensorflowJsWeightsDescr": {
      "additionalProperties": false,
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "The multi-file weights.\nAll required files/folders should be a zip archive.",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        },
        "authors": {
          "anyOf": [
            {
              "items": {
                "$ref": "#/$defs/Author"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
          "title": "Authors"
        },
        "parent": {
          "anyOf": [
            {
              "enum": [
                "keras_hdf5",
                "keras_v3",
                "onnx",
                "pytorch_state_dict",
                "tensorflow_js",
                "tensorflow_saved_model_bundle",
                "torchscript"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
          "examples": [
            "pytorch_state_dict"
          ],
          "title": "Parent"
        },
        "comment": {
          "default": "",
          "description": "A comment about this weights entry, for example how these weights were created.",
          "title": "Comment",
          "type": "string"
        },
        "tensorflow_version": {
          "$ref": "#/$defs/Version",
          "description": "Version of the TensorFlow library used."
        }
      },
      "required": [
        "source",
        "tensorflow_version"
      ],
      "title": "model.v0_5.TensorflowJsWeightsDescr",
      "type": "object"
    },
    "TensorflowSavedModelBundleWeightsDescr": {
      "additionalProperties": false,
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "The multi-file weights.\nAll required files/folders should be a zip archive.",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        },
        "authors": {
          "anyOf": [
            {
              "items": {
                "$ref": "#/$defs/Author"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
          "title": "Authors"
        },
        "parent": {
          "anyOf": [
            {
              "enum": [
                "keras_hdf5",
                "keras_v3",
                "onnx",
                "pytorch_state_dict",
                "tensorflow_js",
                "tensorflow_saved_model_bundle",
                "torchscript"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
          "examples": [
            "pytorch_state_dict"
          ],
          "title": "Parent"
        },
        "comment": {
          "default": "",
          "description": "A comment about this weights entry, for example how these weights were created.",
          "title": "Comment",
          "type": "string"
        },
        "tensorflow_version": {
          "$ref": "#/$defs/Version",
          "description": "Version of the TensorFlow library used."
        },
        "dependencies": {
          "anyOf": [
            {
              "$ref": "#/$defs/FileDescr",
              "examples": [
                {
                  "source": "environment.yaml"
                }
              ]
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Custom dependencies beyond tensorflow.\nShould include tensorflow and any version pinning has to be compatible with **tensorflow_version**."
        }
      },
      "required": [
        "source",
        "tensorflow_version"
      ],
      "title": "model.v0_5.TensorflowSavedModelBundleWeightsDescr",
      "type": "object"
    },
    "TorchscriptWeightsDescr": {
      "additionalProperties": false,
      "properties": {
        "source": {
          "anyOf": [
            {
              "description": "A URL with the HTTP or HTTPS scheme.",
              "format": "uri",
              "maxLength": 2083,
              "minLength": 1,
              "title": "HttpUrl",
              "type": "string"
            },
            {
              "$ref": "#/$defs/RelativeFilePath"
            },
            {
              "format": "file-path",
              "title": "FilePath",
              "type": "string"
            }
          ],
          "description": "Source of the weights file.",
          "title": "FileSource"
        },
        "sha256": {
          "anyOf": [
            {
              "description": "A SHA-256 hash value",
              "maxLength": 64,
              "minLength": 64,
              "title": "Sha256",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "SHA256 hash value of the **source** file.",
          "title": "Sha256"
        },
        "authors": {
          "anyOf": [
            {
              "items": {
                "$ref": "#/$defs/Author"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
          "title": "Authors"
        },
        "parent": {
          "anyOf": [
            {
              "enum": [
                "keras_hdf5",
                "keras_v3",
                "onnx",
                "pytorch_state_dict",
                "tensorflow_js",
                "tensorflow_saved_model_bundle",
                "torchscript"
              ],
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
          "examples": [
            "pytorch_state_dict"
          ],
          "title": "Parent"
        },
        "comment": {
          "default": "",
          "description": "A comment about this weights entry, for example how these weights were created.",
          "title": "Comment",
          "type": "string"
        },
        "pytorch_version": {
          "$ref": "#/$defs/Version",
          "description": "Version of the PyTorch library used."
        }
      },
      "required": [
        "source",
        "pytorch_version"
      ],
      "title": "model.v0_5.TorchscriptWeightsDescr",
      "type": "object"
    },
    "Version": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        }
      ],
      "description": "wraps a packaging.version.Version instance for validation in pydantic models",
      "title": "Version"
    },
    "YamlValue": {
      "anyOf": [
        {
          "type": "boolean"
        },
        {
          "format": "date",
          "type": "string"
        },
        {
          "format": "date-time",
          "type": "string"
        },
        {
          "type": "integer"
        },
        {
          "type": "number"
        },
        {
          "type": "string"
        },
        {
          "items": {
            "$ref": "#/$defs/YamlValue"
          },
          "type": "array"
        },
        {
          "additionalProperties": {
            "$ref": "#/$defs/YamlValue"
          },
          "type": "object"
        },
        {
          "type": "null"
        }
      ]
    }
  },
  "additionalProperties": false,
  "properties": {
    "keras_hdf5": {
      "anyOf": [
        {
          "$ref": "#/$defs/KerasHdf5WeightsDescr"
        },
        {
          "type": "null"
        }
      ],
      "default": null
    },
    "keras_v3": {
      "anyOf": [
        {
          "$ref": "#/$defs/KerasV3WeightsDescr"
        },
        {
          "type": "null"
        }
      ],
      "default": null
    },
    "onnx": {
      "anyOf": [
        {
          "$ref": "#/$defs/OnnxWeightsDescr"
        },
        {
          "type": "null"
        }
      ],
      "default": null
    },
    "pytorch_state_dict": {
      "anyOf": [
        {
          "$ref": "#/$defs/PytorchStateDictWeightsDescr"
        },
        {
          "type": "null"
        }
      ],
      "default": null
    },
    "tensorflow_js": {
      "anyOf": [
        {
          "$ref": "#/$defs/TensorflowJsWeightsDescr"
        },
        {
          "type": "null"
        }
      ],
      "default": null
    },
    "tensorflow_saved_model_bundle": {
      "anyOf": [
        {
          "$ref": "#/$defs/TensorflowSavedModelBundleWeightsDescr"
        },
        {
          "type": "null"
        }
      ],
      "default": null
    },
    "torchscript": {
      "anyOf": [
        {
          "$ref": "#/$defs/TorchscriptWeightsDescr"
        },
        {
          "type": "null"
        }
      ],
      "default": null
    }
  },
  "title": "model.v0_5.WeightsDescr",
  "type": "object"
}

Fields:

• keras_hdf5 (KerasHdf5WeightsDescr | None)
• keras_v3 (KerasV3WeightsDescr | None)
• onnx (OnnxWeightsDescr | None)
• pytorch_state_dict (PytorchStateDictWeightsDescr | None)
• tensorflow_js (TensorflowJsWeightsDescr | None)
• tensorflow_saved_model_bundle (TensorflowSavedModelBundleWeightsDescr | None)
• torchscript (TorchscriptWeightsDescr | None)

Validators:

• check_entries

available_formats property

available_formats: dict[WeightsFormat, SpecificWeightsDescr]

keras_hdf5 pydantic-field

keras_hdf5: KerasHdf5WeightsDescr | None = None

keras_v3 pydantic-field

keras_v3: KerasV3WeightsDescr | None = None

missing_formats property

missing_formats: set[WeightsFormat]

onnx pydantic-field

onnx: OnnxWeightsDescr | None = None

pytorch_state_dict pydantic-field

pytorch_state_dict: PytorchStateDictWeightsDescr | None = (
    None
)

tensorflow_js pydantic-field

tensorflow_js: TensorflowJsWeightsDescr | None = None

tensorflow_saved_model_bundle pydantic-field

tensorflow_saved_model_bundle: (
    TensorflowSavedModelBundleWeightsDescr | None
) = None

torchscript pydantic-field

torchscript: TorchscriptWeightsDescr | None = None

__getitem__

__getitem__(key: WeightsFormat)

Source code in src/bioimageio/spec/model/v0_5.py

def __getitem__(
    self,
    key: WeightsFormat,
):
    if key == "keras_hdf5":
        ret = self.keras_hdf5
    elif key == "keras_v3":
        ret = self.keras_v3
    elif key == "onnx":
        ret = self.onnx
    elif key == "pytorch_state_dict":
        ret = self.pytorch_state_dict
    elif key == "tensorflow_js":
        ret = self.tensorflow_js
    elif key == "tensorflow_saved_model_bundle":
        ret = self.tensorflow_saved_model_bundle
    elif key == "torchscript":
        ret = self.torchscript
    else:
        raise KeyError(key)

    if ret is None:
        raise KeyError(key)

    return ret

__setitem__

__setitem__(
    key: Literal["keras_hdf5"],
    value: KerasHdf5WeightsDescr | None,
) -> None

__setitem__(
    key: Literal["keras_v3"],
    value: KerasV3WeightsDescr | None,
) -> None

__setitem__(
    key: Literal["onnx"], value: OnnxWeightsDescr | None
) -> None

__setitem__(
    key: Literal["pytorch_state_dict"],
    value: PytorchStateDictWeightsDescr | None,
) -> None

__setitem__(
    key: Literal["tensorflow_js"],
    value: TensorflowJsWeightsDescr | None,
) -> None

__setitem__(
    key: Literal["tensorflow_saved_model_bundle"],
    value: TensorflowSavedModelBundleWeightsDescr | None,
) -> None

__setitem__(
    key: Literal["torchscript"],
    value: TorchscriptWeightsDescr | None,
) -> None

__setitem__(
    key: WeightsFormat, value: SpecificWeightsDescr | None
)

Source code in src/bioimageio/spec/model/v0_5.py

def __setitem__(
    self,
    key: WeightsFormat,
    value: SpecificWeightsDescr | None,
):
    if key == "keras_hdf5":
        if value is not None and not isinstance(value, KerasHdf5WeightsDescr):
            raise TypeError(
                f"Expected KerasHdf5WeightsDescr or None for key 'keras_hdf5', got {type(value)}"
            )
        self.keras_hdf5 = value
    elif key == "keras_v3":
        if value is not None and not isinstance(value, KerasV3WeightsDescr):
            raise TypeError(
                f"Expected KerasV3WeightsDescr or None for key 'keras_v3', got {type(value)}"
            )
        self.keras_v3 = value
    elif key == "onnx":
        if value is not None and not isinstance(value, OnnxWeightsDescr):
            raise TypeError(
                f"Expected OnnxWeightsDescr or None for key 'onnx', got {type(value)}"
            )
        self.onnx = value
    elif key == "pytorch_state_dict":
        if value is not None and not isinstance(
            value, PytorchStateDictWeightsDescr
        ):
            raise TypeError(
                f"Expected PytorchStateDictWeightsDescr or None for key 'pytorch_state_dict', got {type(value)}"
            )
        self.pytorch_state_dict = value
    elif key == "tensorflow_js":
        if value is not None and not isinstance(value, TensorflowJsWeightsDescr):
            raise TypeError(
                f"Expected TensorflowJsWeightsDescr or None for key 'tensorflow_js', got {type(value)}"
            )
        self.tensorflow_js = value
    elif key == "tensorflow_saved_model_bundle":
        if value is not None and not isinstance(
            value, TensorflowSavedModelBundleWeightsDescr
        ):
            raise TypeError(
                f"Expected TensorflowSavedModelBundleWeightsDescr or None for key 'tensorflow_saved_model_bundle', got {type(value)}"
            )
        self.tensorflow_saved_model_bundle = value
    elif key == "torchscript":
        if value is not None and not isinstance(value, TorchscriptWeightsDescr):
            raise TypeError(
                f"Expected TorchscriptWeightsDescr or None for key 'torchscript', got {type(value)}"
            )
        self.torchscript = value
    else:
        raise KeyError(key)

check_entries pydantic-validator

check_entries() -> Self

Source code in src/bioimageio/spec/model/v0_5.py

@model_validator(mode="after")
def check_entries(self) -> Self:
    entries = {wtype for wtype, entry in self if entry is not None}

    if not entries:
        raise ValueError("Missing weights entry")

    entries_wo_parent = {
        wtype
        for wtype, entry in self
        if entry is not None and hasattr(entry, "parent") and entry.parent is None
    }
    if len(entries_wo_parent) != 1:
        issue_warning(
            "Exactly one weights entry may not specify the `parent` field (got"
            + " {value}). That entry is considered the original set of model weights."
            + " Other weight formats are created through conversion of the orignal or"
            + " already converted weights. They have to reference the weights format"
            + " they were converted from as their `parent`.",
            value=len(entries_wo_parent),
            field="weights",
        )

    for wtype, entry in self:
        if entry is None:
            continue

        assert hasattr(entry, "type")
        assert hasattr(entry, "parent")
        assert wtype == entry.type
        if (
            entry.parent is not None and entry.parent not in entries
        ):  # self reference checked for `parent` field
            raise ValueError(
                f"`weights.{wtype}.parent={entry.parent} not in specified weight"
                + f" formats: {entries}"
            )

    return self

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

WeightsEntryDescrBase pydantic-model

Bases: FileDescr

Show JSON schema:

{
  "$defs": {
    "Author": {
      "additionalProperties": false,
      "properties": {
        "affiliation": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Affiliation",
          "title": "Affiliation"
        },
        "email": {
          "anyOf": [
            {
              "format": "email",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Email",
          "title": "Email"
        },
        "orcid": {
          "anyOf": [
            {
              "description": "An ORCID identifier, see https://orcid.org/",
              "title": "OrcidId",
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "An [ORCID iD](https://support.orcid.org/hc/en-us/sections/360001495313-What-is-ORCID\n) in hyphenated groups of 4 digits, (and [valid](\nhttps://support.orcid.org/hc/en-us/articles/360006897674-Structure-of-the-ORCID-Identifier\n) as per ISO 7064 11,2.)",
          "examples": [
            "0000-0001-2345-6789"
          ],
          "title": "Orcid"
        },
        "name": {
          "title": "Name",
          "type": "string"
        },
        "github_user": {
          "anyOf": [
            {
              "type": "string"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "title": "Github User"
        }
      },
      "required": [
        "name"
      ],
      "title": "generic.v0_3.Author",
      "type": "object"
    },
    "RelativeFilePath": {
      "description": "A path relative to the `rdf.yaml` file (also if the RDF source is a URL).",
      "format": "path",
      "title": "RelativeFilePath",
      "type": "string"
    }
  },
  "additionalProperties": false,
  "properties": {
    "source": {
      "anyOf": [
        {
          "description": "A URL with the HTTP or HTTPS scheme.",
          "format": "uri",
          "maxLength": 2083,
          "minLength": 1,
          "title": "HttpUrl",
          "type": "string"
        },
        {
          "$ref": "#/$defs/RelativeFilePath"
        },
        {
          "format": "file-path",
          "title": "FilePath",
          "type": "string"
        }
      ],
      "description": "Source of the weights file.",
      "title": "FileSource"
    },
    "sha256": {
      "anyOf": [
        {
          "description": "A SHA-256 hash value",
          "maxLength": 64,
          "minLength": 64,
          "title": "Sha256",
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "SHA256 hash value of the **source** file.",
      "title": "Sha256"
    },
    "authors": {
      "anyOf": [
        {
          "items": {
            "$ref": "#/$defs/Author"
          },
          "type": "array"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "Authors\nEither the person(s) that have trained this model resulting in the original weights file.\n    (If this is the initial weights entry, i.e. it does not have a `parent`)\nOr the person(s) who have converted the weights to this weights format.\n    (If this is a child weight, i.e. it has a `parent` field)",
      "title": "Authors"
    },
    "parent": {
      "anyOf": [
        {
          "enum": [
            "keras_hdf5",
            "keras_v3",
            "onnx",
            "pytorch_state_dict",
            "tensorflow_js",
            "tensorflow_saved_model_bundle",
            "torchscript"
          ],
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The source weights these weights were converted from.\nFor example, if a model's weights were converted from the `pytorch_state_dict` format to `torchscript`,\nThe `pytorch_state_dict` weights entry has no `parent` and is the parent of the `torchscript` weights.\nAll weight entries except one (the initial set of weights resulting from training the model),\nneed to have this field.",
      "examples": [
        "pytorch_state_dict"
      ],
      "title": "Parent"
    },
    "comment": {
      "default": "",
      "description": "A comment about this weights entry, for example how these weights were created.",
      "title": "Comment",
      "type": "string"
    }
  },
  "required": [
    "source"
  ],
  "title": "model.v0_5.WeightsEntryDescrBase",
  "type": "object"
}

Fields:

• sha256 (Sha256 | None)
• source (Annotated[FileSource, AfterValidator(wo_special_file_name)])
• authors (list[Author] | None)
• parent (Annotated[WeightsFormat | None, Field(examples=['pytorch_state_dict'])])
• comment (str)

Validators:

• _validate

authors pydantic-field

authors: list[Author] | None = None

Authors Either the person(s) that have trained this model resulting in the original weights file. (If this is the initial weights entry, i.e. it does not have a parent) Or the person(s) who have converted the weights to this weights format. (If this is a child weight, i.e. it has a parent field)

comment pydantic-field

comment: str = ''

A comment about this weights entry, for example how these weights were created.

parent pydantic-field

parent: Annotated[
    WeightsFormat | None,
    Field(examples=["pytorch_state_dict"]),
] = None

The source weights these weights were converted from. For example, if a model's weights were converted from the pytorch_state_dict format to torchscript, The pytorch_state_dict weights entry has no parent and is the parent of the torchscript weights. All weight entries except one (the initial set of weights resulting from training the model), need to have this field.

sha256 pydantic-field

sha256: Sha256 | None = None

SHA256 hash value of the source file.

source pydantic-field

source: Annotated[
    FileSource, AfterValidator(wo_special_file_name)
]

Source of the weights file.

suffix property

suffix: str

type class-attribute

type: WeightsFormat

weights_format_name class-attribute

weights_format_name: str

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

download

download(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

alias for .get_reader

Source code in src/bioimageio/spec/_internal/io.py

def download(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """alias for `.get_reader`"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

get_reader

get_reader(
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
)

open the file source (download if needed)

Source code in src/bioimageio/spec/_internal/io.py

def get_reader(
    self,
    *,
    progressbar: ProgressbarLike
    | Callable[[], ProgressbarLike]
    | bool
    | None = None,
):
    """open the file source (download if needed)"""
    return get_reader(self.source, progressbar=progressbar, sha256=self.sha256)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

validate_sha256

validate_sha256(force_recompute: bool = False) -> None

validate the sha256 hash value of the source file

Source code in src/bioimageio/spec/_internal/io.py

def validate_sha256(self, force_recompute: bool = False) -> None:
    """validate the sha256 hash value of the **source** file"""
    context = get_validation_context()
    src_str = str(self.source)
    if force_recompute:
        actual_sha = None
    else:
        actual_sha = context.known_files.get(src_str)

    if actual_sha is None:
        if context.perform_io_checks or force_recompute:
            reader = get_reader(self.source, sha256=self.sha256)
            if force_recompute:
                actual_sha = get_sha256(reader)
            else:
                actual_sha = reader.sha256

            context.known_files[src_str] = actual_sha
        elif context.known_files and src_str not in context.known_files:
            # perform_io_checks is False, but known files were given,
            # so we expect all file references to be in there
            raise ValueError(f"File {src_str} not found in `known_files`.")

    if actual_sha is None or self.sha256 == actual_sha:
        return
    elif self.sha256 is None or context.update_hashes:
        self.sha256 = actual_sha
    elif self.sha256 != actual_sha:
        raise ValueError(
            f"Sha256 mismatch for {self.source}. Expected {self.sha256}, got "
            + f"{actual_sha}. Update expected `sha256` or point to the matching "
            + "file."
        )

WithHalo pydantic-model

Bases: Node

Show JSON schema:

{
  "$defs": {
    "SizeReference": {
      "additionalProperties": false,
      "description": "A tensor axis size (extent in pixels/frames) defined in relation to a reference axis.\n\n`axis.size = reference.size * reference.scale / axis.scale + offset`\n\nNote:\n1. The axis and the referenced axis need to have the same unit (or no unit).\n2. Batch axes may not be referenced.\n3. Fractions are rounded down.\n4. If the reference axis is `concatenable` the referencing axis is assumed to be\n    `concatenable` as well with the same block order.\n\nExample:\nAn unisotropic input image of w*h=100*49 pixels depicts a phsical space of 200*196mm\u00b2.\nLet's assume that we want to express the image height h in relation to its width w\ninstead of only accepting input images of exactly 100*49 pixels\n(for example to express a range of valid image shapes by parametrizing w, see `ParameterizedSize`).\n\n>>> w = SpaceInputAxis(id=AxisId(\"w\"), size=100, unit=\"millimeter\", scale=2)\n>>> h = SpaceInputAxis(\n...     id=AxisId(\"h\"),\n...     size=SizeReference(tensor_id=TensorId(\"input\"), axis_id=AxisId(\"w\"), offset=-1),\n...     unit=\"millimeter\",\n...     scale=4,\n... )\n>>> print(h.size.get_size(h, w))\n49\n\n\u21d2 h = w * w.scale / h.scale + offset = 100 * 2mm / 4mm - 1 = 49",
      "properties": {
        "tensor_id": {
          "description": "tensor id of the reference axis",
          "maxLength": 32,
          "minLength": 1,
          "title": "TensorId",
          "type": "string"
        },
        "axis_id": {
          "description": "axis id of the reference axis",
          "maxLength": 16,
          "minLength": 1,
          "title": "AxisId",
          "type": "string"
        },
        "offset": {
          "default": 0,
          "title": "Offset",
          "type": "integer"
        }
      },
      "required": [
        "tensor_id",
        "axis_id"
      ],
      "title": "model.v0_5.SizeReference",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "properties": {
    "halo": {
      "description": "The halo should be cropped from the output tensor to avoid boundary effects.\nIt is to be cropped from both sides, i.e. `size_after_crop = size - 2 * halo`.\nTo document a halo that is already cropped by the model use `size.offset` instead.",
      "minimum": 1,
      "title": "Halo",
      "type": "integer"
    },
    "size": {
      "$ref": "#/$defs/SizeReference",
      "description": "reference to another axis with an optional offset (see [SizeReference][])",
      "examples": [
        {
          "axis_id": "a",
          "offset": 5,
          "tensor_id": "t"
        }
      ]
    }
  },
  "required": [
    "halo",
    "size"
  ],
  "title": "model.v0_5.WithHalo",
  "type": "object"
}

Fields:

• halo (Annotated[int, Ge(1)])
• size (Annotated[SizeReference, Field(examples=[{'tensor_id': 't', 'axis_id': 'a', 'offset': 5}])])

halo pydantic-field

halo: Annotated[int, Ge(1)]

The halo should be cropped from the output tensor to avoid boundary effects. It is to be cropped from both sides, i.e. size_after_crop = size - 2 * halo. To document a halo that is already cropped by the model use size.offset instead.

size pydantic-field

size: Annotated[
    SizeReference,
    Field(
        examples=[
            {"tensor_id": "t", "axis_id": "a", "offset": 5}
        ]
    ),
]

reference to another axis with an optional offset (see SizeReference)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

ZeroMeanUnitVarianceDescr pydantic-model

Bases: NodeWithExplicitlySetFields

Subtract mean and divide by variance.

Examples:

Subtract tensor mean and variance - in YAML

preprocessing:
  - id: zero_mean_unit_variance

- in Python

>>> preprocessing = [ZeroMeanUnitVarianceDescr()]

Show JSON schema:

{
  "$defs": {
    "ZeroMeanUnitVarianceKwargs": {
      "additionalProperties": false,
      "description": "key word arguments for [ZeroMeanUnitVarianceDescr][]",
      "properties": {
        "axes": {
          "anyOf": [
            {
              "items": {
                "maxLength": 16,
                "minLength": 1,
                "title": "AxisId",
                "type": "string"
              },
              "type": "array"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "The subset of axes to normalize jointly, i.e. axes to reduce to compute mean/std.\nFor example to normalize 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x')\nresulting in a tensor of equal shape normalized per channel, specify `axes=('batch', 'x', 'y')`.\nTo normalize each sample independently leave out the 'batch' axis.\nDefault: Scale all axes jointly.",
          "examples": [
            [
              "batch",
              "x",
              "y"
            ]
          ],
          "title": "Axes"
        },
        "eps": {
          "default": 1e-06,
          "description": "epsilon for numeric stability: `out = (tensor - mean) / (std + eps)`.",
          "exclusiveMinimum": 0,
          "maximum": 0.1,
          "title": "Eps",
          "type": "number"
        }
      },
      "title": "model.v0_5.ZeroMeanUnitVarianceKwargs",
      "type": "object"
    }
  },
  "additionalProperties": false,
  "description": "Subtract mean and divide by variance.\n\nExamples:\n    Subtract tensor mean and variance\n    - in YAML\n    ```yaml\n    preprocessing:\n      - id: zero_mean_unit_variance\n    ```\n    - in Python\n    >>> preprocessing = [ZeroMeanUnitVarianceDescr()]",
  "properties": {
    "id": {
      "const": "zero_mean_unit_variance",
      "title": "Id",
      "type": "string"
    },
    "kwargs": {
      "$ref": "#/$defs/ZeroMeanUnitVarianceKwargs"
    }
  },
  "required": [
    "id"
  ],
  "title": "model.v0_5.ZeroMeanUnitVarianceDescr",
  "type": "object"
}

Fields:

• id (Literal['zero_mean_unit_variance'])
• kwargs (ZeroMeanUnitVarianceKwargs)

id pydantic-field

id: Literal["zero_mean_unit_variance"] = (
    "zero_mean_unit_variance"
)

implemented_id class-attribute

implemented_id: Literal["zero_mean_unit_variance"] = (
    "zero_mean_unit_variance"
)

kwargs pydantic-field

kwargs: ZeroMeanUnitVarianceKwargs

__pydantic_init_subclass__ classmethod

__pydantic_init_subclass__(**kwargs: Any) -> None

Source code in src/bioimageio/spec/_internal/common_nodes.py

@classmethod
def __pydantic_init_subclass__(cls, **kwargs: Any) -> None:
    explict_fields: dict[str, Any] = {}
    for attr in dir(cls):
        if attr.startswith("implemented_"):
            field_name = attr.replace("implemented_", "")
            if field_name not in cls.model_fields:
                continue

            assert (
                cls.model_fields[field_name].get_default() is PydanticUndefined
            ), field_name
            default = getattr(cls, attr)
            explict_fields[field_name] = default

    cls._fields_to_set_explicitly = MappingProxyType(explict_fields)
    return super().__pydantic_init_subclass__(**kwargs)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

ZeroMeanUnitVarianceKwargs pydantic-model

Bases: KwargsNode

key word arguments for ZeroMeanUnitVarianceDescr

Show JSON schema:

{
  "additionalProperties": false,
  "description": "key word arguments for [ZeroMeanUnitVarianceDescr][]",
  "properties": {
    "axes": {
      "anyOf": [
        {
          "items": {
            "maxLength": 16,
            "minLength": 1,
            "title": "AxisId",
            "type": "string"
          },
          "type": "array"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "description": "The subset of axes to normalize jointly, i.e. axes to reduce to compute mean/std.\nFor example to normalize 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x')\nresulting in a tensor of equal shape normalized per channel, specify `axes=('batch', 'x', 'y')`.\nTo normalize each sample independently leave out the 'batch' axis.\nDefault: Scale all axes jointly.",
      "examples": [
        [
          "batch",
          "x",
          "y"
        ]
      ],
      "title": "Axes"
    },
    "eps": {
      "default": 1e-06,
      "description": "epsilon for numeric stability: `out = (tensor - mean) / (std + eps)`.",
      "exclusiveMinimum": 0,
      "maximum": 0.1,
      "title": "Eps",
      "type": "number"
    }
  },
  "title": "model.v0_5.ZeroMeanUnitVarianceKwargs",
  "type": "object"
}

Fields:

• axes (Annotated[Sequence[AxisId] | None, Field(examples=[('batch', 'x', 'y')])])
• eps (Annotated[float, Interval(gt=0, le=0.1)])

axes pydantic-field

axes: Annotated[
    Sequence[AxisId] | None,
    Field(examples=[("batch", "x", "y")]),
] = None

The subset of axes to normalize jointly, i.e. axes to reduce to compute mean/std. For example to normalize 'batch', 'x' and 'y' jointly in a tensor ('batch', 'channel', 'y', 'x') resulting in a tensor of equal shape normalized per channel, specify axes=('batch', 'x', 'y'). To normalize each sample independently leave out the 'batch' axis. Default: Scale all axes jointly.

eps pydantic-field

eps: Annotated[float, Interval(gt=0, le=0.1)] = 1e-06

epsilon for numeric stability: out = (tensor - mean) / (std + eps).

__contains__

__contains__(item: str) -> bool

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __contains__(self, item: str) -> bool:
    return item in self.__class__.model_fields

__getitem__

__getitem__(item: str) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def __getitem__(self, item: str) -> Any:
    if item in self.__class__.model_fields:
        return getattr(self, item)
    else:
        raise KeyError(item)

dict_from_kwargs classmethod

dict_from_kwargs(
    *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def dict_from_kwargs(
    cls: Callable[P, T], *args: P.args, **kwargs: P.kwargs
) -> dict[str, Any]:
    assert not args, "Did not expected any args"
    return dict(kwargs)

get

get(item: str, default: Any = None) -> Any

Source code in src/bioimageio/spec/_internal/common_nodes.py

def get(self, item: str, default: Any = None) -> Any:
    return self[item] if item in self else default  # ruff: ignore[SIM401]

model_validate classmethod

model_validate(
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"]
    | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext
    | Mapping[str, Any]
    | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self

Validate a pydantic model instance.

Parameters:

Name	Type	Description	Default
obj	Any | Mapping[str, Any]	The object to validate.	required
strict	bool | None	Whether to raise an exception on invalid fields.	None
from_attributes	bool | None	Whether to extract data from object attributes.	None
context	ValidationContext | Mapping[str, Any] | None	Additional context to pass to the validator.	None

Raises:

Type	Description
ValidationError	If the object failed validation.

Returns:

Type	Description
Self	The validated description instance.

Source code in src/bioimageio/spec/_internal/node.py

@classmethod
def model_validate(
    cls,
    obj: Any | Mapping[str, Any],
    *,
    strict: bool | None = None,
    extra: Literal["allow", "ignore", "forbid"] | None = None,
    from_attributes: bool | None = None,
    context: ValidationContext | Mapping[str, Any] | None = None,
    by_alias: bool | None = None,
    by_name: bool | None = None,
) -> Self:
    """Validate a pydantic model instance.

    Args:
        obj: The object to validate.
        strict: Whether to raise an exception on invalid fields.
        from_attributes: Whether to extract data from object attributes.
        context: Additional context to pass to the validator.

    Raises:
        ValidationError: If the object failed validation.

    Returns:
        The validated description instance.
    """
    __tracebackhide__ = True

    if context is None:
        context = get_validation_context()
    elif isinstance(context, collections.abc.Mapping):
        context = ValidationContext(**context)

    assert not is_mapping(obj) or is_kwargs(obj), obj

    # TODO: pass on extra with pydantic >=2.12
    if extra is not None:
        warnings.warn("`extra` argument is currently ignored")

    with context:
        # use validation context as context manager for equal behavior of __init__ and model_validate
        return super().model_validate(
            obj, strict=strict, from_attributes=from_attributes
        )

convert_axes

convert_axes(
    axes: str,
    *,
    shape: Sequence[int]
    | _ParameterizedInputShape_v0_4
    | _ImplicitOutputShape_v0_4,
    tensor_type: Literal["input", "output"],
    halo: Sequence[int] | None,
    size_refs: Mapping[_TensorName_v0_4, Mapping[str, int]],
)

Source code in src/bioimageio/spec/model/v0_5.py

def convert_axes(
    axes: str,
    *,
    shape: Sequence[int] | _ParameterizedInputShape_v0_4 | _ImplicitOutputShape_v0_4,
    tensor_type: Literal["input", "output"],
    halo: Sequence[int] | None,
    size_refs: Mapping[_TensorName_v0_4, Mapping[str, int]],
):
    ret: list[AnyAxis] = []
    for i, a in enumerate(axes):
        axis_type = _AXIS_TYPE_MAP.get(a, a)
        if axis_type == "batch":
            ret.append(BatchAxis())
            continue

        scale = 1.0
        if isinstance(shape, _ParameterizedInputShape_v0_4):
            if shape.step[i] == 0:
                size = shape.min[i]
            else:
                size = ParameterizedSize(min=shape.min[i], step=shape.step[i])
        elif isinstance(shape, _ImplicitOutputShape_v0_4):
            ref_t = str(shape.reference_tensor)
            if ref_t.count(".") == 1:
                t_id, orig_a_id = ref_t.split(".")
            else:
                t_id = ref_t
                orig_a_id = a

            a_id = _AXIS_ID_MAP.get(orig_a_id, a)
            if not (orig_scale := shape.scale[i]):
                # old way to insert a new axis dimension
                size = int(2 * shape.offset[i])
            else:
                scale = 1 / orig_scale
                if axis_type in ("channel", "index"):
                    # these axes no longer have a scale
                    offset_from_scale = orig_scale * size_refs.get(
                        _TensorName_v0_4(t_id), {}
                    ).get(orig_a_id, 0)
                else:
                    offset_from_scale = 0
                size = SizeReference(
                    tensor_id=TensorId(t_id),
                    axis_id=AxisId(a_id),
                    offset=int(offset_from_scale + 2 * shape.offset[i]),
                )
        else:
            size = shape[i]

        if axis_type == "time":
            if tensor_type == "input":
                ret.append(TimeInputAxis(size=size, scale=scale))
            else:
                assert not isinstance(size, ParameterizedSize)
                if halo is None:
                    ret.append(TimeOutputAxis(size=size, scale=scale))
                else:
                    assert not isinstance(size, int)
                    ret.append(
                        TimeOutputAxisWithHalo(size=size, scale=scale, halo=halo[i])
                    )

        elif axis_type == "index":
            if tensor_type == "input":
                ret.append(IndexInputAxis(size=size))
            else:
                if isinstance(size, ParameterizedSize):
                    size = DataDependentSize(min=size.min)

                ret.append(IndexOutputAxis(size=size))
        elif axis_type == "channel":
            assert not isinstance(size, ParameterizedSize)
            if isinstance(size, SizeReference):
                warnings.warn(
                    "Conversion of channel size from an implicit output shape may be"
                    + " wrong"
                )
                ret.append(
                    ChannelAxis(
                        channel_names=[f"channel{i}" for i in range(size.offset)]
                    )
                )
            else:
                ret.append(
                    ChannelAxis(channel_names=[f"channel{i}" for i in range(size)])
                )
        elif axis_type == "space":
            if tensor_type == "input":
                ret.append(SpaceInputAxis(id=AxisId(a), size=size, scale=scale))
            else:
                assert not isinstance(size, ParameterizedSize)
                if halo is None or halo[i] == 0:
                    ret.append(SpaceOutputAxis(id=AxisId(a), size=size, scale=scale))
                elif isinstance(size, int):
                    raise NotImplementedError(
                        f"output axis with halo and fixed size (here {size}) not allowed"
                    )
                else:
                    ret.append(
                        SpaceOutputAxisWithHalo(
                            id=AxisId(a), size=size, scale=scale, halo=halo[i]
                        )
                    )

    return ret

generate_covers

generate_covers(
    inputs: Sequence[tuple[InputTensorDescr, NDArray[Any]]],
    outputs: Sequence[
        tuple[OutputTensorDescr, NDArray[Any]]
    ],
) -> list[FileDescr]

Source code in src/bioimageio/spec/model/v0_5.py

def generate_covers(
    inputs: Sequence[tuple[InputTensorDescr, NDArray[Any]]],
    outputs: Sequence[tuple[OutputTensorDescr, NDArray[Any]]],
) -> list[FileDescr]:
    def squeeze(
        data: NDArray[Any], axes: Sequence[AnyAxis]
    ) -> tuple[NDArray[Any], list[AnyAxis]]:
        """apply numpy.ndarray.squeeze while keeping track of the axis descriptions remaining"""
        if data.ndim != len(axes):
            raise ValueError(
                f"tensor shape {data.shape} does not match described axes"
                + f" {[a.id for a in axes]}"
            )

        axes = [deepcopy(a) for a, s in zip(axes, data.shape) if s != 1]
        return data.squeeze(), axes

    def normalize(
        data: NDArray[Any], axis: tuple[int, ...] | None, eps: float = 1e-7
    ) -> NDArray[np.float32]:
        data = data.astype("float32")
        data -= data.min(axis=axis, keepdims=True)
        data /= data.max(axis=axis, keepdims=True) + eps
        return data

    def to_2d_image(data: NDArray[Any], axes: Sequence[AnyAxis]):
        original_shape = data.shape
        original_axes = list(axes)
        data, axes = squeeze(data, axes)

        # take slice fom any batch or index axis if needed
        # and convert the first channel axis and take a slice from any additional channel axes
        slices: tuple[slice, ...] = ()
        ndim = data.ndim
        ndim_need = 3 if any(isinstance(a, ChannelAxis) for a in axes) else 2
        has_c_axis = False
        for i, a in enumerate(axes):
            s = data.shape[i]
            assert s > 1
            if (
                isinstance(a, (BatchAxis, IndexInputAxis, IndexOutputAxis))
                and ndim > ndim_need
            ):
                data = data[slices + (slice(s // 2 - 1, s // 2),)]
                ndim -= 1
            elif isinstance(a, ChannelAxis):
                if has_c_axis:
                    # second channel axis
                    data = data[slices + (slice(0, 1),)]
                    ndim -= 1
                else:
                    has_c_axis = True
                    if s == 2:
                        # visualize two channels with cyan and magenta
                        data = np.concatenate(
                            [
                                data[slices + (slice(1, 2),)],
                                data[slices + (slice(0, 1),)],
                                (
                                    data[slices + (slice(0, 1),)]
                                    + data[slices + (slice(1, 2),)]
                                )
                                / 2,  # TODO: take maximum instead?
                            ],
                            axis=i,
                        )
                    elif data.shape[i] == 3:
                        pass  # visualize 3 channels as RGB
                    else:
                        # visualize first 3 channels as RGB
                        data = data[slices + (slice(3),)]

                    assert data.shape[i] == 3

            slices += (slice(None),)

        data, axes = squeeze(data, axes)
        assert len(axes) == ndim
        # take slice from z axis if needed
        slices = ()
        if ndim > ndim_need:
            for i, a in enumerate(axes):
                s = data.shape[i]
                if a.id == AxisId("z"):
                    data = data[slices + (slice(s // 2 - 1, s // 2),)]
                    data, axes = squeeze(data, axes)
                    ndim -= 1
                    break

            slices += (slice(None),)

        # take slice from any space or time axis
        slices = ()

        for i, a in enumerate(axes):
            if ndim <= ndim_need:
                break

            s = data.shape[i]
            assert s > 1
            if isinstance(
                a, (SpaceInputAxis, SpaceOutputAxis, TimeInputAxis, TimeOutputAxis)
            ):
                data = data[slices + (slice(s // 2 - 1, s // 2),)]
                ndim -= 1

            slices += (slice(None),)

        del slices
        data, axes = squeeze(data, axes)
        assert len(axes) == ndim

        if (has_c_axis and ndim != 3) or (not has_c_axis and ndim != 2):
            raise ValueError(
                f"Failed to construct cover image from shape {original_shape} with axes {[a.id for a in original_axes]}."
            )

        if not has_c_axis:
            assert ndim == 2
            data = np.repeat(data[:, :, None], 3, axis=2)
            axes.append(ChannelAxis(channel_names=list("RGB")))
            ndim += 1

        assert ndim == 3

        # transpose axis order such that longest axis comes first...
        axis_order: list[int] = [int(i) for i in np.argsort(list(data.shape))]
        axis_order.reverse()
        # ... and channel axis is last
        c = next(i for i in range(3) if isinstance(axes[i], ChannelAxis))
        axis_order.append(axis_order.pop(c))
        axes = [axes[ao] for ao in axis_order]
        data = data.transpose(axis_order)

        # h, w = data.shape[:2]
        # if h / w  in (1.0 or 2.0):
        #     pass
        # elif h / w < 2:
        # TODO: enforce 2:1 or 1:1 aspect ratio for generated cover images

        norm_along = (
            tuple(i for i, a in enumerate(axes) if a.type in ("space", "time")) or None
        )
        # normalize the data and map to 8 bit
        data = normalize(data, norm_along)
        data = (data * 255).astype("uint8")

        return data

    def create_diagonal_split_image(im0: NDArray[Any], im1: NDArray[Any]):
        assert im0.dtype == im1.dtype == np.uint8
        assert im0.shape == im1.shape
        assert im0.ndim == 3
        N, M, C = im0.shape
        assert C == 3
        out = np.ones((N, M, C), dtype="uint8")
        for c in range(C):
            outc = np.tril(im0[..., c])
            mask = outc == 0
            outc[mask] = np.triu(im1[..., c])[mask]
            out[..., c] = outc

        return out

    if not inputs:
        raise ValueError("Missing test input tensor for cover generation.")

    if not outputs:
        raise ValueError("Missing test output tensor for cover generation.")

    ipt_descr, ipt = inputs[0]
    out_descr, out = outputs[0]

    ipt_img = to_2d_image(ipt, ipt_descr.axes)
    out_img = to_2d_image(out, out_descr.axes)

    cover_folder = Path(mkdtemp())
    if ipt_img.shape == out_img.shape:
        covers = [cover_folder / "cover.png"]
        imwrite(covers[0], create_diagonal_split_image(ipt_img, out_img))
    else:
        covers = [cover_folder / "input.png", cover_folder / "output.png"]
        imwrite(covers[0], ipt_img)
        imwrite(covers[1], out_img)

    return [FileDescr(source=c) for c in covers]

get_halos

get_halos(
    tensors: Mapping[TensorId, TensorDescr],
) -> dict[TensorId, dict[AxisId, tuple[int, int]]]

Get all input and output halos from tensor descriptions.

Note

• Input halos are to be padded
• Output halos are to be cropped

Source code in src/bioimageio/spec/model/v0_5.py

def get_halos(
    tensors: Mapping[TensorId, TensorDescr],
    /,
) -> dict[TensorId, dict[AxisId, tuple[int, int]]]:
    """Get all input and output halos from tensor descriptions.

    Note:
        - Input halos are to be padded
        - Output halos are to be cropped
    """
    halos: dict[TensorId, dict[AxisId, tuple[int, int]]] = {}
    for descr in tensors.values():
        if isinstance(descr, InputTensorDescr):
            continue
        for axis in descr.axes:
            if not isinstance(axis, WithHalo):
                continue

            ref_scale = next(
                a
                for a in tensors[axis.size.tensor_id].axes
                if a.id == axis.size.axis_id
            ).scale

            # set output halo (to be cropped)
            halos.setdefault(descr.id, {})[axis.id] = (axis.halo, axis.halo)
            # set input halo (to be padded)
            pad_width = int(axis.halo / axis.scale * ref_scale)
            halos.setdefault(axis.size.tensor_id, {})[axis.size.axis_id] = (
                pad_width,
                pad_width,
            )

    return halos

validate_tensors

validate_tensors(
    tensors: Mapping[
        TensorId, tuple[TensorDescr, NDArray[Any] | None]
    ],
    tensor_origin: Literal[
        "source", "test_tensor"
    ] = "source",
    *,
    pad_inputs: bool | Literal["allow"] = True,
    crop_outputs: bool | Literal["allow"] = True,
)

Validate all inputs (and optionally output tensors) against their tensor descriptions.

Parameters:

Name	Type	Description	Default
tensors	Mapping[TensorId, tuple[TensorDescr, NDArray[Any] | None]]	Mapping of tensor id to a tuple of tensor description and optional numpy array.	required
tensor_origin	Literal['source', 'test_tensor']	String to use in error messages to indicate the origin of the tensors being validated.	'source'
pad_inputs	bool | Literal['allow']	Wether to apply/allow padding of inputs before shape comparison	True
crop_outputs	bool | Literal['allow']	Wether to apply/allow cropping of outputs before shape comparison.	True

Source code in src/bioimageio/spec/model/v0_5.py

def validate_tensors(
    tensors: Mapping[TensorId, tuple[TensorDescr, NDArray[Any] | None]],
    tensor_origin: Literal[
        "source", "test_tensor"
    ] = "source",  # for more precise error messages
    *,
    pad_inputs: bool | Literal["allow"] = True,
    crop_outputs: bool | Literal["allow"] = True,
):
    """Validate all inputs (and optionally output tensors) against their tensor descriptions.

    Args:
        tensors: Mapping of tensor id to a tuple of tensor description and optional numpy array.
        tensor_origin: String to use in error messages to indicate the origin of the tensors being validated.
        pad_inputs: Wether to apply/allow padding of inputs before shape comparison
        crop_outputs: Wether to apply/allow cropping of outputs before shape comparison.
    """
    all_tensor_axes: dict[TensorId, dict[AxisId, tuple[AnyAxis, int | None]]] = {}

    def e_msg_location(d: TensorDescr):
        return f"{'inputs' if isinstance(d, InputTensorDescr) else 'outputs'}[{d.id}]"

    for descr, array in tensors.values():
        if array is None:
            axis_sizes = {a.id: None for a in descr.axes}
        else:
            try:
                axis_sizes = descr.get_axis_sizes_for_array(array)
            except ValueError as e:
                raise ValueError(f"{e_msg_location(descr)} {e}")

        all_tensor_axes[descr.id] = {a.id: (a, axis_sizes[a.id]) for a in descr.axes}

    # get halos to be padded/cropped to validate against halo-adjusted sizes
    io_halos = get_halos({k: v[0] for k, v in tensors.items()})

    for descr, array in tensors.values():
        if array is None:
            continue

        if descr.dtype in ("float32", "float64"):
            invalid_test_tensor_dtype = array.dtype.name not in (
                "float32",
                "float64",
                "uint8",
                "int8",
                "uint16",
                "int16",
                "uint32",
                "int32",
                "uint64",
                "int64",
            )
        else:
            invalid_test_tensor_dtype = array.dtype.name != descr.dtype

        if invalid_test_tensor_dtype:
            raise ValueError(
                f"{tensor_origin} data type '{array.dtype.name}' does not"
                + f" match described {e_msg_location(descr)}.dtype '{descr.dtype}'"
            )

        if array.min() > -1e-4 and array.max() < 1e-4:
            raise ValueError(
                "Output values are too small for reliable testing."
                + f" Values <-1e5 or >=1e5 must be present in {tensor_origin}"
            )

        for a in descr.axes:
            actual_size = all_tensor_axes[descr.id][a.id][1]

            if actual_size is None:
                continue

            if a.size is None:
                continue

            # add padding width to actual tensor size
            total_axis_halo = sum(io_halos.get(descr.id, {}).get(a.id, (0, 0)))
            if isinstance(descr, InputTensorDescr):
                # pad input halos
                actual_size_with_halo = actual_size + total_axis_halo
                if pad_inputs is True:
                    check_sizes = {actual_size_with_halo}
                    size_hint = " (after padding input halo)"
                elif pad_inputs == "allow":
                    check_sizes = {actual_size, actual_size_with_halo}
                    size_hint = " (with or without padding input halo)"
                elif pad_inputs is False:
                    check_sizes = {actual_size}
                    size_hint = ""
                else:
                    assert_never(pad_inputs)

            elif isinstance(descr, OutputTensorDescr):
                # crop output halos
                actual_size_with_halo = max(0, actual_size - total_axis_halo)
                if crop_outputs is True:
                    check_sizes = {actual_size_with_halo}
                    size_hint = " (after cropping output halo)"
                elif crop_outputs == "allow":
                    check_sizes = {actual_size, actual_size_with_halo}
                    size_hint = " (with or without cropping output halo)"
                elif crop_outputs is False:
                    check_sizes = {actual_size}
                    size_hint = ""
                else:
                    assert_never(crop_outputs)
            else:
                assert_never(descr)

            del actual_size  # make sure we explicitly use unchanged or halo-adjusted size from here on

            if isinstance(a.size, int):
                if a.size not in check_sizes:
                    raise ValueError(
                        f"{e_msg_location(descr)}.axes[{a.id}]: {tensor_origin} axis "
                        + f"has incompatible size {check_sizes}{size_hint}, expected {a.size}"
                    )
            elif isinstance(a.size, (ParameterizedSize, DataDependentSize)):
                _ = try_all_raise_last(
                    (partial(a.size.validate_size, s) for s in check_sizes),
                    f"{e_msg_location(descr)}.axes[{a.id}]: {tensor_origin} axis ",
                )
            elif isinstance(a.size, SizeReference):
                ref_tensor_axes = all_tensor_axes.get(a.size.tensor_id)
                if ref_tensor_axes is None:
                    raise ValueError(
                        f"{e_msg_location(descr)}.axes[{a.id}].size.tensor_id: Unknown tensor"
                        + f" reference '{a.size.tensor_id}', available: {list(all_tensor_axes)}"
                    )

                ref_axis, ref_size = ref_tensor_axes.get(a.size.axis_id, (None, None))
                if ref_axis is None or ref_size is None:
                    raise ValueError(
                        f"{e_msg_location(descr)}.axes[{a.id}].size.axis_id: Unknown tensor axis"
                        + f" reference '{a.size.tensor_id}.{a.size.axis_id}, available: {list(ref_tensor_axes)}"
                    )

                if a.unit != ref_axis.unit:
                    raise ValueError(
                        f"{e_msg_location(descr)}.axes[{a.id}].size: `SizeReference` requires"
                        + " axis and reference axis to have the same `unit`, but"
                        + f" {a.unit}!={ref_axis.unit}"
                    )

                if (
                    expected_size := (
                        ref_size * ref_axis.scale / a.scale + a.size.offset
                    )
                ) not in check_sizes:
                    raise ValueError(
                        f"{e_msg_location(descr)}.{tensor_origin}: axis '{a.id}' of size"
                        + f" {check_sizes} invalid for referenced size {ref_size};"
                        + f" expected {expected_size}"
                    )
            else:
                assert_never(a.size)