In each down layer, the input is first downsampled using maxpooling, then passed through a convolution block, unless this is the top

Returns the encoding (down) part of a layer of the network. This typically will downsample data at some point in its structure. Its o

Get spatial shape for fake data according to the specified shape pattern. It supports `int` number and `string` with formats like: "32", "32

Calculate the amplitude of the fourier transformations representation of the images Args: images: Images that are to und

Extract hidden states from BertLayer output. Compatible with both older transformers (returns a tuple) and newer transformers >=5.0 (

Compute the mean of the posterior at timestep t. Args: timestep: current timestep. x0: the noise-free input.

Get the input and output information defined in the metadata. Args: parser: a ConfigParser of the given bundle. prefix: a pr

Select one of the indices within slice containing val. Args: val : value for comparison dim : dimension in w

Calculate the power of the unit factor with respect to the base unit

Abstract callback function for the processing logic of 1 iteration in Ignite Engine. Need subclass to implement different logics, lik

Args: idx: the index of the input data sequence.

Loads TRT plan from disk and activates its execution context.

Load an Algo object from a legacy pickle file. This is an internal function to support backward compatibility with pickle files. Gated b

Helper function to pad inputs.

Positionally encode points that are normalized to [0,1].

Pop the kwargs used to define ImageSave class for the ensemble output. Args: kwargs: image writing parameters for the en

Prepare the command for distributed job running using bcprun. Args: script: the script to run in the distributed job. cmd_pr

Prepare the command for subprocess to run the script with the given arguments. Args: cmd: the command or script to run in the distri

Prepare the command for multi-gpu/multi-node job execution using torchrun. Args: cmd: the command or script to run in the distribute

linear resample

Resolve and return one ``ConfigItem`` of ``id``, cache the resolved result in ``resolved_content``. If it has unresolved references,

Run the command with bcprun. Args: cmd: the command to run. Typically it is prepared by ``_prepare_cmd_bcprun``. kwargs: the

Run the command with torchrun. Args: cmd: the command to run. Typically it is prepared by ``_prepare_cmd_torchrun``. kwargs:

Helper function to introduce a given intensity at given location. Args: k: intensity array to alter. idx: in

Compute the square root of a matrix.

Strip the optional PyTorch >= 2.13 ``, bias=True|False`` repr fragment. Args: text: Layer string representation to normalize. Return

All gather the buffers across distributed ranks for aggregating. Each buffer will be concatenated as a PyTorch Tensor.

Fetch single data item from `self.data`.

Return whether a traced transform entry matches this transform. Matching succeeds when the traced ID matches this instance, when the ID

Wrap a parsed dict/list in a :class:`_ConfigProxy` so nested access keeps chaining; pass scalars through. Args: parser: the owning :

Complex dot product between tensors x1 and x2: sum(x1.*x2)

Complex dot product between tensor x and itself

Add the factory function to this object under the given name, with optional description.

Besides the default predefined properties, some 3rd party applications may need the bundle definition to provide additional propertie

Besides the default predefined properties, some 3rd party applications may need the bundle definition to provide additional propertie

Extract the affine matrix from a pending transform item.

Functional implementation of affine. This function operates eagerly or lazily according to ``lazy`` (default ``False``). Args:

returns the total average value (averaged across processes) Args: to_numpy: whether to convert to numpy array. Defaults

Execute reduction logic for the output of `compute_average_surface_distance`. Args: reduction: define mode of reduction

Execute reduction logic for the output of `compute_hausdorff_distance`. Args: reduction: define mode of reduction to the

Typically `y_pred` and `y` are stored in the cumulative buffers at each iteration, This function reads the buffers and computes the A

Returns the aggregated loss value across multiple iterations. Args: reduction: define mode of reduction to the metrics,

Execute reduction logic for the output of `compute_iou`. Args: reduction: define mode of reduction to the metrics, will

Aggregate values for merging. Args: values: a tensor of shape BCHW[D], representing the values of inference output.

Aggregate values for merging. Args: values: a tensor of shape BCHW[D], representing the values of inference output.

This function applies affine matrices to the boxes Args: boxes: bounding boxes, Nx4 or Nx6 torch tensor or ndarray. The box mode is

apply_pending_transforms is called with either a tensor or a dictionary, some entries of which contain tensors. When operating on a dict

`np.argsort` with equivalent implementation for torch. Args: a: the array/tensor to sort. axis: axis along which to sort. Re

Args: engine: Ignite Engine, it can be a trainer, validator or evaluator.

Args: engine: Ignite Engine, it can be a trainer, validator or evaluator.

Args: engine: Ignite Engine, it can be a trainer, validator or evaluator.

Args: engine: Ignite Engine, it can be a trainer, validator or evaluator.

Convert the bounding boxes of the current mode to corners. Args: boxes: bounding boxes, Nx4 or Nx6 torch tensor

Builds a sin-cos position embedding based on the given grid size, embed dimension, spatial dimensions, and temperature. Reference: https://gi

Calculate the output tensor shape when applying a convolution to a tensor of shape `inShape` with kernel size `kernel_size`, stride value `st

Utility function to cast all tensors in a tuple from from_dtype to to_dtype

Checks which center points are within boxes Args: boxes: bounding boxes, Nx4 or Nx6 torch tensor or ndarray. The box mode is assumed

convert `params` into '--key_1=value_1 --key_2=value_2 ...

Check whether the required properties are existing in the bundle workflow. If no workflow type specified, return None, otherwise, ret

Check whether the required properties are existing in the bundle workflow. If the optional properties have reference in the config, w

Clone data independent of type. Args: data (NdarrayTensor): This can be a Pytorch Tensor or numpy array. Returns: Any: