fiftyone.utils.flash

module documentation

(source)

Utilities for working with Lightning Flash.

Function	`apply_flash_model`	Applies the given `Lightning Flash model` to the samples in the collection.
Function	`compute_flash_embeddings`	Computes embeddings for the samples in the collection using the given `Lightning Flash model`.
Function	`_do_export_array`	Undocumented
Function	`_export_arrays`	Undocumented
Function	`_get_output`	Undocumented
Constant	`_MODEL_TO_DATAMODULE_MAP`	Undocumented
Constant	`_SUPPORTED_MODELS`	Undocumented

def apply_flash_model(samples, model, label_field='predictions', confidence_thresh=None, store_logits=False, batch_size=None, num_workers=None, output_dir=None, rel_dir=None, transform_kwargs=None, trainer_kwargs=None): (source) ¶

Applies the given Lightning Flash model to the samples in the collection.

Parameters
samples	a `fiftyone.core.collections.SampleCollection`
model	a `flash:flash.core.model.Task`
label_field:"predictions"	the name of the field in which to store the model predictions. When performing inference on video frames, the "frames." prefix is optional
confidence_thresh:`None`	an optional confidence threshold to apply to any applicable labels generated by the model
store_logits:`False`	whether to store logits for the model predictions. This is only supported when the provided `model` has logits
batch_size:`None`	an optional batch size to use. If not provided, a default batch size is used
num_workers:`None`	the number of workers for the data loader to use
output_dir:`None`	an optional output directory in which to write segmentation images. Only applicable if the model generates segmentations. If none is provided, the segmentations are stored in the database
rel_dir:`None`	an optional relative directory to strip from each input filepath to generate a unique identifier that is joined with `output_dir` to generate an output path for each segmentation image. This argument allows for populating nested subdirectories in `output_dir` that match the shape of the input paths. The path is converted to an absolute path (if necessary) via `fiftyone.core.storage.normalize_path`
transform_kwargs:`None`	an optional dict of transform kwargs to pass into the created data module used by some models
trainer_kwargs:`None`	an optional dict of kwargs used to initialize the `Trainer`. These can be used to, for example, configure the number of GPUs to use and other distributed inference parameters

def compute_flash_embeddings(samples, model, embeddings_field=None, batch_size=None, num_workers=None, transform_kwargs=None, trainer_kwargs=None): (source) ¶

Computes embeddings for the samples in the collection using the given Lightning Flash model.

This method only supports applying an :ref:`ImageEmbedder <flash:image_embedder>` to an image collection.

If an embeddings_field is provided, the embeddings are saved to the samples; otherwise, the embeddings are returned in-memory.

Parameters
samples	a `fiftyone.core.collections.SampleCollection`
model	a `flash:flash.core.model.Task`
embeddings_field:`None`	the name of a field in which to store the embeddings
batch_size:`None`	an optional batch size to use. If not provided, a default batch size is used
num_workers:`None`	the number of workers for the data loader to use
transform_kwargs:`None`	an optional dict of transform kwargs to pass into the created data module used by some models
trainer_kwargs:`None`	an optional dict of kwargs used to initialize the `Trainer`. These can be used to, for example, configure the number of GPUs to use and other distributed inference parameters
Returns
`one` of the following	`None`, if an `embeddings_field` is provided a `num_samples x num_dim` array of embeddings, if `embeddings_field` is None

def _do_export_array(label, input_path, filename_maker): (source) ¶

Undocumented

def _export_arrays(label, input_path, filename_maker): (source) ¶

Undocumented

def _get_output(model, confidence_thresh, store_logits): (source) ¶

Undocumented

_MODEL_TO_DATAMODULE_MAP = (source) ¶

Undocumented

Value

{fi.ObjectDetector: fi.ObjectDetectionData,
 fi.ImageClassifier: fi.ImageClassificationData,
 fi.SemanticSegmentation: fi.SemanticSegmentationData,
 fv.VideoClassifier: fv.VideoClassificationData}

_SUPPORTED_MODELS = (source) ¶

Undocumented

Value

(fi.ImageClassifier, fi.ObjectDetector, fi.SemanticSegmentation)