fiftyone.utils.eta.ETAModelConfig

class documentation

class ETAModelConfig(fom.ModelConfig): (source)

Meta-config class that encapsulates the configuration of an eta.core.learning.Model that is to be run via the ETAModel wrapper.

Example:

import fiftyone.core.models as fom

model = fom.load_model({
    "type": "fiftyone.utils.eta.ETAModel",
    "config": {
        "type": "eta.detectors.YOLODetector",
        "config": {
            "model_name": "yolo-v2-coco"
        }
    }
})

Parameters
type	the fully-qualified class name of the `fiftyone.core.models.Model` subclass, which must be `ETAModel` or a subclass of it
config	a dict containing the `eta.core.learning.ModelConfig` for the ETA model

Method	`confidence_thresh.setter`	Undocumented
Property	`confidence_thresh`	The confidence threshold of the underlying `eta.core.model.Model`.

@confidence_thresh.setter
def confidence_thresh(self, confidence_thresh): (source) ¶

Undocumented

@property
confidence_thresh = (source) ¶

The confidence threshold of the underlying eta.core.model.Model.

Note that this may not be defined for some models.