steps.ultralytics.model.trainer

trainer

Functions:

Name	Description
train_ultralytics_model	Train an Ultralytics model using the provided dataset collection and training context.

train_ultralytics_model(model, dataset_collection)

Train an Ultralytics model using the provided dataset collection and training context.

This step: - Retrieves the active training context to access hyperparameters, augmentation settings, and experiment metadata. - Initializes an UltralyticsModelTrainer to handle the training logic. - Runs the training pipeline on the dataset collection. - Sets the latest run directory and locates the best model weights after training. - Saves the trained model weights as an artifact in the experiment.

Parameters:

Name	Type	Description	Default
model	UltralyticsModel	The model instance to be trained.	required
dataset_collection	DatasetCollection[TBaseDataset]	The dataset collection used for training, typically including 'train', 'val', and optionally 'test' datasets.	required

Returns:

Name	Type	Description
UltralyticsModel	UltralyticsModel	The trained model with updated internal state and trained weights.