neodroidvision.multitask.multi_task_learning_scheduler.cache_backbone_results

neodroidvision.multitask.multi_task_learning_scheduler.cache_backbone_results(model: Module) → None

Cache and freeze the backbone part of the model and only update heads