PointCloudObjectDetector¶
- class flash.pointcloud.detection.model.PointCloudObjectDetector(num_classes, backbone='pointpillars_kitti', backbone_kwargs=None, loss_fn=None, optimizer='Adam', lr_scheduler=None, metrics=None, learning_rate=None, lambda_loss_cls=1.0, lambda_loss_bbox=1.0, lambda_loss_dir=1.0)[source]¶
The
PointCloudObjectDetector
is aClassificationTask
that classifies pointcloud data.- Parameters
num_classes¶ (
int
) – The number of classes (outputs) for thisTask
.backbone¶ (
Union
[str
,Tuple
[Module
,int
]]) – The backbone name (or a tuple ofnn.Module
, output size) to use.backbone_kwargs¶ (
Optional
[Dict
]) – Any additional kwargs to pass to the backbone constructor.loss_fn¶ (
Optional
[TypeVar
(LOSS_FN_TYPE
,Callable
,Mapping
,Sequence
,None
)]) – The loss function to use. IfNone
, a default will be selected by theClassificationTask
depending on themulti_label
argument.optimizer¶ (
TypeVar
(OPTIMIZER_TYPE
,str
,Callable
,Tuple
[str
,Dict
[str
,Any
]],None
)) – Optimizer to use for training.lr_scheduler¶ (
Optional
[TypeVar
(LR_SCHEDULER_TYPE
,str
,Callable
,Tuple
[str
,Dict
[str
,Any
]],Tuple
[str
,Dict
[str
,Any
],Dict
[str
,Any
]],None
)]) – The LR scheduler to use during training.metrics¶ (
Optional
[TypeVar
(METRICS_TYPE
,Metric
,Mapping
,Sequence
,None
)]) – Any metrics to use with thisTask
. IfNone
, a default will be selected by theClassificationTask
depending on themulti_label
argument.learning_rate¶ (
Optional
[float
]) – The learning rate for the optimizer.lambda_loss_cls¶ (
float
) – The value to scale the loss classification.lambda_loss_bbox¶ (
float
) – The value to scale the bounding boxes loss.lambda_loss_dir¶ (
float
) – The value to scale the bounding boxes direction loss.
- classmethod available_finetuning_strategies(cls)¶
Returns a list containing the keys of the available Finetuning Strategies.
- classmethod available_lr_schedulers(cls)¶
Returns a list containing the keys of the available LR schedulers.
- classmethod available_optimizers(cls)¶
Returns a list containing the keys of the available Optimizers.
- classmethod available_outputs(cls)¶
Returns the list of available outputs (that can be used during prediction or serving) for this
Task
.Examples
..testsetup:
>>> from flash import Task
>>> print(Task.available_outputs()) ['preds', 'raw']