YAHP interface for ComposerDeepLabV3.


These classes are used with yahp for YAML-based configuration.

class composer.models.deeplabv3.deeplabv3_hparams.DeepLabV3Hparams(initializers=<factory>, num_classes=None, backbone_arch='resnet101', is_backbone_pretrained=True, backbone_url='', use_plus=True, sync_bn=True)[source]#

Bases: composer.models.model_hparams.ModelHparams

YAHP interface for ComposerDeepLabV3.

  • num_classes (int) โ€“ Number of classes in the segmentation task.

  • backbone_arch (str, optional) โ€“ The architecture to use for the backbone. Must be either ['resnet50', 'resnet101']. Default: 'resnet101'.

  • is_backbone_pretrained (bool, optional) โ€“ If True, use pretrained weights for the backbone. Default: True.

  • backbone_url (str, optional) โ€“ Url used to download model weights. If empty, the PyTorch url will be used. Default: ''.

  • use_plus (bool, optional) โ€“ If True, use DeepLabv3+ head instead of DeepLabv3. Default: True.

  • sync_bn (bool, optional) โ€“ If True, replace all BatchNorm layers with SyncBatchNorm layers. Default: True.

  • initializers (List[Initializer], optional) โ€“ Initializers for the model. [] for no initialization. Default: [].