XDER CE#
Arguments#
Options
- --alphafloat
Help: Penalty weight.
Default:
None
- --betafloat
Help: Penalty weight.
Default:
None
- --gammafloat
Help: Weight for logit update
Default:
0.85
- --constr_etafloat
Help: Regularization weight for past/future constraints
Default:
0.01
- --constr_marginfloat
Help: Margin for past/future constraints
Default:
0.3
- --align_bn0|1|True|False -> bool
Help: Use BatchNorm alignment
Default:
1
Rehearsal arguments
Arguments shared by all rehearsal-based methods.
- --buffer_sizeint
Help: The size of the memory buffer.
Default:
None
- --minibatch_sizeint
Help: The batch size of the memory buffer.
Default:
None
Classes#
- class models.xder_ce.XDerCe(backbone, loss, args, transform, dataset=None)[source]#
Bases:
ContinualModel
Continual learning via eXtended Dark Experience Replay with cross-entropy on future heads.