GEM

Arguments

Options

--gammafloat

Help: Margin parameter for GEM.

• Default: 0.5

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.gem.Gem(backbone, loss, args, transform, dataset=None)

Bases: ContinualModel

Continual learning via Gradient Episodic Memory.

COMPATIBILITY: List[str] = ['class-il', 'domain-il', 'task-il']

NAME: str = 'gem'

end_task(dataset)

static get_parser(parser)

Return type:

ArgumentParser

observe(inputs, labels, not_aug_inputs, epoch=None)

Functions

models.gem.overwrite_grad(params, newgrad, grad_dims)

This is used to overwrite the gradients with a new gradient vector, whenever violations occur. pp: parameters newgrad: corrected gradient grad_dims: list storing number of parameters at each layer

models.gem.project2cone2(gradient, memories, margin=0.5, eps=0.001)

Solves the GEM dual QP described in the paper given a proposed gradient “gradient”, and a memory of task gradients “memories”. Overwrites “gradient” with the final projected update.

input: gradient, p-vector input: memories, (t * p)-vector output: x, p-vector

models.gem.store_grad(params, grads, grad_dims)

This stores parameter gradients of past tasks. pp: parameters grads: gradients grad_dims: list with number of parameters per layers