ATTRICLIP#

Arguments#

Options

--num_promptint

Help: num_prompt

  • Default: 10

--text_promptint

Help: text_prompt

  • Default: 3

--freeze_clipint

Help: freeze_clip

  • Default: 1

DISCLAIMER: AttriCLIP does not reproduce the results in the paper (https://arxiv.org/pdf/2305.11488). Unfortunately, the original implementation (https://github.com/bhrqw/AttriCLIP) did not reproduced the results either and is no longer available. This is a known issue (see https://github.com/bhrqw/SADA/issues/3).

This implementation is based on that code and on the information provided in the paper.

Classes#

class models.attriclip.Attriclip(backbone, loss, args, transform, dataset=None)[source]#

Bases: ContinualModel

Continual Learning via Progressive Neural Networks.

COMPATIBILITY: List[str] = ['class-il', 'domain-il', 'task-il', 'general-continual']#
NAME: str = 'attriclip'#
begin_task(dataset)[source]#
forward(x)[source]#
static get_parser(parser)[source]#
Return type:

ArgumentParser

observe(inputs, labels, not_aug_inputs, epoch=0)[source]#