itsShnik/adaptively-finetuning-transformers

Issues

• To experiment with attention pooling

  #33 opened 4 years ago by itsShnik
  0

• Implement SpotTune Res

  #31 opened 4 years ago by itsShnik
  0

• Experiment with dropping layers.

  #27 opened 4 years ago by itsShnik
  1

• Global-K Variant

  #14 opened 4 years ago by itsShnik
  1

• Add other strategies in LXMERT

  #21 opened 4 years ago by itsShnik
  1

• Experiment with Standard + SpotTune on VLBERT

  #25 opened 4 years ago by itsShnik
  1

• Experiment with policy network inputs

  #26 opened 4 years ago by itsShnik
  0

• Add Visualization Material and Wandb Integration in LXMERT

  #20 opened 4 years ago by itsShnik
  1

• Implement SpotTune strategies on LXMERT

  #15 opened 4 years ago by itsShnik
  1

• Parallel Net in LXMERT

  #18 opened 4 years ago by itsShnik
  1

• Add Gumbel Softmax in LXMERT

  #19 opened 4 years ago by itsShnik
  1

• LXMERT Policy Net

  #16 opened 4 years ago by itsShnik
  1

• Add VQAcpV2 dataset

  #12 opened 4 years ago by itsShnik
  1

• Wandb Integration

  #3 opened 4 years ago by itsShnik
  1

• Visualization Material

  #8 opened 4 years ago by itsShnik
  1

• Other finetuning policies

  #4 opened 4 years ago by itsShnik
  1

• Create parallel blocks for each of the transformer encoder layers

  #2 opened 4 years ago by itsShnik
  1

• Create a small policy network

  #1 opened 4 years ago by itsShnik
  1