dniku/rl-attention

Issues

• Visualize attention in callback

  #17 opened 5 years ago by dniku
  1

• Make sure that vanilla architecture earns high reward on multiple envs

  #12 opened 5 years ago by dniku
  1

• Export trained models, evaluation results and training stats from Colab

  #10 opened 5 years ago by dniku
  2

• Implement entropy loss from paper

  #16 opened 5 years ago by dniku
  0

• Implement entropy loss

  #14 opened 5 years ago by dniku
  1

• Save metrics after training

  #7 opened 5 years ago by dniku
  2

• Review of attention model

  #13 opened 5 years ago by dniku
  1

• Fix non-determinism

  #2 opened 5 years ago by dniku
  1

• Implement attention model from LTIAA

  #3 opened 5 years ago by dniku
  0

• Ensure that agent earns high reward on Colab

  #11 opened 5 years ago by dniku
  1

• Implement visualization of attention map

  #5 opened 5 years ago by dniku
  0

• Make code runnable on Colab

  #4 opened 5 years ago by dniku
  0

• Specify network architecture in our code

  #8 opened 5 years ago by dniku
  1

• Implement skeleton based on stable-baselines

  #1 opened 5 years ago by dniku
  1