mtanghu/LEAP

Issues

• Encoder formulation of LEAP

  #11 opened 3 years ago by mtanghu
  0

• Constant memory LEAP

  #12 opened 3 years ago by mtanghu
  0

• Theory to show LEAP is Turing Complete?

  #13 opened 3 years ago by mtanghu
  0

• Latest embeddings techniques (Infinite Context Length?)

  #14 opened 3 years ago by mtanghu
  0

• RNN formulation

  #15 opened 3 years ago by mtanghu
  0

• Explainability support

  #16 opened 3 years ago by mtanghu
  0

• Equations not formatted correctly

  #20 opened 3 years ago by mtanghu
  0

• Running experiments/measurement on the Rescaled Dot-Product

  #17 opened 3 years ago by mtanghu
  1

• Current querying doesn't make much sense

  #18 opened 3 years ago by mtanghu
  1

• Improving Documentation

  #9 opened 3 years ago by mtanghu
  0

• NaN loss when using attention mask or not using attention mask but training for long time

  #6 opened 3 years ago by annasajkh
  6

• Rescaling Attention values experiments/different implementations

  #10 opened 3 years ago by mtanghu
  2

• Finish project renovation with new QKV formulation

  #8 opened 3 years ago by mtanghu
  1

• Move development section to issues or projects

  #4 opened 3 years ago by mtanghu
  2

• Readme citations should be bibtex rather than APA

  #3 opened 3 years ago by mtanghu
  0

• Can't seem to get all inline equations to render in README

  #1 opened 3 years ago by mtanghu
  0