vwxyzjn/lm-human-preference-details

Issues

• Reward Shape

  #28 opened a year ago by QiyaoWei
  1

• right_to_left_pad optimization

  #26 opened a year ago by vwxyzjn
  6

• Creating a jax implementation

  #5 opened a year ago by vwxyzjn
  0

• Question about KL divergence computation

  #25 opened a year ago by Maxtoq
  3

• A question about `normalize_after`

  #8 opened a year ago by liutianlin0121
  3

• Add accelerate to poetry dependencies

  #9 opened a year ago by liutianlin0121
  2

• Questions about `left_padding_to_right_padding`

  #6 opened a year ago by liutianlin0121
  4